3 PDF

2005 AMA Winter Educators’ Conference
Marketing Theory and

Applications
Editors
Kathleen Seiders, Boston College
Glenn B. Voss, North Carolina State University
Track Chairs
Jeff Inman, University of Pittsburgh
Cheryl Nakata, University of Illinois at Chicago
K. Sivakumar, Lehigh University
Paul Bloom, University of North Carolina, Chapel Hill
Steve Hoeffler, University of North Carolina, Chapel Hill
Ronald C. Goodstein, Georgetown University
Brian Wansink, University of Illinois
David Henard, North Carolina State University
Mitzi Montoya-Weiss, North Carolina State University
Sandy Jap, Emory University
Rick Andrews, Louisiana State University
Danny Weathers, Louisiana State University
Peter Golder, New York University
Mike Ahearne, University of Connecticut
Eli Jones, University of Houston
Jagdip Singh, Case Western Reserve University
Deepak Sirdeshmukh, North Carolina State University
Linda Ferrell, University of Wyoming
O.C. Ferrell, Colorado State University
Rebecca Slotegraaf, Indiana University
Rosann Spiro, Indiana University
Volume 16
311 S. Wacker Drive • Chicago, Illinois 60606 • (312) 542 - 9000

© Copyright 2005, American Marketing Association
Printed in the United States of America
Publications Director: Francesca Van Gorp Cooley

Project Coordinator: Charles Chandler
Cover Design: Jeanne Nemcek
Typesetter: Marie Steinhoff
ISSN: 1054-0806
ISBN: 0-87757-314-X
All rights reserved. No part of the material protected by this copyright

notice may be reproduced or utilized in any form or by any means,
including photocopying and recording, or by any information storage
or retrieval system without the written permission of the American
Marketing Association.
Preface and Acknowledgements
The theme for the 2005 Winter Marketing Educators’ Conference is “Understanding Diverse and
Emerging Markets, Technologies, and Strategies.” The multicultural marketplace is the new mainstream and
sophisticated marketing technologies provide a better understanding of buyer behavior and the ability to reach
diverse segments more efficiently. This year’s conference includes a variety of innovative special sessions
presenting key thought leaders and paper sessions addressing research themes that are critical to marketing
scholars and educators.
Many dedicated individuals contributed to this conference. In particular, we acknowledge the outstanding
work of the Conference Track Chairs, whose stellar performances produced a truly distinguished program.
Reinforcing the role of the Winter Educator’s Conference as a premier research conference, the Track Chairs
worked to shape high-level special sessions and maintained a lofty bar for competitive paper sessions. The
Conference Track Chairs are as follows:
Consumer Behavior Jeff Inman, University of Pittsburgh

Global Marketing: Cheryl Nakata, University of Illinois – Chicago
K. Sivakumar, Lehigh University
Marketing and Society Paul Bloom, University of North Carolina – Chapel Hill
Steve Hoeffler. University of North Carolina – Chapel Hill
Brand Marketing and Communications Ronald C. Goodstein, Georgetown University
Brian Wansink, University of Illinois – Urbana-Champaign
Marketing, Technology, and Innovations David Henard, North Carolina State University
Mitzi Montoya-Weiss, North Carolina State University
Inter-Organizational Issues Sandy Jap, Emory University
Marketing Research Rick Andrews, Louisiana State University
Danny Weathers, Louisiana State University
Marketing Strategy Peter Golder, New York University
Selling and Sales Management Mike Ahearne, University of Houston
Eli Jones, University of Houston
Services & Relationship Marketing Jagdip Singh, Case Western Reserve University
Deepak Sirdeshmukh, North Carolina State University
Instructional Innovations Linda Ferrell, University of Wyoming
O.C. Ferrell, Colorado State University
Special Interest Group Rebecca Slotegraaf, Indiana University
Rosann Spiro, Indiana University
The program reflects the important contributions of the conference reviewers (listed on pp. iv-vii). All
those who submitted papers and special session proposals are to be thanked, as are members of the Academic
Council and AMA journal editors whose leadership in creating the featured Conference Special
Sessions was instrumental.
We are exceedingly grateful to Denise Smart, AMA Academic Council President, for her advice and
support throughout the conference planning process. Also, we acknowledge the significant implementation
efforts of the AMA staff: Nicole Morris, our Program Manager; Charles Chandler; Francesca Van Gorp
Cooley; and Pat Goodrich. Thanks also to Marie Steinhoff for typesetting the proceedings.
With deep appreciation, we thank all contributors for making this a conference that is worthy of our field.
Kathleen Seiders Glenn B. Voss

Boston College North Carolina State University
iii
2005 AMA Winter Educators’ Conference List of Reviewers
A Mary J. Bitner, Arizona State

University
Bruce Clark, Northeastern
University
Manoj Agarwal, Binghamton Carolyn Bonifield, University of Lucette B. Comer, Purdue
University Vermont University
Sanjeev Agarwal, Iowa State Greg Bonner, Villanova University Robert Cosenza, University of
University Doug Bowman, Emory University Mississippi
Shanita Akintonde, Columbia Kevin Bradford, University of Nicole E. Coviello, University of
College Notre Dame Auckland
Joe Alba, University of Florida Michael Brady, Florida State Bill Cron, Texas Christian
Terri Albert, University of Hartford University University
Dana Alden, University of Hawaii Steve Brown, University of Lawrence Cunningham, University
Alan Andreasen, Georgetown Houston of Colorado – Denver
University Tom Brown, Oklahoma State Kerry P. Curtis, Golden Gate
Kersi Antia, University of Western University University
Ontario Eileen L. Bridges, Kent State
John Antil, University of Delaware
Syed Anwar, Western Texas A&M
University
Nancy Buchan, University of
D
University Wisconsin – Madison Peter Dacin, Queens University
Neeraj Arora, University of Margo Buchanan-Oliver, Kofi Dadzie, Georgia State
Wisconsin University of Auckland University
Zeynep Arsel, University of Cheryl Buff, Siena College Robert Dahlstrom, University of
Wisconsin Al Burns, Louisiana State Kentucky
Kwaku Atuahene-Gima, City University Rajiv Dant, Clarkson University
University (Hong Kong) Jim Burroughs, University of Prakash Das, Queen’s University
Craig Atwater, Temple University Virginia Kirk Davidson, Mount St. Mary’s
Anne-Francoise Audrain, Rouen Ronald Bush, University of West College
School of Management Florida Marion Debruyne, Emory
Jon Austin, Cedarville University University
Catherine Axinn, Ohio University David Dekker, University of
C Nijmegen
Benedict Dellaert, Universiteit
B Susan Cadwallader, Texas A&M
University Maastricht
Julie A. Baker, University of Richard Caldarola, American Carol Demoranville, Northern
Texas – Arlington Intercontinental University Illinois University
Sridhar Balasubramanian, Meg Campbell, University of Kalpesh Desai, University of
University of North Carolina Colorado Buffalo
Soumava Bandyopadhyay, Lamar Les Carlson, Clemson University Sameer Deshpande, University of
University Forrest Carter, Michigan State Lethbridge
Yeqing Bao, University of University Andrea Dixon, University of
Alabama Goutam Challagalla, Georgia Cincinatti
Fleura Bardhi, University of Institute of Technology Susan Douglas, New York
Nebraska – Lincoln Fiona Chan, University of Hong University
Michael Basil, University of Kong Kent Drummond, University of
Lethbridge Rajesh Chandy, University of Wyoming
Boris Becker, Oregon State Minnesota Sean Dwyer, Louisiana Tech
University Joseph Chang, University of
Joseph Bellizzi, Arizona State
University West
Regina E
Amar Cheema, Washington Ike Ekeledo, Northeastern Illinois
Neeraj Bharadwaj, University of University
Texas University
John Cherry, Southeast Missouri Anita Elberse, Harvard University
Sundar Bharadwaj, Emory State University
University Reham Eltantawy, Florida State
Lawrence B. Chonko, Baylor University
University
iv
Sunil Erevelles, University of Xin He, University of Pittsburgh Harold Kassarjian, University of
North Carolina – Charlotte Tim Heath, Miami University California – Los Angeles
Sevgin Eroglu, Georgia State Geraldine R. Henderson, Carol Kaufman-Scarborough,
University University of Texas – Austin Rutgers University
Neil Herndon, University of Patrick Kaufmann, Boston
F Missouri – Columbia
Louise Heslop, Carleton University
University
Erdener Kaynak, Penn State
Lawrence Feick, University of
Janet Hoek, Massey University University
Pittsburgh
Susan Hogan, Emory University Jeremy Kees, University of
Reto Felix, University of
Hartmut Holzmueller, University of Arkansas
Monterrey
Dortmund Stephen Keysuk Kim, Oregon State
Karen Fernandez, University of
Lee Kam Hon, Chinese University University
Auckland
of Hong Kong Adwait Khare, University of
Rosie Ferraro, Duke University
Heather Honea, San Diego State Houston
Linda Ferrell, University of
Mark Houston, University of Jaehwan Kim, University of
Wyoming
Missouri Colorado
Leslie Fine, Ohio State University
Carol Howard, Oklahoma City Noreen Klein, Virginia Tech
Bob Fisher, Western Ontario
University University
Carlos Flavian, Universidad de
Frederick Hoyt, Illinois Wesleyan Thomas Klein, University of
Zaragoza
University Toledo
Thomas Foscht, University of Graz
Yili Huang, University of Illinois – Yu Jun Koernig, University of
Ellen Foxman, Bentley College
Chicago Illinois – Chicago
Dan Freeman, University of
John Hulland, University of Praveen Kopalle, Dartmouth
Delaware
Pittsburgh College
Frank Fu, University of Houston
Mike Hutt, Arizona State Scott Koslow, University of
University Waikato
G Bruce Hutton, University of Robert V. Kozinets, Northwestern
Larry Garber, Appalachian State Denver University
University Michael Hyman, New Mexico Ram Krishnan, California
Nitika Garg, University of State University Polytechnic State University
Mississippi Hyojkin Kwak, Drexel University
Mrinal Ghosh, University of I
Michigan
Subin Im, San Francisco State L
Joan Giese, Washington State
University Dan Ladik, Suffolk University
University
Scott Inks, Ball State University Michel Laroche, Concordia
Mary Gilly, University of
Koert van Ittersum, Georgia University
California – Irvine
Institute of Technology Debra Laverie, Texas Tech
Charlotte Greig, Golden Gate
Vish Iyer, University of Northern University
University
Colorado Ruby Lee, University of Nevada
Andy Grein, City University of
James H. Leigh, Texas A&M
New York
Raj Grewal, Penn State University J University
Tom Leigh, University of Georgia
Stephen J. Grove, Clemson Cheryl Jarvis, Arizona State Katherine Lemon, Boston College
University University Patrick Lentz, University of
Julie Guidry, Texas A&M Rama Jayanti, Cleveland State Dortmund
University University Steven V. LeShay, Wilmington
Devon Johnson, Northeastern College
H University
Chris Joiner, George Mason
Lewis Lim, Indiana University
Charles Lindsey, Indiana
John Hadjimarcou, University of
Marilyn Jones, Bond University Susan Lloyd, American University
Texas – El Paso
Som Hanvanich, Xavier University Terry Loe, Kennesaw State
Angela Hausman, University of K University
Brian Lofman, Ramapo College of
Texas – Pan American Rajiv Kashyap, William Peterson New Jersey
University
v
Ritu Lohtia, Georgia State
University
Venkatapparao Mummalaneni,
Virginia State University
Q
Peggy Sue Loroz, Gonzaga Patrick Murphy, University of William Qualls, University of
University Notre Dame Illinois
Xueming Luo, University of Jim L. Murrow, Drury University
Texas – Arlington R
N Priya Raghubir, University of
M Kent Nakamoto, Virginia Tech California – Berkeley
Doug MacLachlan, University of Om Narasimhan, University of Priyali Rajagopal, Ohio State
Washington Minnesota University
Humaira Mahi, Michigan State Ed Nijssen, University of Nijmegen Deva Rangarajan, Vlerick-Leuven
University Rakesh K. Niraj, University of Gent, Belgium
Alan Malter, University of Arizona South California Peter Rea, Baldwin-Wallace
Elliot Maltz, Willamette University Thomas G. Noordewier, University College
Detelina Marinova, Case Western of Vermont Kristy Reynolds, Louisiana State
Reserve University Patricia Norberg, Quinnipiac University
James Maskulka, Lehigh University Su Bom Rhee, Santa Clara
Greg Rich, Bowling Green State
Charlotte H. Mason, University of
North Carolina
O University
Charla K. Mathwick, Portland State Gillian Oakenfull, Miami Keith Richards, University of
University University Houston
Shashi Matta, University of Matthew P. O’Brien, University of Nora J. Rifon, Michigan State
Southern California Arizona University
Debbie McAlister, Texas State Elie Ofek, Harvard University Ed Rigdon, Georgia State
University – San Marcos Shintaro Okazaki, Universidad University
Michael S. McCarthy, Miami Autónoma de Madrid Aric Rindfleisch, University of
University Ulrich Orth, Oregon State Wisconsin
David Mick, University of Virginia University Deborah E. Rosen, University of
Sam Min, University of South David Ortinau, University of South Rhode Island
Dakota Florida Bill Ross, Penn State University
Vikas Mittal, University of Abhijit Roy, Loyola College
Pittsburgh P Don Roy, Middle Tennessee State
University
Risto Moisio, University of Photis M. Panayides, Cyprus
Nebraska – Lincoln Subroto Roy, University of New
International Institute of Haven
Michael Mokwa, Arizona State Management
University Salvador Ruiz, Universidad de
Veronika Papyrina, University of Murcia
Sangkil Moon, North Carolina Western Ontario
State University Leonard Parsons, Georgia Institute
Bill Moore, University of Utah
David Moore, University of
of Technology S
Vanessa Patrick, University of Joel Saegert, University of Texas
Michigan Southern California
Elizabeth Moore, University of at San Antonio
Joann Peck, University of Jeff Sager, University of North
Notre Dame Wisconsin
Neil Morgan, University of North Texas
Antony Peloso, Arizona State Kare Sandvik, Buskerud University
Carolina University
Rob Morgan, University of Nicola Sauer, University of
Robert Peterson, University of Mannheim
Alabama Texas – Austin
David Mothersbaugh, University of Paul Sauer, Canisius College
Jeffrey Podoshen, Prado Saeed Samiee, University of Tulsa
Alabama Jaideep Prabhu, Imperial College
Susan Mudambi, Temple Sanjit Sengupta, San Francisco
Devashish Pujari, McMaster State University
Paulo H. Muller, University Reshma Shah, Emory University
Ellen Pullins, University of Toledo Tim Silk, University of Florida
Federal do Paraná
vi
Jim Simpson, University of Janet Tinoco, University of Central Rebecca M. Wells, University of
Alabama – Huntsville Florida Dayton
Alina Sorescu, Texas A&M Julie Toner (Schrader), Bellarmine Patricia West, Ohio State
University University University
Jelena Spanjol, Texas A&M Chris White, Michigan State
University
Richard Spreng, Michigan State
U University
Tiffany Barnett White, University
University David Urban, Virginia of Illinois at Urbana
Srinivas Sridharan, University of Commonwealth University Champaign
Western Ontario John Wong, Iowa State University
Raji Srinivasan, University of
Texas – Austin
V Arch Woodside, Boston College
Rajiv Vaidyanathan, University of
Srini Srinivasan, Drexel University
James A. Stephens, Emporia State Minnesota Y
University Raj Venkatesan, University of Attila Yaprak, Wayne State
Rodney L. Stump, Morgan State Connecticut University
University Alladi Venkatesh, University of Jun Ye, Case Western Reserve
Ursula Y. Sullivan, University of California – Irvine University
Illinois Peter Verhoef, Erasmus University Virginia Yonkers, Siena College
Tracy Suter, Oklahoma State Rotterdam Boonghee Yoo, Hofstra University
University Zannie Voss, Duke University Gergana Yordanova, University of
Scott D. Swain, Boston University Pittsburgh
W Cliff Young, University of
Colorado
T Frank Wadsworth, Indiana
University Southeast
Debu Talukdar, University of
Buffalo Kirk Wakefield, Baylor University Z
John F. Tanner, Jr, Baylor Wakiuru Wamwara-Mbugua, Gail Zank, Texas State
University Wright State University University – San Marcos
Kimberly A. Taylor, Florida Florian Wangenheim, University of Allen Zhang, University of Texas –
International University Dortmund San Antonio
Stephen S. Tax, University of Charles Weinberg, University of Zhu Zhen, Babson College
Victoria British Columbia George Zinkhan, University of
Georgia
vii
TABLE OF CONTENTS
PREFACE AND ACKNOWLEDGMENTS iii
LIST OF REVIEWERS iv
TABLE OF CONTENTS viii
COMPANY-CONSUMER AND CONSUMER-CONSUMER SOCIAL EFFECTS
The Impact of the Type, Frequency, and Quality of Customer Contact on Customer
Satisfaction
Marshall Rice 1
Mirror, Mirror, on the Wall, Am I What I Consume After All: A Framework for
Ethnicity Based Consumption, a Social Identity Perspective
Tracy R. Harmon 7
Cultural Influence on Word-of-Mouth Communication

Desmond Lam, Dick Mizerski, Alvin Lee 9
ORGANIZATIONAL IMPACT ON SERVICE DELIVERY
The Effects of Authenticity Rift on Firm Performance

Zannie Giraud Voss, Glenn B. Voss, Daniel M. Cable 11
Positioning in Service Firms: Model Development and Some Basic Normative

Guidelines
Charles Blankson, Stavros P. Kalafatis 13
Antecedents and Consequences of Role Clarity in Explaining Employee-Perceived

Service Quality in Call Centers
Avinandan Mukherjee, Neeru Malhotra 15
INSTRUCTIONAL INNOVATION IN MARKETING EDUCATION
The Status of Cross-Functional Education in Undergraduate Marketing Curricula

Within Management Education
Victoria L. Crittenden, Elizabeth J. Wilson, Cameron Duffy 18
On Different Teaching Pedagogies: What Happens to Your Course Evaluations?

Alma Mintu-Wimsatt, Kendra Ingram, Mary Anne Milward, Courtney Russ 20
An Assessment of a Consumer Behavior Multiple-Choice Question Taxonomy

John R. Dickinson 22
PRODUCT PLACEMENTS EFFECTS
A Comparison of Consumers’ Responses to Traditional Advertising and Product

Placement Strategies: Implications for Advertisers
Terry Daugherty, Harsha Gangadharbatla 24
New Brand Worlds: A Comparison of College Student Attitudes Toward Brand

Placements in Four Media
Yongjun Sung, Federico de Gregorio 26
viii
Brands in Action: The Role of Brand Placements in Building Consumer-Brand
Identification
Andrew T. Stephen, Leonard V. Coote 28
PROCESS AND BEHAVIOR
The Marketing Cocompetition Process and Strategic Alliance Instability: A System

Dynamics Model
Anna Shaojie Cui, Roger J. Calantone 30
Focal Supplier Opportunism in Retailer Category Management

Neil A. Morgan, Anna Kaleka, Richard A. Gooner 32
AFFECT-RELATED CONSUMER RESEARCH
“No Greater Satisfaction Than to Vindicate Expectation” How Affective Expectations

Shape Consumption Experience
Andrew K.C. Wong 34
Customer Delight: An Attempt to Comprehend the Dimensions That Compose the

Construct and its Behavioral Consequences
Stefânia Ordovás de Almeida, Walter Meucci Nique 36
Towards a Conceptual Framework of E-Confusion

Vincent-Wayne Mitchell, Gianfranco Walsh 44
UNDERSTANDING CUSTOMER-FIRM EXCHANGE PROCESSES
Differential Impacts of CRM on Consumer Responses: Toward an Integrative

Framework
Frederick Hong-Kit Yim 46
Exploring the Phenomenon of Buyer-Seller Mismatches in Business-to-Business

Relationships
Christopher P. Blocker 47
Consumer Trust Norms in Multi-Channel Firms: The Role of Trust in Technology

and Firm Commitment to Privacy Protection in Technology-Based Service Delivery
Devon Johnson 58
EXPLORATIONS OF CULTURE AND ETHNOCENTRISM DIMENSIONS
Validation and Application of a Bi-Dimensional Long-Term Orientation Scale

William O. Bearden, R. Bruce Money, Jennifer L. Nevins 59
Consumer Socialization of Third Culture Kids in a Cosmopolitan City

Alfred Y. Sit, Haksin Chan 61
Consumer Ethnocentrism in the German Market

Heiner Evanschitzky, Florian V. Wangenheim 63
NEW PRODUCT CONSUMER DECISION PROCESSES
Identifying Information Search Patterns in a Web-Based Environment: Development

of a Search Pattern Index
Morris K. George, Girish N. Punj 65
ix
Perceived Risk and Consumer Innovativeness Hierarchy: An Empirical Study of
Resistance to High Technology Product Adoption
Tanawat Hirunyawipada, Mohammadali Zolfagharian 73
Perceived Entitativity as a Moderator of Family Brand Evaluations

Joseph W. Chang, Yung-Chien Lou 75
MARKET-BASED DRIVERS OF FIRM PERFORMANCE
The Performance Implications of Synergistic Knowledge Resource Effects in

Differing Environmental Conditions
David A. Griffith, Stephanie M. Noble, Qimei Chen 77
The Impact of Market Characteristics on Order-of-Brand Entry Strategy: An Empirical

Study
Danielle A. Chmielewski, Bryan A. Lukas, Robert E. Widing II 79
Organizational Culture Antecedents of Market-Driven Positional Advantage and

Organizational Performance Consequences
Artur Baldauf, David W. Cravens, Christian Bischof 87
DEFENSIVE AND PROACTIVE RESPONSES TO SERVICE FAILURE AND

COMPLAINTS
See No Evil, Hear No Evil, Speak No Evil: A Study of Defensive Organizational Behavior
Towards Customer Complaints
Christian Homburg, Andreas Fürst 89
The Rest of the Iceberg: An Examination of Noncomplaining Service Customers

Clay M. Voorhees, Michael K. Brady, David M. Horowitz 91
From Empathy to Forgiveness: A Prosocial Perspective in Service Failure and Recovery

Research
Felix T. Tang 92
STRATEGIC CONCERNS IN GLOBAL MILIEUS
International Marketing Alliance Dynamics: Empirical Findings from the

Pharmaceutical Industry
Sengun Yeniyurt, Janell D. Townsend, Erin Cavusgil 94
A Cross-National Study of Consumer-Firm Exchange Relationships Within the

Context of Market Milieus
Patrick Lentz, Deepak Sirdeshmukh, Ed Nijssen, Hartmut H. Holzmüller,
Jagdip Singh 96
CONVERGENT COMMUNICATIONS: MOVING BEYOND ADVERTISING
An Examination of IMC at the Tactical Level: Differences Across Time and Product
Type
Stephen J. Grove, Les Carlson, Michael J. Dorsch, Christopher D. Hopkins 98
Assessing the Effects of In-School Point of Purchase and Sampling on the Choice of
a Healthy Food Option
Dafina Rexha, Katherine Mizerski, Richard Mizerski 99
x
Testing Why Adults Purchase Fast Food Cartoon Character Toy Premiums
Claire Lambert, Richard Mizerski 101
COMPETITIVE INTERACTION AND STRATEGIC DECISIONS
Multimarket Contact and the Moderating Role of Dominant Local Players: A Conceptual
Overview
Sweta Chaturvedi Thota 103
The Nature of Co-Opetition: Literature Review and Propositions

Pilsik Choi 105
Technology Versus People: Two Schools of Thought on Pricing Capability Development

Lewis K.S. Lim, Rebecca J. Slotegraaf, Rockney G. Walters 107
GOVERNANCE
Dimensions and Outcomes of Relational Exchange in a Business-to-Business Context:

A Meta-Analysis
Mohammadali Zolfagharian, Rajasree K. Rajamma 109
Entry Mode and Level of Equity: A Simultaneous Examination of Foreign Direct

Investment Governance
Sudha Mani, Kersi D. Antia, Aric Rindfleisch 111
Towards an Understanding of the Governance of Complex Networks of Relationships:

Uncertainty, Governance System Choices, and Performance Outcomes
Andrew T. Stephen, Leonard V. Coote 112
INNOVATIVENESS AND LEARNING
Market Driving Relationship Marketing for Radical Innovations

Helder J. Sebastiao 114
An Investigation of Perceptual Factors Influencing Consumer’s Intention to Adopt

Radical Versus Incremental New Products
Audhesh K. Paswan, Lisa C. Troy 117
It’s All about Learning: What Firms Can Learn from Consumer Pioneers
Yun Ye 118
SPECIAL SESSION: CULTURE, COMPETITION, AND CONSUMERS’

REACTIONS TO BRANDS
Will Consumers Prefer Global or Local Brands? The Role of Identity Accessibility in
Consumer Preference for Global Versus Local Brands
Yinlong Zhang, Lawrence Feick, Vikas Mittal 125
CONSUMER JUDGMENTS AND MOTIVATIONS IN SERVICES
Why Do Conference Goers Return? A Model of Intentions to Attend and Recommend

Annie H. Liu, Mark P. Leach, Robert D. Winsor 126
Influence of Other Customers: A Scale Development

E. Deanne Brocato, Susan B. Kleiser 128
xi
ENCOURAGING YOUNG PEOPLE TO BEHAVE BETTER
Beyond Just Being There: An Examination of the Impact of Attitudes, Materialism, and
Self-Esteem on the Quality of Helping Behavior in Youth Volunteers
Elten Briggs, Tim Landry, Charles Wood, Todd Arnold 130
A Cross-Cultural Examination of the Relationship Between Materialism and Individual

Values
William Kilbourne, Marko Grünhagen, Janice Foley 132
M.E.A.L. Time: How Nutritional Disclosure Affects Gender Evaluations of Fast Food
Menu Items
Kenneth W. Bates, Kyle A. Huggins 134
LEARNING FROM CUSTOMERS AND EDUCATING CUSTOMERS IN THE

NEW DOMINANT LOGIC OF MARKETING
Unlocking Value Through Customer Education

Thorsten Hennig-Thurau, Peter C. Honebein, Benoit Aubert 136
Organizational Learning and Dynamic Marketing Capabilities: Implications for

Organizational Performance
Linda M. Foley, Douglas W. Vorhies, Victoria D. Bush 138
INNOVATION ADOPTION
Adopting RFID Technology: Does the Manager’s Attitude Matter?

Gilbert N. Nyaga, Roger J. Calantone, Thomas J. Page 140
A Synergistic Model for New Product Success

Russell Adams 147
Income Elasticity of Household’s Demand for Communication and its Products: Global
Measurement and its Marketing Implications
Min Lu, Yanbin Tu 148
CONTEMPORARY PRICE-RELATED CONSUMER RESEARCH
Computational Estimation of Partioned Prices: Another Heuristic Moves into the

Marketing Neighborhood
William J. Jones, Devon S. DelVecchio, Terry L. Childers 154
The Effects of Magnitude Representation Encoding Interference and Order of Price

Exposure in Comparative Price Advertising
Keith S. Coulter, Robin A. Coulter 155
Influences on What Consumers Know and What They Think They Know Regarding
the Persuasive Aspects of Pricing-Related Selling Tactics
Jay P. Carlson, William O. Bearden, David M. Hardesty 157
MEASURE ADVERTISING SUCCESS
Creativity in Advertising: Purchase Intent and Brand Attitude Effects

Brian D. Till, Daniel W. Baack 159
xii
Where’s the Affect? An Investigation of the Effect of Three Advertising Scales on
Attitude to the Ad
Arjun Chaudhuri 161
Marketing Communication and Company Brand Attitude

Marc Weinberger, Dale Taoping Tzeng, Paul Bottomley, Harlan Spotts 168
RELATIONSHIP STRATEGIES: INTRA-FIRM, INTER-FIRM, AND SOCIETAL
Organizational Antecedents to and Outcomes of Marketing Strategy Development Styles:

A Contingency Model
J. Chris White, Jeffrey S. Conant, Raj Echambadi 177
Coordinating Marketing and Sales: Exploration of a Neglected Interface

Christian Homburg, Ove Jensen 179
THE ROLES OF CROSS FUNCTIONAL TEAMS AND PRODUCT QUALITY

IN INNOVATION
The Effect of Interactional Justice on the Performance of Cross-Functional Product

Development Teams
Tianjiao Qiu, Deborah Rupp, William Qualls 181
Cross-Functional Integration and New Product Performance: A Meta-Analysis

Tanawat Hirunyawipada, Archna Vahie 183
The Contingent Effect of Product Quality on New Product Performance: A Conceptual

Model
Kwaku Atuahene-Gima, Gloria Barczak 185
CONSUMPTION OF SERVICES
Consumer Experience of Social Power During Service Consumption: An Exploratory

Study
Kalyani Menon, Harvir Bansal 187
What Drives Customer Success? Consumer Perceptions of Enabling and Restraining

Forces Associated with Performing Consumption Tasks
Pete C. Honebein 194
The Effect of Event Valence on Wait Management Strategies

Elizabeth G. Miller, Barbara E. Kahn, Mary Frances Luce 196
MARKETING RESEARCH: NEW PERSPECTIVES ON DATA ANALYSIS
Seemingly Unrelated Regression: An Alternative to Traditional Bridging in Conjoint

Analysis
Niels J. Blunch 198
Within-Informant Bias in Marketing Research

James R. Brown, Anjala S. Krishen, Pushkin Kachroo, Chekitan S. Dev 200
Building Formative Construct Measures: The Example of Corporate Reputation

Sabrina Helm 202
xiii
STRATEGIC ISSUES IN ELECTRONIC MARKETING
Effects of Online Store Attributes on Customer Satisfaction and Loyalty

Miao Zhao, Ruby Roy Dholakia 204
Organizational Factors Related to Effective Customer Information Systems Practices

Debra Zahay 206
PUBLIC POLICY MEETS CONSUMER RESEARCH
Marketing Considerations in Weight Control: Preliminary Findings

Angela Hausman 207
Product Involvement and Place Attachment: Insights from the Environmental

Psychology Literature
Merlyn A. Griffiths 209
Emergency Contraception: Expectations of Product Need and Use

Andrew M. Parker, Melanie A. Gold 216
CUSTOMER SATISFACTION AND LOYALTY DYNAMICS
Investigating Drivers of Customer Defection: A Relative Weight Approach

Thomas Hollmann, Cheryl Burke Jarvis 218
Investigating the Moderators of the Customer Satisfaction-Loyalty Link: Evidence

from Retailing
Heiner Evanschitzky, Gianfranco Walsh 220
The Interplay of Cognition and Affect in the Formation of Customer Satisfaction:

A Dynamic Perspective
Christian Homburg, Nicole Koschate, Wayne D. Hoyer 222
BUYER AND ORGANIZATIONAL DYNAMICS IN THE GLOBAL SETTING
Marketing Six Sigma: Zero Defects in Intercultural Service Quality

Martin C. Reimann, Ulrich F. Luenemann 223
The Impact of Experiential Knowledge and Creativity on Performance of International

Project
Taewon Suh, Hongxin Zhao, Seung H. Kim, Mark J. Arnold, Mueun Bae 234
COPING WITH A CHANGING TECHNOLOGICAL AND ETHICAL CLIMATE
Change and the Marketing Organization

Kelly D. Martin, Jean L. Johnson 236
Privacy Concerns and Customers’ Willingness to Provide Information: A Review with

Implications for Future Research
Mona Srivastava, Robert Harmon 238
Shifts in Workplace Ethics: Opportunities for Conflict?

Paul L. Sauer, Paul Chao 240
xiv
IMPROVING SALESPERSON PERFORMANCE THROUGH NEW SKILLS AND
TECHNOLOGY
Interaction Between the Salesperson and Customer: A Framework for Improving the
Sales Outcome
Elizabeth Hemphill, Chris Dubelaar, Steven Goodman, Gus Geursen 242
EXTRINSIC REWARDS AND CONSUMER CHOICE
Coupons: The Inside Scoop

Somjit Barat 253
What Next? Explaining Repurchase Decisions After Joining a Loyalty Program

Shirley Y. Cheng, Jessica Y. Kwong 254
Not All Deals Are Created Equal: Two Different Roles of Sales Promotion
Dongwoo Shin, James H. Leigh 256
ONLINE AUCTIONS, ADVERTISING, AND RESEARCH
The Effects of Reserve Prices on Bidding Behavior in Online Auctions

Marla Royne Stafford, Ashley Kilburn, Barbara B. Stern 258
Advertising Goes Mobile: Explaining Attitude Toward M-Advertising

Parissa Haghirian 260
FOSTERING SERVICE RELATIONSHIPS
Regulatory Focus and Relationship Marketing Success

Maria Sääksjärvi, Johanna Gummerus 262
Predicting Usage Level and Upgrading Behavior of Service Customers: A Model for
Lifetime Value Estimation at Early Relationship Stages
Florian V. Wangenheim 268
INTERNATIONAL ADVERTISING AND BRAND POSITIONING ISSUES
The Impact of Perceived Language Status on Product and Service Quality Expectations
Melissa Maier Bishop 270
Effects of Positioning a Foreign Brand as a Domestic Brand in Countries with

Developed (U.S.) Versus Transitioning (Romania) Market Economies
Lada V. Kurpis, Simona Stan, Carmen Barb 272
ISSUES IN MANAGING THE BRAND
Can Self-Affirmation Reduce Prejudice Expression Toward Stereotyped Brands?

Huimin Xu 279
Self and Brand Image Congruence: Driving Consumer Value

Adam Marquardt 286
The Spillover Effects of Product-Harm Crises in a Brand Portfolio

Jing Lei, Niraj Dawar, Jos Lemmink 288
xv
CREATING SUCCESS WITH SOFT ASSETS
Market Driven Intangibles and Sustainable Performance Advantages

Matti Tuominen, Sheelagh Matear, Sami Kajalo, Saara Hyvönen, Arto Rajala,
Kristian Möller, Gordon E. Greenley, Graham J. Hooley 290
Understanding Creative Campaign Implementation: An Investigation of its Antecedents

Atlanta L. Stoyle, Leonard V. Coote 292
RELATIONSHIP STRUCTURE
Value-Based Differentiation in Business Relationships: Gaining and Maintaining Key

Supplier Status
Wolfgang Ulaga, Andreas Eggert 294
Affect and Conation in Business-to-Business Relationships: An Empirical Analysis of

Loyalty Lifecycle Sequence
Chad Ruel Allred 296
The Effect of Total and Asymmetric Specific Asset Investment on Supplier-Buyer

Relationship: A Structural Model
Taewon Suh, Henry Yu Xie, Ik-Whan G. Kwon 298
AUTHOR INDEX 300
xvi
THE IMPACT OF THE TYPE, FREQUENCY, AND QUALITY OF
CUSTOMER CONTACT ON CUSTOMER SATISFACTION
Marshall Rice, York University, Toronto
ABSTRACT satisfaction. For example, a customer’s level of satisfac-

tion with a company (and it’s products and services) is
This paper presents an investigation of the impact of likely impacted by such factors as interaction with the
81 distinct types of customer contact on customer satisfac- corporate Web site, exposure to marketing communica-
tion. Data from 8,836 respondents support the finding that tions, participation in corporate events and training, inter-
customer satisfaction is improved by increasing contact action with sales representatives and many other contacts.
points. In addition, certain types of contact and quality of
contact are shown to be important. This paper presents an exploratory investigation of
the impact of a larger number of customer contacts than is
INTRODUCTION found in the academic literature. Specifically, this paper
examines the impact of the type, frequency and quality of
Over the past 20 years, customer satisfaction has 81 distinct contact points on customer satisfaction. By
become a key concept in marketing (Harvey 1998). Aca- examining 81 different contact points, we hope to provide
demic research has focused on important customer satis- some preliminary insights into the following types of
faction issues that include such topics as the conceptual- managerial and research questions:
ization and measurement of satisfaction constructs, (Fou-
rier and Mick 1999; Smith 1999), the impact of satisfac- 1. Can a company improve overall satisfaction by in-
tion on customer retention (Rust and Zahorik 1993), creasing its frequency of contact with customers?
customer satisfaction and profitability (Gurau and Ranch-
hod 2002), loyalty and customer satisfaction (Gronholdt, 2. Are certain types of customer contact more valuable
Martensen, and Kristensen 2000), the role of value in than others in improving overall satisfaction with a
customer satisfaction (Day 2002) to name a few. Al- company?
though the subject of customer satisfaction has been
extensively studied, research shows mixed findings and 3. Does frequency of contact or quality of contact have
complex relationships between the antecedents and out- a stronger relationship with overall satisfaction with
comes of business having more versus less-satisfied cus- a company?
tomers (Szymanski and Henard 2001).
RESEARCH DESIGN AND
In addition, the impact of customer contact has been DATA COLLECTION
extensively examined in the CRM literature (Winer 2001).
Research has shown that regular contact with customers For this study, the researchers analyzed data provided
can improve profits (Reinartz and Kumar 2000), decrease by a large company in the technology sector (the name of
defection/increase retention (Verhoef 2003; Weinstein the company is not revealed to protect proprietary data).
2002) and increase the perceived relationship investment A total of 8,836 people were intercepted at the corporate
(De Wulf, Odekerken-Schroder, and Lacobucci 2001). Web site and participated in the research which was
Further, considerable research shows that increased cus- administered via an online survey. Respondents were
tomer contact by activities such as loyalty programs typical of those who come to the Web site (i.e., a mix of
improve acquisition, retention and customer development IT Professionals and Developers and General Users). As
(Stone, Bearman, Butscher, and Gilbert 2004), while an incentive to participate, respondents were able to enter
brand building contact improves loyalty (McAlexander, a contest to win a $100 gift certificate from Amazon.com.
Schouten, and Koening 2002). Data was collected via an Internet population for several
reasons. First, experience and usage of many of the
Research on customer satisfaction has generally fo- variables that we examined (i.e., subscription to Internet
cused on understanding the impact of a limited number of services, participation in online events etc.) are not com-
product or firm attributes. In particular, the research has mon phenomenon and it would be difficult to obtain these
tended to focus on the impact of service and product respondents via a simple random telephone sample. Sec-
quality on overall customer satisfaction. It is reasonable to ond, the instrument was quite lengthy which would have
assume, however, that many other contacts that an orga- made it prohibitive to attempt to administer the survey via
nization has with its customers can have an impact on traditional data collection methods.
American Marketing Association / Winter 2005 1

The 81 contact points that were examined in this Respondents also indicated their level of satisfaction
paper were arrived at through extensive discussions with with each type of contact using nine point scales (where 1
decision makers at the company that provided the data. In indicated that they were “not at all satisfied” and 9
the discussion, the principal researcher and corporate indicated that they were “very satisfied”). Respondents
decision makers began with an informal listing of possible were also asked to rate their overall satisfaction with the
contact variables that can have an impact on satisfaction company, its products, service and support organization
and that could be measured. This list was then judgmentally using the same nine point scale. This overall satisfaction
reduced to produce the final 81 contact points. question served as the basis (dependent variable) for
much of the analysis that is presented in this paper. To
Respondents answered a series of closed-ended ques- answer the research questions, analysis was performed
tions that asked them to identify which of 81 unique types using a combination of cross-tabulation (X2 analysis),
of contact they have had with the company. These 81 analysis of variance (ANOVA) and regression.
contact points were then categorized for analysis into nine
distinct types of customer contact as listed in Table 1. Research Hypothesis
Cronbach’s Alpha was used to test the reliability of the
scales that make up the nine individual types of contact. The following research hypothesis are investigated in this
All nine items displayed an Alpha of > .85 which is paper:
considered good (Nunnaly 1978) and indicates that the
individual contact points were measuring a unidimen- H1: Increases in the amount of customer contact points
sional construct within each group. will correlate positively with overall company satis-
faction.
TABLE 1
Classification of Types of Contact
Category of Type of Contact Number of Contacts Within This Group
1. Corporate Web sites visited 16 contact points
2. Received Information about the company/ 14 contact points: such as reading online reviews
the company Products about products, company Products seeing print ads,
meeting a sales representative, receiving direct mail,
etc.
3. Initiated Contact with the company 13 contact points: such as contacting the company by
telephone, e-mail communication, contacting support,
etc.
4. Subscribed to Newsletter(s) 11 contact points: a total of 11 newsletters were

offered and monitored
5. Participated in Training 7 contact points: such as taking an online course, using

training materials, etc.
6. Participated in Events/Community Activities 7 contact points: such as participating in an online

chat, attending a conference, etc.
7. Planned and Deployed corporate solutions 6 contact points: such as using company products,
upgrading a product/service, etc.
8. Subscribed to Services 5 contact points: a total of five services offered by the

company were available
9. Product/Service Purchase or Download 2 contact points: purchasing product/service or down-

loading product/service

H2: Certain types of customer contacts are more valuable was significantly different across the three categories of
(i.e., have more impact) than others in improving frequency of contact. Those respondents who had the
overall satisfaction. highest levels of contact with the company had a signifi-
cantly higher mean score on satisfaction with the com-
H3: Quality of contact will have a stronger relationship pany as compared to those respondents who had lower
with overall satisfaction than frequency of contact. levels of contact. Similarly, those respondents who had
medium levels of contact had significantly higher satis-
RESULTS AND DISCUSSION faction with the company as compared to those respon-
dents in the light contact category. The mean values of
H1: Increases in the amount of customer contact will overall satisfaction with the company (on a nine point
correlate positively with overall company satisfac- scale with nine indicating “very satisfied”) for each of
tion. these groups are shown in Table 3. A Duncan’s Multiple
Range test confirmed significant differences among the
Respondents were categorized into three levels of mean values shown in Table 3.
frequency of contact, based on the number of contact
points (out of a possible 81 total number of contacts) they This positive relationship between frequency of con-
had with the company. Table 2 details the definition of the tact and increased overall satisfaction with the company
three categories of frequency of contact: Light Contact, can also be seen in the percentage of respondents who
Medium Contact, and Heavy Contact. were classified as highly satisfied (HSAT is defined as
those respondents who indicated that their satisfaction
The frequency of contact into light (10 or fewer was either 8 or 9 on a 9 point scale) and those who are
contacts), medium (11 to 20 contacts) and heavy contact classified as “not satisfied” (NSAT is defined as those
(21+ contacts) groups was determined judgmentally by respondents who indicated that their satisfaction was 1, 2,
key decision makers at the company that provided the 3, or 4) on overall satisfaction with the company across the
data. Specifically, decision makers were asked how many three categories of frequency of contact. As seen in Table
contacts they believe constituted “light,” “medium,” and 4, as the level of contact with the company increases, the
“heavy” contact with their customers. percentage of respondents who fall into the HSAT cat-
egory goes up significantly, and the percentage of those in
Using analysis of variance (ANOVA), it was possible the NSAT category goes down accordingly (Chi-square =
to determine that overall satisfaction with the company 119.4, p. < .0001).
TABLE 2
Categories of Frequency of Contact
Frequency of Contact Definition % of Sample
Light Contact 10 or fewer contacts 37%
Medium Contact 11 to 20 contacts 34%
Heavy Contact 21+ contacts 29%
TABLE 3
Impact on Satisfaction based on Frequency of Contact
Frequency of Contact Satisfaction With the Company (Mean)
Light Contact 6.4
Medium Contact 6.7
Heavy Contact 7.1
(F = 78.61, Significance: < .0001)

H2: Certain types of customer contact will be more valu- who had that contact and those who didn’t, were com-
able (i.e., have more impact) than others in improving pared to determine if a significant difference existed.
overall satisfaction. Table 5 summarizes the change in the mean score of
overall satisfaction with the company as a result of each
Analysis of variance (ANOVA) was used to deter- type of customer contact.
mine which types of contact were associated with the
greatest change in overall satisfaction with the company. As Table 5 shows, the most valuable types of contact
For each type of contact, the mean scores on overall for the sample overall, based on change in the mean scores
satisfaction with the company between those respondents of overall satisfaction with the company are:
TABLE 4
HSAT and NSAT by Frequency of Contact
Frequency of Contact % HSAT % DSAT
Light Contact 36% 17%
Medium Contact 41% 14%
Heavy Contact 49% 10%
X2 shows significant differences between all groups: (X2 = 119.4 p. < .001)
TABLE 5
Change in Overall Satisfaction With the Company By Type of Contact
Satisfaction of Satisfaction of Change in

Those Those Not Satisfaction as
Contacted Contacted a Result of ANOVA
Type of Contact (Mean) (Mean) Contact Results
Received Info. About The Company 6.78 5.98 +0.80 F = 132.66

p < .0001
Visited Web sites 6.75 6.23 +0.52 F = 57.43

p < .0001
Product/Service Purchase/Download 6.78 6.28 +0.50 F = 80.19

p < .0001
Subscribed to Newsletter(s) 7.01 6.53 +0.48 F = 108.53

p < .001
Participated in Events/Communities 6.99 6.61 +0.38 F = 47.41

p < .0001
Subscribed to Services 6.97 6.64 +0.33 F = 27.67

p < .0001
Planned/ Deployed corporate solutions 6.76 6.56 +0.20 F = 18.40

p < .0001
Participated in Training 6.89 6.60 +0.29 F = 36.76

P < .0001
Initiated Contact with the company 6.72 6.50 +0.22 F = 10.95

P < .001

a. Received information about the company or the com- should it direct resources to trying to improve satisfaction
pany products with some key types of customer contact? In order to study
this question, the survey asked respondents to rate their
b. Visited Web sites satisfaction with contacts they had experienced on a nine-
point scale (1= not at all satisfied, 9 = very satisfied). A
c. Product or service purchase or download regression analysis was then preformed with overall sat-
isfaction with the company as the dependent variable and
It is important to note, however, that satisfaction was aggregate satisfaction measures for the categories of con-
higher in all categories if the respondent had been contact that had the largest number of contacts (received
tacted. information about the company, visited Web sites, prod-
uct or service download, planned or deployed solutions,
H3: Quality of contact will have a stronger relationship initiated contact with the company).
with overall satisfaction than frequency of contact.
The results of this regression clearly show that satis-
From a managerial viewpoint it would be useful to faction with key customer contact points (i.e., quality of
know whether frequency of contact or quality of contact contact) is more strongly related than frequency of contact
has a greater impact on customer satisfaction. That is, to overall satisfaction (R2 = .52, all variables significant
given limited resources should organizations direct those at the .0001 level). Complete regression results are shown
resources to trying to increase the number of contacts or in Table 6. Given these results, it suggests that organiza-
TABLE 6
Regression Results
Type of Contact Parameter Estimate Significance
Initiated Contact with the company .193 p < .0001
Received Info. about The company .282 p < .0001
Product/Service Purchase/Download .209 p < .0001
Planned/ Deployed Corporate solutions .172 p < .0001
Visited Web sites .110 p < .0001
Frequency of contact .132 p < .0001
tions should consider putting more resources into improv- overall company satisfaction. However, while the number
ing the quality of contact rather than increasing the total of contacts is clearly important more impact is seen by
number of contacts. increasing the quality of each contact.
CONCLUSION As this is a exploratory study with only one company,

it is suggested that future research focus on different
This paper is an exploratory study that attempted to companies and sectors in an attempt to see whether these
show the relationship between the amount and quality of patterns exist in other industries. In addition, future re-
customer contact and it’s impact on satisfaction. As indi- search should investigate whether different taxonomy/
cated in the paper, the research hypothesis were largely consumer groups for the same company respond in simi-
confirmed. Specifically, the data show that increases in lar ways to both frequency and type of consumer contacts.
the amount of contact points is correlated positively with

REFERENCES 28 (1), 9–16.
Reinartz, Werner J. and V. Kumar (2000), “On the Prof-
Day, Ellen (2002), “The Role of Value in Consumer itability of Long-Life Customers in a Noncontractual
Satisfaction,” Journal of Consumer Satisfaction, Dis- Setting: An Empirical Investigation and Implications
satisfaction, and Complaining Behavior, 15, 22–33. for Marketing,” Journal of Marketing, 64 (4), 176–
De Wulf, Kristof, Gaby Odekerken-Schroder, and Dawn 33.
Lacobucci (2001), “Investments in Consumer Rela- Rust, R.T. and A.J. Zahorik (1993), “Customer Satisfac-
tionships: A Cross-Country and Cross-Industry Ex- tion, Customer Retention, and Market Share,” Jour-
ploration,” Journal of Marketing, 65 (October), (4), nal of Retailing, 69 (2), 193–215.
33–51 Smith, A. (1999), “Some Problems When Adopting
Fournier, Susan and David Mick (1999), “Rediscovering Churchill’s Paradigm for the Development of Ser-
Satisfaction,” Journal of Marketing, 63, (October), vice Quality Measurement Scales,” Journal of Busi-
(4), 5–24. ness Research, 46, 109–20.
Gronholdt, Lars, Anne Martensen, and Kai Kristensen Stone, Merlin, David Bearman, Stephan A. Butscher, and
(2000), “The Relationship Between Customer Satis- David Gilbert (2004), “The Effect of Retail Customer
faction and Loyalty: Cross-Industry Differences,” Loyalty Schemes – Detailed Measurement or Trans-
Total Quality Management, 11, (July), (4–6), S509– forming Marketing,” Journal of Targeting, Mea-
S515. surement and Analysis for Marketing, 12 (3), 305–19
Gurau, Calin and Ashok Ranchhod (2002), “How to Szymanski, David M. and David H. Henard (2001),
Calculate the Value of a Customer–Measuring Cus- “Customer Satisfaction: A Meta-Analysis of the Em-
tomer Satisfaction: A Platform for Calculating, Pre- pirical Evidence,” Journal of the Academy of Market-
dicting and Increasing Customer Profitability,” Jour- ing Science, 29 (1), 16–36.
nal of Targeting, Measurement, and Analysis for Verhoef, Peter C. (2003), “Understanding the Effect of
Marketing, 10 (March), (3), 203–20. Customer Relationship Management Efforts on Cus-
Harvey, J. (1998), “Service Quality: A Tutorial,” Journal tomer Retention and Customer Share Development,”
of Operations Management, 16, 583– 97. Journal of Marketing, 67, (4), 30
McAlexander, James H., John W. Schouten, and Harold Weinstein, Art (2002), “Customer Specific Strategies –
F. Koening (2002), “Building Brand Community,” Customer Retention: A Usage Segmentation and
Journal of Marketing, 66, (January), 38–55. Customer Value Approach,” Journal of Targeting,
Nunnaly, J. (1978), Psychometric Theory. New York: Measurement and Analysis for Marketing, 10 (3),
McGraw-Hill 259–69.
Parasuraman, A. and D. Grewal (2000), “Serving Cus- Winer, Russell S. (2001), “A Framework for Customer
tomers and Consumers Effectively in the Twenty- Relationship Management,” California Management
First Century: A Conceptual Framework and Over- Review, (Summer), 89–105.
view,” Journal of the Academy of Marketing Science,
For further information contact

Marshall Rice
Marketing Area
Schulich School of Business
York University
4700 Keele Street
Toronto, Ontario
Canada M3J-1P3
Phone: 416.736.2100, Ext. 58241
E-Mail: mrice@schulich.yorku.ca

MIRROR, MIRROR, ON THE WALL, AM I WHAT I CONSUME AFTER
ALL: A FRAMEWORK FOR ETHNICITY BASED CONSUMPTION,
A SOCIAL IDENTITY PERSPECTIVE
Tracy R. Harmon, University of South Florida, Tampa
SUMMARY identity are: (1) human cognitive capacities are limited,

(2) individuals think about others in order to interact, and
The study of ethnicity as it relates to consumption is (3) individual struggle for understanding in order to
fairly recent in the study of consumer behavior and predict and control their outcomes. As a result individuals
marketing (Ogden et al. 2004). Within the marketing tend to categorize information about themselves, others,
literature, many studies have shown how culture signifi- objects, and situations before engaging in memory or
cantly impacts consumer’s perceptions and behavior in inferential processes, relevant to social interactions and
the global marketplace (McCracken 1986). Deshpande consumption.
et al. (1986) suggests that the concept of strength of ethnic
identification leads us to believe the existence of funda- Gilly and Penaloza (1999) state that marketers inves-
mental differences between members of a particular eth- tigate culture indirectly, resulting in references to dispar-
nic group. Herche and Balasubramanian (1994) discov- ate cultural groups and marketing practices, with less
ered that consumers who belonged to a specific ethnic attention to generalized adaptation processes. Their point
group were likely to display analogous shopping behav- highlights what is missing form the ethnic identity con-
iors. Thus, the focus of this research does not address sumption research, the underpinning foundation of ethnic
aggregate national cultures; rather the intent is to under- identity, which precedes consumption, which is addressed
stand the impact of strength of ethnic identity within a with the proposed framework. Therefore ethnicity-based
national boundary. This paper therefore seeks to explain consumption is more meaningful than the use of a sign, or
the impact of intra-national cultural differences in ethnic symbol that is consumed, it is a self-concept of image, that
identity within the specific context of the United States on cannot be expunged by the presence of a global market-
consumption behavior. place.
This framework does not seek to underscore the role Based upon the current research trends in consumer
of consumption in the construction of a socially based acculturation and the lack of fit of acculturation to micro-
ethnic identity, but examines the role of a socially based cultures within a nation, as well as within an ethnic group,
ethnic identity as an antecedent to the act of consumption. a framework is presented that expands the study of con-
Attempts to assimilate the ethnic self into a one-dimen- sumer buyer behavior. This lack of fit leads to an impor-
sional “melting-pot” ideology fail to address the complex- tant question relevant to marketing: Does the degree of
ities of personal and social identities among diverse pop- ethnic identity affect consumption practices? The frame-
ulations. Thus, the primary assumptions are (1) ethnic work suggests consumption practices are moderated by
identity precedes consumer consumption, (2) as the the degree of ethnic identity formation. The framework
strength of ethnic identity changes, so does consumption includes taking into the consideration the core self, iden-
practices, and (3) ethnicity based consumption does not tifying the strength of a consumer’s ethnic identity using
aid in the construction of a socially-based ethnic identity. Phinney’s (1989) key dimensions which are influenced
This research proposes that the strength of ethnic identi- by socialization processes, and determining whether pur-
fication displayed by consumers is the focal construct of chase decisions vary by degree of ethnic identification
interest during the investigation of ethnicity based con- during the consumption process. The proposed frame-
sumption practices. The proposed study will relate the work considers the collective nature of ethnic identifica-
four components of ethnicity (1) ethnic awareness, (2) tion, which considers the elements of the self, motive,
ethnic self-identification, (3) ethnic attitudes, and (4) ethnicity, and factors of socialization. These four compo-
ethnic behaviors (Phinney 1990) to consumption practic- nents capture the core of the individual while accommo-
es. We expect the differences to be moderated by the dating the situational and contextual factors of intersect-
degree of formation of an ethnically based social identity. ing identities. The proposed framework is divided into
two distinct halves, which capture the core ethnic self, and
Social cognition is one of the prevailing perspectives the apparent ethnic self.
in social psychology which provides the theoretical foun-
dations of conventional conceptions of identity. The as- After consideration of the core ethnic self and the
sumptions which underlie social cognitive theories of apparent ethnic self, which is influenced by social interac-

tions, the formation of ethnic identity emerges. The frame- frequently. Individuals who are in the foreclosed and
work indicates four defined levels of ethnic identity moratorium stages of development may development a
development, which are termed diffuse, foreclosed, mora- multicultural attitude of consumption, and may partici-
torium, and achieved ethnic identities. A diffused ethnic pate in more frequent culture swapping. Naturally the
identity has little or no exploration of ethnicity, and lacks levels of ethnic identification will vary within ethnic
a clear understanding of the issues. A foreclosed ethnic groups and among subcultures within these groups. How-
identity has little exploration but a clearer understanding ever these levels of identification will allow marketers to
of their ethnicity. Whereas, an ethnic identity in morato- better understand the consumption practices within a
rium denotes evidence of exploration of ethnicity, but particular ethnic group.
some confusion about the meaning of ethnic groups.
Lastly, an achieved ethnic identity is one that has been This framework helps to mitigate the premise of
explored, understood and accepted. contradictory consumption practices among those who
possess a subjectively, strong ethnic self-identity, by
It would be expected that during the diffuse stage, an providing a better understanding of the factors that help
individual would be the least likely to reference his or her shape a consumer’s ethnic identity. This will enhance the
ethnic identity during consumption. Whereas, an indi- ability to develop marketing communications appropriate
vidual who has achieved development of their ethnic for the multicultural market.
identity, would readily access their ethnic identity more
For further information contact:

Tracy R. Harmon
Marketing Department
University of South Florida
4202 East Fowler Avenue – BSN 3222
Tampa, FL 33620
Phone: 813.74.6184
E-Mail: tharmon@coba.usf.edu

CULTURAL INFLUENCE ON WORD-OF-MOUTH COMMUNICATION
Desmond Lam, University of Western Australia, Australia
Dick Mizerski, University of Western Australia, Australia
Alvin Lee, University of Western Australia, Australia
SUMMARY individuals. Culture, in particular, can have a strong

influence on one’s word-of-mouth behavior.
The power of word-of-mouth has considerable docu-
mentation since the 1960s. Word-of-mouth has been It is well documented that culture can have a strong
widely reported to be many times more influential than influence on consumers’ thoughts and actions. Hence,
information from prints, radio, and personal selling. De- culture can potentially have a significant influence on
spite the importance and influence of word-of-mouth, it consumers’ word-of-mouth behavior through its influ-
has remained one of the most neglected marketing areas. ence on individual values and group norms. In one of the
In fact, many companies are still struggling to develop most widely cited work, Hofstede (1980) found many
effective marketing programs that encourage consumer differences between the perceptions and the working
word-of-mouth communication. The authors of this cur- styles of individuals in 53 countries. Hofstede identified
rent study believe that a sound understanding of factors four basic dimensions of differences between national
influencing word-of-mouth such as those relating to cul- cultures, namely, individualism, masculinity, uncertainty
ture may help to create more proactive and targeted avoidance, and power distance. In this current study,
promotional programs toward stimulating consumer word- Hofstede’s dimensions were employed to examine indi-
of-mouth. vidual-level or within-culture differences.
Consumers may engage in word-of-mouth for a num- A survey was conducted on a convenience sample of
ber of intrinsic reasons. Word-of-mouth among consum- 228 total respondents from two universities in Australia.
ers is also affected by other external factors. The influence Each respondent was given a questionnaire comprising a
of culture appears to be the most important external number of items on issues relating to their cultural values
factors, particularly in the context of international market- and word-of-mouth behavior. These items were measured
ing. While there are many studies on the impact of culture on a 5-point Likert scale from 1 (strongly disagree) to 5
on marketing, very few examined the effect of culture on (strongly agree). The data was subjected to confirmatory
consumers’ word-of-mouth communication. This study factor analysis (CFA) and fitted to a structural equation
attempts to examine whether and how culture can influ- model with AMOS 6.0. The results gave strong support to
ence consumers’ word-of-mouth behavior. In particular, the influence of masculinity and power distance cultural
special attention will be given to distinguish word-of- dimensions on individuals’ word-of-mouth behavior. The
mouth communication with people of strong ties (defined influence of masculinity on out-group word-of-mouth
as in-group) from communication with people of weaker was significant and positive. Individuals high in mascu-
ties (defined as out-group). Consequently, this article linity are expected to be more assertive and aggressive in
provides both theoretical and empirical contribution to the their approach to communication. As such, they are more
word-of-mouth literature. Ultimately, it is the intention of likely to exchange product information with weaker ties
the authors to help companies to identify markets or such as out-groups. The results of this study have sup-
societies that may be more receptive to word-of-mouth ported this hypothesis. In addition, power distance has
marketing. significant positive influence on in-group word-of-mouth
and negative influence on out-group word-of-mouth. In-
The reasons why one engages in word-of-mouth have dividuals with high power distance are more likely to
been extensively researched for about 40 years. Word-of- engage in word-of-mouth within their in-groups than with
mouth activity has been shown to influence a variety of their out-groups. Those with low power distance are more
consumer conditions, from awareness, expectations, per- likely to feel less inhibited and, as such, will more likely
ceptions, attitudes, behavioral intentions to actual behav- engage in out-group word-of-mouth. These phenomena
iors. Past research found that consumers engaged in word- were observed in this current study. However, the study
of-mouth mainly for altruistic, product involvement, and did not find any evidence of influence of individualism on
self-enhancement reasons. The frequency and intensity of either in-group or out-group word-of-mouth. At the same
word-of-mouth may also depend on situations, service time, contrary to expectations, uncertainty avoidance ap-
quality, types of products and markets, social networks, peared to have little impact on both in-group and out-
social class, individual personality, and culture of the group word-of-mouth.

It is important to note that any extrapolation of the receptive to the use of word-of-mouth compared to others.
results must be made cautiously given that this research This study examined and differentiated those cultural
was conducted on a single country and was only repre- values/dimensions that may be more likely to encourage
sented by a sample of higher-education student popula- word-of-mouth from those that are less likely to do so.
tion. Future research will attempt to broaden the sample Both word-of-mouth and mass media influence the adop-
frame across several countries to improve generalization. tion of new products. The results from this research can
Also, the support of the influence of culture does not rule provide businesses with greater insight into using word-
out the explanations of other factors that are not covered of-mouth as a tool for marketing, especially in the interna-
in this study. Consumers’ word-of-mouth behavior may tional context, by examining critical factors that can affect
also change depending upon consumption contexts and word-of-mouth. For example, if companies understand
on the types of products they consume. Research into the factors that affect business referral behaviors, they can
these areas will likely yield a more comprehensive insight then try to create an environment to induce more customer
into the word-of-mouth construct. referrals. The results will also enable companies to esti-
mate the potential impact of product or company-related
Word-of-mouth is indeed a major force in the market- negative and positive word-of-mouth across different
place that should not be taken for granted. Individuals national cultures.
with certain cultural values/dimensions may be more

Desmond Lam
Faculty of Economics and Commerce
University of Western Australia
Social Sciences Building South
35 Stirling Highway
WA 6009, Australia
Phone: +61.8.6488.2890
FAX: +61.8.6488.1055
E-Mail: desmond@ecel.uwa.edu.au

THE EFFECTS OF AUTHENTICITY RIFT ON FIRM PERFORMANCE
Zannie Giraud Voss, Duke University, Durham
Glenn B. Voss, North Carolina State University, Raleigh
Daniel M. Cable, University of North Carolina at Chapel Hill, Chapel Hill
SUMMARY nonprofit professional theatre industry as the context. We

operationalized organizational identity using five organi-
This research examines whether internal disagree- zational value dimensions relevant to the nonprofit pro-
ments relating to organizational identity – authenticity fessional theatre industry (Voss, Cable, and Voss 2000),
rift – influence firm performance. An authentic organiza- and we collected measures from two respondents (i.e., the
tional identity represents a true, unique and singular managing director and marketing director) at each theatre.
statement of core organizational values and beliefs held We tested the hypotheses by conducting polynomial re-
by all organizational insiders and communicated to all gression analyses that modeled firm performance as a
stakeholders. Rift in the authenticity of organizational function of the main and quadratic effect of each value
identity represents clefts or differences in the core organi- dimension reported by each respondent and the interac-
zational values and beliefs held by organizational insid- tion between the managing director and marketing direc-
ers. These internal differences signal confusion regarding tor value dimension report. We used two distinct mea-
the organization’s core values and lead to divergent goals sures of firm performance: customer support, measured
and strategies. as the theatre’s total earned revenue and overall financial
performance, measured as net income.
The notion of multiple organizational identities has
existed since early explorations by Albert and Whetten Rift had a negative impact on firm performance for
(1985), who proposed that dual identity organizations four of the five values. These findings offer evidence that
possess both normative and utilitarian orientations, and multiple definitions of identity may hinder rather than
that organizations may begin with a single identity but help the firm. Surprisingly, rift had a positive effect on
acquire multiple identities over the course of time. Thus, firm performance for the market value dimension, defined
firms may have a constellation of values with some values as the organization’s commitment to customer satisfac-
dominating (Gioia, Schultz, and Corley 2000; Voss, Cable, tion. Firm performance was lower when both marketing
and Voss 2000). As long as the identity of the organiza- and managing directors reported that market values were
tion – however complex the set of core values and be- either extremely important or not at all important. This
liefs – is commonly held and understood, it is still singular suggests that “extreme” market theatres either ignore
and congruent. customers in pursuit of artistic experimentation or are too
focused on current customer preferences at the expense of
When authenticity rift exists, an organization pos- creating exciting new art. Balance appears to be achieved
sesses diverse and competing responses to the questions when the marketing director reported that market values
“Who are we?” and “What are our core values and were extremely important while the managing director
beliefs?” Rift occurs when multiple identities create con- reported that market values were not important. This
fusion throughout the organization as to the firm’s iden- result questions the wisdom of diffusing market values
tity or when different factions within the organization throughout an arts organization.
actively espouse different organizational priorities and
values (Albert and Whetten 1985; Golden-Biddle and While much research exists on organizational iden-
Rao 1997; Pratt and Foreman 2000). Conceptual argu- tity, this study is unique in its empirical examination of the
ments support alternative hypotheses for the relationship effect of internal discrepancies on firm performance. In
between authenticity rift and firm performance, including short, skillful management of a singular, clear, authentic
(1) no relationship, (2) a negative relationship, or (3) a organizational identity may be more advantageous than
positive relationship. multiple, distinctive identities presented to different stake-
holders. References available upon request.
To test these alternative hypotheses, we conducted a
longitudinal, two-wave empirical study using the U.S.

Zannie Giraud Voss
Duke University
Box 90680, 206 Bivins Building
Durham, NC 27708–0680
Phone: 919.660.3347
FAX: 919.684.8906
E-Mail: zannie@duke.edu

POSITIONING IN SERVICE FIRMS: MODEL DEVELOPMENT AND
SOME BASIC NORMATIVE GUIDELINES
Charles Blankson, Long Island University, New York
Stavros P. Kalafatis, Kingston University, United Kingdom
SUMMARY objectives (functional and symbolic). Furthermore, this

study finds that services are managed within two broad
Despite the growing activities and interest attached to life cycle stages (fortification and membership) out of
the concept of positioning and the fact that the subject is seven (i.e., primal, consolidation, latent, deposition, for-
considered to be one of the key elements of modern tification, membership, fallow). Moreover, they employ
marketing management (Porter 1996; Kotler 1997), there “the name” (i.e., the brand name) as the dominant posi-
appears to be a paucity of documented strategic position- tioning strategy.
ing models capable of being applied by managers and
advertising executives. This article deals with the actual In conclusion, the general patterns of the components
process of managing the concept of positioning. It at- of the model, i.e., decisions and activities, are described
tempts to put forward normative guidelines through the and underline the comprehensiveness and robustness of
formulation, development and operationalization of a the new model. Concerning the definition of the decisions
comprehensive composite strategic positioning frame- and activities incorporated in the framework, on the basis
work. Using a triangulation research methodology, a of inductive reasoning, the following three phases and the
conceptual positioning framework that is the composite of related managerial decisions and activities have been
two extant positioning frameworks (Brand Concept Im- identified:
age Management (BCM) by Park et al. (1986) and the
Generic Positioning Framework (GPF) proposed by Phase 1 Definition, by management, of the over-
Hooley et al. (1998)) is formulated, and given the diffi- all positioning aim(s).
culty in the positioning of services (Assael 1985; Zeithaml
and Bitner 1996), and the wide spectrum of services Phase 2 Identification of positioning objective(s),
available, it was decided to carry out the research in the which are deemed as, appropriate in order
United Kingdom plastic card service firms (e.g., credit to achieve the desired positioning aim(s).
cards, charge cards, store cards, debit cards industry). The
proposed framework thus represents the composite of the Phase 3 – (i) Decisions related to the selection of spe-
BCM and GPF positioning framework. The rationale for cific positioning strategies which reflect
combining these two models (see Jacoby 1978; Wright both the life cycle stage of the offering
and Kearns 1998) stems from the need for the develop- and the objectives identified in phase 2.
ment of a comprehensive framework that incorporates the
management of positioning over time and at the same Phase 3 – (ii) Management (i.e., implementation and
time, ensures marketing synergy, i.e., congruence of re- monitoring) of the positioning related
lated activities. activities.
From the findings, the paper puts forward some basic Secondly, based on literature-derived descriptions
normative guidelines and in the process, reveals that and in-depth face-to-face interview, the managerial deci-
services pursue two key positioning aims (profit and sions and activities were operationalized leading to the
market share; profit and status) and two main positioning following normative guidelines.
Positioning Aim(s) Positioning Objective(s) Life Cycle Stages (LCS)
Profit & Market Share Functional Primal

Profit & Status Symbolic Consolidation
Latent Position
Deposition
Fortification
Membership
Fallow

Thirdly, the resultant guidelines were applied in the positioning aim(s), objective(s), life cycle stages, and
U.K. plastic card services domain and in this process, data strategies. The latter has been distilled into a simple
were collected using a combination of face-to-face inter- summary and is presented in Table 1. Our study calls for
views (executives/experts), survey (target group), and new ways of thinking about and conceptualizing the
content analysis (company communications). These have application of positioning in service organizations.
been used in the descriptions of the four card brands’
TABLE 1
Summary of Application of the Comprehensive Strategic Positioning Model
Overall Positioning Strategies

Employed, i.e., Overlapping in
Executives/Experts,
Card Brand Positioning Positioning LCS Communications,
Aim(s) Objective(s) and Target Group
Credit Card: Profit and Market Functional Fortification Reliability and

Visa Share and Membership The Brand Name
Charge Card: Profit and Status Symbolic Membership Top of the Range
Amex and The Brand Name
Store Card: Profit and Status Functional Fortification Service, Value for Money,
M&S and The Brand Name
Debit Card: Profit and Market Functional Fortification The Brand Name
Switch Share

Charles Blankson
Department of Marketing
College of Management
Long Island University – C. W. Post Campus
Brookville, NY 11548–1300
Phone: 516.299.3094
FAX: 516.299.3917
E-Mail: Charles.Blankson@liu.edu

ANTECEDENTS AND CONSEQUENCES OF ROLE CLARITY IN
EXPLAINING EMPLOYEE-PERCEIVED SERVICE QUALITY
IN CALL CENTERS
Avinandan Mukherjee, Montclair State University, New Jersey
Neeru Malhotra, Aston University, United Kingdom
SUMMARY to influence decisions about their job (Teas 1983). Super-

visory consideration refers to leader behaviors concerned
Role clarity as perceived by frontline service employ- with promoting the comfort and well-being of the subor-
ees is the extent to which they receive and understand dinates (Boshoff and Mels 1995). Finally, team support –
information required to do the job in the manner that is support from co-workers – not only provides an outlet to
expected by the management (Kelly and Hise 1980; Teas service burnouts arising from difficult service encounters,
et al. 1979). Role clarity is critical to delivering service but also acts as a channel for disseminating practical
quality in service organizations. This is particularly true in knowledge and information relating to the jobs of the
call centers, where customer contact employees are fre- frontline employees (Sergeant and Frenkel 2000). Key
quently subjected to conflicting demands of cost effi- consequences of role clarity are organizational commit-
ciency and customer service, thus leading to lower per- ment, job satisfaction, and service quality. Job satisfac-
ceptions of role clarity. In the absence of adequate role tion and organizational commitment also affect service
clarity, call center representatives (CCR) could end up quality. Organizational (affective) commitment refers to
misguiding customers, providing wrong information to the employee’s emotional attachment to, identification
them, transferring calls unnecessarily, putting calls on with and involvement in the organization (Meyer and
hold or asking the customer to call later, all of which Allen 1991). Job satisfaction refers to the extent to which
would lead to poor service quality. Lack of role clarity can the employees feel satisfied with the kind of work they do
also have negative effects on job satisfaction and organi- and with the nature of their job. Service quality is the result
zational commitment (Ruyter et al. 2001). Hence, the key of human interaction between the service provider and the
purpose of this study is to investigate how frontline staff’s customer. Service quality as perceived by frontline em-
role clarity in telephonic voice-to-voice service encoun- ployees is chosen as the performance consequence of their
ters affects their perception of the service quality deliv- role clarity, as “customer contact employees are well
ered by them. placed to effectively judge the quality of services that they
deliver” (Sergeant and Frenkel 2000, p. 19). We adapted
We also examine in this research the influence of the SERVQUAL instrument (Parasuraman et al. 1988) to
selected antecedents and consequences of role clarity in include only those dimensions that relate to the employ-
explaining service quality. The conceptual model is pro- ees’ service quality. As far as possible, measurement
vided in Figure 1. According to Singh (1993), the study of items for the variables were drawn from standardized
relationships between organizational factors and role clarity scales well established in literature having acceptable
is rooted in the path-goal theory of leadership (House reliabilities. However, some items were adapted and the
1971) and the job characteristics model (Hackman and shortened versions of some scales were used based on a 2-
Oldham 1976). Key antecedents of role clarity are five stage pre-test. All items were linked to five point Likert
variables mostly relating to organizational job design type scale ranging from “strongly agree” to “strongly
characteristics – two job-related variables (feedback and disagree.”
autonomy), two supervisory variables (participation and
supervisory consideration), and one social variable (team- We based our study on an “in-house” call center of a
support). Feedback refers to the degree to which carrying major retail bank, where customer satisfaction is one of
out the work activities required by the job results in the the main objectives. Self-administered anonymous ques-
individual obtaining direct and clear information about tionnaires were mailed to their “Head of Customer Ser-
the effectiveness of his/her performance (Hackman and vices” responsible for call centers who further forwarded
Oldham 1976). Autonomy is the degree to which the job them to the respective CCRs. Questionnaires were distrib-
provides substantial freedom, independence, and discre- uted to 710 call center employees. Three hundred eighty
tion to the individual in scheduling the work and in questionnaires were returned to the researchers, provid-
determining the procedures to be used in carrying it out ing a response rate of 53.5 percent. These in turn yielded
(Hackman and Oldham 1976). Participation in decision- 342 useable questionnaires.
making refers to the degree to which employees are able

A structural equation model is developed and tested Our study helps to understand the nature and signifi-
on this sample of 342 call center representatives. A two- cance of role clarity in customer service. Role clarity
stage approach was followed (Anderson and Gerbing affects service quality positively, and directly as well as
1990). First, the measurement model was estimated and indirectly through organizational commitment and job
standardized regression coefficients obtained, and sec- satisfaction. Although job satisfaction does not affect
ond, the structural model was estimated. The final model service quality directly, it does so indirectly through
satisfied all three criteria (absolute fit, incremental fit and organizational commitment which has a positive effect on
parsimonious fit) to determine goodness-of-fit for the service quality. Further, managers need to act upon the
model. Of the 11 hypotheses, all except H2, H4, and H11 key antecedents of role clarity in call centers. Detailed
were accepted (see Figure 1). scripting and setting clear expectations in call centers
would improve role clarity. Effective feedback based on
Our research reveals that role clarity plays a critical random call recording becomes vital in a highly mecha-
role in explaining service quality. Further, feedback, nized call-center environment. Participation in decision-
participation and team support positively influence role making by the customer contact employees on issues
clarity, which in turn increases job satisfaction, organiza- concerning their jobs and going beyond scripts contribute
tional commitment and service quality. However, au- to role clarity. Team support, in terms of helpful and
tonomy and supervisory consideration have no signifi- supportive team workers, further assists in role clarity by
cant effect on role clarity. Our research suggests that disseminating useful information concerning various is-
boundary personnel in service firms should strive for sues in their jobs that are not explicitly known or in-
higher perceived role clarity to be able to deliver higher structed, and through sharing each other’s experiences
service quality. Hence, this study establishes linkages and learning from them. Overall, our research suggests
between internal marketing and external marketing in that boundary personnel in service firms should strive for
service firms, demonstrating that customer contact em- higher perceived role clarity to be able to deliver higher
ployees, who are clear of their job roles, feel satisfied and service quality.
committed and deliver better service quality to customers.
FIGURE 1
The Research Framework with Hypothesized Relationships and Structural Model Coefficients
Feedback
H1 (0.414*)
Autonomy
H2
H7 (0.12*) Organizational
Commitment
Participation H3 (0.19*)
Role
Role Service
clarity
Clarity H9 (0.226*) Quality
H4 H8 (0.652*)
Supervisory
H4
Consideration H6 (0.295*)
Job H11
Satisfaction
H5 (0.12*)
Team
Support
*Significant at p < .01, std parameter estimate within parenthesis.

REFERENCES Human Resource Management Review, 11 (1), 61–
89.
Anderson, J.C. and D.W. Gerbing (1990), “Structural Parasuraman, A., V.A. Zeithaml, and L.L. Berry (1988),
Equation Modeling in Practice: A Review and Rec- “SERVQUAL: A Multiple Item Scale for Measuring
ommended Two-Step Approach,” Psychological Consumer Perceptions of Service Quality,” Journal
Bulletin, 103 (3), 411–23. of Retailing, 64 (1), 12–40.
Boshoff, C. and G. Mels (1995), “A Causal Model to Ruyter, K. De, M. Wetzels, and R. Feinberg (2001), “Role
Evaluate the Relationships Among Supervision, Role Stress in Call Centres: Its Effects on Employee Per-
Stress, Organizational Commitment, and Internal Ser- formance and Satisfaction,” Journal of Interactive
vice Quality,” European Journal of Marketing, 29 Marketing, 15 (2), 23–35.
(2), 23–42. Sergeant A. and S. Frenkel (2000), “When Do Customer
Hackman, J.R. and G.R. Oldham (1976), “Motivation Contact Employees Satisfy Customers?” Journal of
Through the Design of Work: Test of a Theory,” Service Research, 3 (1), 18–34.
Organizational Behaviour and Human Performance, Singh, J. (1993), “Boundary Role Ambiguity: Facets,
16, 250–79. Determinants, and Impacts,” Journal of Marketing,
House, R.J. (1971), “A Path-Goal Theory of Leadership 57 (April), 11–31.
Effectiveness,” Administrative Science Quarterly, Teas, R.K., J.G. Wacker, and R.E. Hughes (1979), “A
321–39. Path Analysis of Causes and Consequences of
Kelly, J.P. and R.T. Hise (1980), “Role Conflict, Role Salesmen’s Perceptions of Role Clarity,” Journal of
Clarity, Job Tension, and Job Satisfaction in the Marketing Research, 16 (August), 335–69.
Brand Manager Position,” Journal of the Academy of ____________ (1983), “Supervisory Behaviour, Role
Marketing Science, 8 (2), 120–37. Stress, and the Job Satisfaction of Industrial Sales-
Meyer, J.P. and N.J. Allen (1991), “A Three-Component people,” Journal of Marketing Research, 20 (Febru-
Conceptualization of Organizational Commitment,” ary), 84–91.
For further information please contact:

Avinandan Mukherjee
Montclair State University
Montclair, NJ 07043
Phone: 973.655.5126
FAX: 973.655.7673
E-Mail: mukherjeeav@mail.montclair.edu

THE STATUS OF CROSS-FUNCTIONAL EDUCATION IN
UNDERGRADUATE MARKETING CURRICULA WITHIN
MANAGEMENT EDUCATION
Victoria L. Crittenden, Boston College, Chestnut Hill
Elizabeth J. Wilson, Suffolk University, Boston
Cameron Duffy, Boston College, Chestnut Hill
SUMMARY tions and teamwork that is facilitated and enhanced

by advances in technology.
It has been 20 years since Behrman and Levin sug-
gested that real-world business problems “do not yield to 2. Graduates with integrative academic experiences that
a single-discipline solution” (1984, p. 142). During this build on strong functional expertise have a competi-
20-year period, there have been numerous calls for inte- tive advantage in the job market and a greater chance
gration across college and university curriculums (cf., of long-term success.
Association of American Colleges 1985; Boyer 1987) and
within colleges of business in particular (Porter and Regarding the integration of marketing and other
McKibbin 1988). Schelfhaudt and Crittenden (2005) re- functional areas, Crittenden (2003) and Barber et al.
port, however, that while most MBA curricula have been (2001) suggest that the marketing function is in a bound-
revised to accommodate the trend toward cross-func- ary-spanning role in facilitating increased functional inte-
tional integration, the curricula of undergraduate business gration. It is not surprising, then, that many cross-func-
programs frequently fail to adapt to the changing needs of tional, academic examples include interactions with mar-
modern organizations. To meet these needs, employees keting. As suggested by Alden et al. (1991), these ex-
must be skilled in organizational flexibility, teamwork, amples cover a wide range of integration models, both
and cross-functional communication and collaboration within and outside the business school. Additionally, the
(Smart and Barnum 2000; Sheth 2002). There is concern functions included in the academic examples are consis-
as to whether undergraduate students will matriculate tent with the results of a 2002 study which found that
with the people and technical skills needed in today’s participation in new product development included mar-
workplace (cf., Newman 1999; Parker 2003). keting, quality, manufacturing, engineering, and purchas-
ing (Goldense and Schwartz 2002).
Marketing has been suggested as the broadest area in
business management and, as such, the boundary-span- Research Questions and Methodology
ning function with a company (Crittenden 2003). Aca-
demically, Barber et al. (2001, p. 240) suggested that the The current research addresses three major questions
marketing department in a college of business “is in a with respect to the business school’s marketing depart-
powerful position to serve an important role in guiding ment and its integration with other functional areas for
and binding together other areas” in the facilitation of undergraduate management education:
cross-functional teaching and learning. Following this
practitioner and academic line of logic, the study reported 1. Does integration occur between marketing and other
here assesses the state of cross-functional management functional areas?
education from the perspective of the chairs of marketing
departments in universities across the United States. 2. How is the integration accomplished?
Cross-Functional Education 3. What are the impediments to integration?
Several researchers have examined the need for cross- In the spring of 2004, a “Status of Cross-Functional
functional integration in the business school curriculum Business Education in the United States” survey was sent
(e.g., Alden et al. 1991; DeMoranville, Aurand, and to marketing department chairs at colleges and universi-
Gordon 2000; Schelfhaudt and Crittenden 2005). Based ties across the United States. The purpose of the survey
on these research findings, there are two major reasons for was to benchmark the state of cross-functional under-
integration within the business school program: graduate marketing education by addressing the three
major research questions identified above. The survey
1. Cross-functional integration is the norm in business, resulted in a 15 percent response rate, slightly below the
resulting in increased interdepartmental communica- 18 to 57 percent response rates found in previous studies

utilizing academic directories as the sampling frame lable by individual faculty members and less time-inten-
(Andrus, Laughlin, and Norvell 1995). sive than more comprehensive programs. In terms of
integration in the marketing classroom, respondents re-
Results ported some difficulty at finding materials, agreed that
teaching experience is an important resource for effective
It appears that marketing educators are participating delivery of cross-functional content, and suggested that
in some level of cross-functional education. Interestingly, industry experience is a plus in cross-functional teaching.
cross-functional education is more of a priority at private More importantly, educators may not be doing a good job
schools than public schools. The majority of integration conveying the importance of cross-functional learning.
takes place in Geiger and Dangerfield’s (1996) “integrat- References available upon request
ing project curriculum” model. This level is more control-

Vicky Crittenden
Boston College
450 Fulton Hall
Chestnut Hill, MA 02467
Phone: 617.552.0430
FAX: 617.552.6677
E-Mail: victoria.crittenden@bc.edu

ON DIFFERENT TEACHING PEDAGOGIES: WHAT HAPPENS TO YOUR
COURSE EVALUATIONS?
Alma Mintu-Wimsatt, Texas A&M University – Commerce, Commerce
Kendra Ingram, Texas A&M University – Commerce, Commerce
Mary Anne Milward, Texas A&M University – Commerce, Commerce
Courtney Russ, Texas A&M University – Commerce, Commerce
SUMMARY The results suggest that the traditional students con-

sistently ranked the professor the highest in all five
A recent conversation with several colleagues evaluative criteria. Looking at each of the evaluative
prompted quite an animated discussion regarding various criteria, only the variable “grading policies” produced
modes of delivering business education and the role of slightly different results. That is, for “teaching skills,”
instructional technology. Some vehemently claimed that “rapport with students,” “knowledge of material,” and
the traditional face-to-face pedagogy is by far the most “presentation skills” – the traditional class was signifi-
effective means of teaching students. Others purported cantly different (at p < .01) from ITV and Internet-based
that technology mediated distance learning modes are just methods. No statistical differences were found between
as effective. Interestingly enough, regardless of how each ITV and Internet-based groups on the said four variables.
colleague felt about various teaching pedagogies – ulti-
mately, a consensus was reached. That is, the group Regarding the variable “grading policies,” ITV stu-
agreed that their teaching evaluations had “gone haywire” dents ranked the professor the lowest. ITV result was
with distance education classes despite the advances in statistically different from traditional and Internet-based
instructional technology. This consensus opinion prompted at p < .01. Meanwhile, no statistical difference was found
us to investigate how teaching pedagogies affect students’ in the “grading policies” between the traditional and
course evaluations. That is, does delivery method affect Internet-based classes.
how students evaluate their instructors?
This findings in this study compared how one sea-
The primary purpose of this study was to compare soned professor’s evaluative rating [who has been recog-
MBA students’ evaluative perceptions of their professor nized with several university teaching awards] was sig-
in three different classroom contexts: traditional, ITV and nificantly affected by the mode of delivery. It appears that
Internet-Based Marketing Management course. This study instructors are rated better in the traditional face-to-face
investigated how students’ evaluation of the instructor context. However, it is noteworthy to mention that while
was affected when different teaching pedagogies are students rated both ITV and Internet-based modes lower,
used. MBA students evaluated their instructor based on the rating itself was still relatively favorable (i.e., means
five evaluation criteria: (1) teaching skills, (2) rapport of less than 2 in a scale of 1–5).
with students, (3) grading policies, (4) knowledge of
materials, and (5) presentation skills Given today’s student needs and university budget-
ary constraints, it is a foregone conclusion that instruc-
MBA students enrolled in the Marketing Manage- tional technology is here to stay. Unfortunately, while
ment course during their first year of the graduate program technology can significantly improve teaching effective-
served as the sample groups for this study. Most of the ness [through the use of e-mails, power point presenta-
students are in the 25–30 age range and have worked for tions, web pages, etc.] it can also provide many chal-
at least two years. The students were requested to com- lenges. Perhaps, there is none more intimidating to educa-
plete a standardized University evaluation form. For the tors than the seemingly adverse impact it has on student
Internet-based students, a second evaluation form was course evaluations. After all, most tenure and promotion
utilized to assess online technology-related issues. Course decisions hinge on students’ evaluative ratings of their
evaluation forms were distributed during the last week of professors.
classes.

Alma Mintu-Wimsatt
Department of Marketing & Management
Texas A&M University – Commerce
Commerce, TX 75429
Phone: 903.886.5698
FAX: 903.886.5702
E-Mail: alma_wimsatt@tamu-commerce.edu

AN ASSESSMENT OF A CONSUMER BEHAVIOR MULTIPLE-CHOICE
QUESTION TAXONOMY
John R. Dickinson, University of Windsor, Ontario
SUMMARY H2: The mean percent of correct answers to questions

classified as recall is greater than the mean percent of
This study conducts an assessment of the taxonomy correct answers to questions classified as applied.
into which multiple-choice questions are classified in a
question bank accompanying a widely adopted consumer H3: There is a significant interaction between classified
behavior text (Solomon, Zaichkowsky, and Polegato [SZP] question difficulty and classified question cognitive
2002). The study provides not only an assessment of that skill on percent of correct answers.
particular question bank, but also a pro forma for similar
assessments of other question banks. Multiple-choice H4: The mean point-biserial correlation of questions clas-
questions are classified by SZP on two dimensions: ques- sified as moderate is greater than the mean point-
tion difficulty (easy, moderate, difficult) and “. . . the main biserial correlation of questions classified as either
cognitive skill it is designed to test. . .” (Forrest 2002, easy or difficult.
“Preface”) (applied, recall).
H5: The mean point-biserial correlation of questions clas-
“The analysis of multiple-choice items typically be- sified as recall is not significantly different from the
gins with the computation of a difficulty and a discrimina- mean point-biserial correlation of questions classi-
tion index for every item.” (Aiken 1991, p. 78) In this fied as applied.
study question difficulty was operationalized as the per-
cent of correct responses for a given question. The mean H6: There is no significant interaction between classified
percent was 55.7 percent, ranging from 2.4 percent to question difficulty and classified question cognitive
100.0 percent. Discrimination was operationalized as the skill on discriminating ability.
point-biserial correlation between a student’s total exam
score (excluding the focal question) and the dichotomy of Data were drawn from two midterm examinations
whether the student answered the question correctly or (40 and 42 students, respectively) covering chapters one
incorrectly. The mean point-biserial correlation was 0.218, through eight and two noncumulative final examinations
ranging from -0.364 to 0.669. (41 and 42 students, respectively) covering chapters nine
through seventeen. For each examination, six questions
H1: The mean percent of correct answers is inversely were selected at random from each chapter on a systematic
related to classified question difficulty. basis, for a total of 204 questions.
TABLE 1
Factorial Analysis of Variance Significance Levels
Criterion
Factor Measured Discrimination

Difficultya Indexb
Classified Difficulty 0.031 0.876
Cognitive Skill 0.103 0.011
Difficulty x Cognitive Skill 0.983 0.948
a
Arc sine transformed
b
Fisher’s r to z transformed

H1 through H3 and H4 through H6 were analyzed for the respective criteria of percent answered correctly
using a 3 (difficulty levels) x 2 (cognitive skills) factorial and point-biserial correlation are presented in Table 2.
analysis of variance. Significance levels, i.e., p-values, for References available upon request.
all statistical tests are presented in Table 1. Mean values
TABLE 2
Mean Criterion Values
Percent Correct Point-Biserial Correlations
Factor Level Recall Applied Row Recall Applied Row
Easy 65.25 59.65 63.38 .251 .172 .225
Moderate 56.24 50.30 54.02 .251 .170 .221
Difficult 54.03 48.77 52.28 .231 .169 .210
Column 57.68 51.98 .243 .170
H1 and H2 were supported, H3 was not supported. H4 and H5 were not supported, H6 was supported.
John R. Dickinson
Odette School of Business
University of Windsor
Windsor, Ontario
Canada N9B 3P4
Phone: 519.243.4232, Ext. 3104
E-Mail: bjd@uwindsor.ca

A COMPARISON OF CONSUMERS’ RESPONSES TO TRADITIONAL
ADVERTISING AND PRODUCT PLACEMENT STRATEGIES:
IMPLICATIONS FOR ADVERTISERS
Terry Daugherty, University of Texas, Austin
Harsha Gangadharbatla, University of Texas, Austin
SUMMARY (Petty and Cacioppo 1983). Thus, information is pro-

cessed actively when deemed relevant and relegated to a
Product placement is a form of advertising and pro- secondary status when relevance is perceived as low. For
motion in which brands are placed in television shows, the most part, product placement is more likely processed
movies, or other entertainment content to generate visibil- through the peripheral route since the message is second-
ity and achieve audience exposure. Proponents of product ary to that of the media content. For instance, while
placement cite several potential advantages for embrac- watching movies or television shows the likelihood that a
ing this method, such as a long shelf life, prominent viewer is more focused on the placed product than on the
exposure, and enhanced realism. For example, even years programming content is low. While elaboration of mes-
after a movie is released or a television show airs, adver- sages can vary depending consumer motivation, the im-
tisers are still able to receive some level of benefit from plications for processing product placement is important
DVD/video releases and television re-runs (Morton 2002). when measuring the effectiveness of this tactic given the
Furthermore, product placement allows advertisers to inherent differences between various types of media and
escape “commercial zapping” because consumers cannot advertising. Gupta and Lord (1998) compared product
skip over products placed within media content. In fact, placement with traditional advertising and found that
television networks believe that by 2007, there will be prominent placements outperformed traditional advertis-
24.7 million DVR owners (Digital Video Recorders) ing. However, this work does not lend itself toward
resulting in a loss of about $6.6 billion in advertising understanding product placement occurring in other me-
revenue. However, there are common disadvantages as- dia modalities, such as television, computer games and
sociated with product placement, such as the lack of music videos (Nelson 2002).
control over how products are portrayed or incorporated
into a scene or storyline. Further, advertisers have no To explore these issues, five hundred adults were
influence over how successful media programming will randomly selected from an online panel and asked to
be, making it difficult to predict where to place brands for complete a thirty-item survey. Subsequently, a total of
maximum exposure. Nevertheless, advertisers continue 227 completed questionnaires were collected resulting in
to spend as much as $20 million to have their products a response rate of 45.4 percent (227 out of 500). The
appear in media content (Grover 2004). variables measured centralized around four categories:
opinions about advertising, opinions about multiple forms
While a considerable amount of research has ex- of product placement, perceived advertising effective-
plored the effects of product placement, key issues remain ness, and perceived effectiveness of multiple forms of
for understanding this strategic tactic. For instance, much product placement.
of the research investigating product placement to date
has advocated or used memory-based tests (either recall or The findings suggest that in general product place-
recognition) to assess effectiveness (Brennan and Dubas ment is less effective in generating brand awareness than
1999; Eddington 1991; Pracejus 1995; Russell 1998; traditional advertising, but more effective in stimulating
Weaver and Oliver 2000) with very little work done interest and generating immediate or short-term product
comparing the effectiveness of product placement versus purchase intentions. Further, while attitudes toward prod-
traditional advertising. Therefore, the purpose of this uct placement in movies and television were elevated over
study is to explore consumer attitudes toward the various those in video games and music lyrics, more work is
forms of product placement (i.e., movies, television, video needed in this area before any conclusions can be made
games, music lyrics, etc.) and investigate how these differ discounting these nontraditional alternatives.
from traditional advertising.
Academic theorists are poised to expand our knowl-
According to the Elaboration Likelihood Model, edge in this area as research exploring these issues has
consumers process information through a central route both important industry and theoretical implications for
when they are highly involved or interested in content and explaining the nature of product placement in media.
a peripheral route when involvement or interest is low Empirical research is needed though to fully conceptual-

ize this tactic and identify exactly how product placement and lead to future research in this area. While advertisers
impacts affective (attitude) responses rather than simply are continuing to search for alternative tactics to reach
cognitive (memory). The objective of this study was not their target audiences, media professionals need to be
to suggest definitive attitudinal or behavioral preferences cautious in how they approach these methods, as more
over product placement or traditional advertising. Rather, work is needed to extend our understanding of product
the goal was simply to expand this growing body of work placement. References available upon request.

Terry Daugherty
Department of Advertising
University of Texas at Austin
1 University Station A1200
Austin, TX 78712
Phone: 512.471.8917
FAX: 512.471.7018
E-Mail: Terry.Daugherty@mail.utexas.edu

NEW BRAND WORLDS: A COMPARISON OF COLLEGE STUDENT
ATTITUDES TOWARD BRAND PLACEMENTS IN FOUR MEDIA
Yongjun Sung, The University of Georgia, Athens
Federico de Gregorio, The University of Georgia, Athens
SUMMARY (attitude toward advertising in general and brand involve-

ment).
Discussions of brand placement in the popular press
and academic literature tend to predominantly revolve Employing a convenience sample of college students
around film. However, recent content analytic work (e.g., (n = 437), the results of this study reveal that opinions
de Gregorio and Sung 2004; Ferraro and Avery 2000; regarding movie and TV show placements tend to be
Friedman 1991) demonstrate that brand appearances within rather similar and more positive (perhaps due to perceived
other media are prevalent and continue to increase in similarities of their characteristics) than those dealing
incidence as time progresses. Brand placement in other with music and video games. Respondents tend to per-
media is not a recent phenomenon. Advertisers produced ceive music and video game placements as: more inappro-
and sponsored television shows in the 1950s such as priate, less effective enhancers of content realism, inferior
Texaco Star Theater (McCarthy 2001), the Sega video sources of brand information, less influential in purchase
game company placed Marlboro ads in its early racing behavior, and more unethical and misleading.
games (Emery 2002), and a song from 1903 entitled
“Under the Anheuser-Busch” asked listeners to “Come, As no previous study has looked at content genre
come, drink some Budwise [sic] with me” (Agenda, Inc. within the context of brand placement, our incorporation
2003). throws further light on and expands the knowledge of
consumer perceptions of placement strategy. Finding of
Although not voluminous, there has been a steady the study indicates that film and television have more
stream of academic research since the late 1980s on the genres considered appropriate for brand integrations than
strategy of brand placement. Although scholarly brand music or video games. It is interesting to note that ani-
placement research has been ongoing since the late 1980s mated fare for both movies and television is considered
(Steortz 1987), it has largely focused on the context of particularly inappropriate for brand placements by more
films, with a dearth of investigations of the practice in than 50 percent of respondents. It is also notable that genre
other media. Some commonly investigated factors have appropriateness for music tends to be split based on
included: extent and type of placement (e,g., Devanathan “mainstream-ness” – country, rock, and hip hop being
et al. 2002; Sapolsky and Kinney 1994), audience recall considered appropriate (genres generally well represented
and recognition of placed brands (e.g., Gupta and Lord on the weekly mainstream music charts) and blues, jazz,
1998; Ong and Meri 1994), and qualitative explorations Christian, and classical/opera inappropriate (genres rarely
of consumers assimilation and interpretation of the mean- on the mainstream music charts). For video games, sports
ing of brands placed (DeLorme and Reid 1999). is clearly the genre of choice among respondents (it is
possible that racing was considered a type of sport, thereby
In addition, a distinct sub-stream of the literature has explaining the closeness of percentages) and is also viewed
been devoted to the gauging of audience opinions about as especially appropriate for brand placement as a televi-
and attitudes toward brand placements in films. While sion genre. This would seem to point to an association of
numerous attitudinal studies of brand placement in films sports with branding, which is of intuitive sense as sports
exist, to the authors knowledge only a single, qualitative events tend to be brimming with sponsorships and adver-
study has yet examined attitudes toward brand placement tisements.
in multiple media (DeLorme 1998). This exploratory
study builds on and contributes to previous work by In addition, the present study provides evidence that
serving as the first quantitative investigation of attitudinal those who are more positively disposed towards advertis-
responses to brand placement in films, television shows, ing in general are also likely to be more accepting of its
popular songs, and video games. Its overall objectives are specific forms (in this case placed within media content).
to examine and compare the attitudes of college student Further, there exists a positive relationship between in-
consumers with regards to brand placement across differ- volvement with brands and positive perceptions of place-
ent media. In addition, this study also investigates two ment strategy. Thus, our results provide preliminary evi-
potential antecedents of attitudes toward brand placement dence that two ways advertisers can potentially segment

their brand placement audience is according to acceptance more detailed and precise points of segmentation for the
of advertising in general and level of involvement with two criteria than our separation of high/low and more/less.
brands. Greater research is needed, however, to develop

Yongjun Sung
Department of Advertising and Public Relations
Grady College of Journalism and Mass Communication
The University of Georgia
Athens, GA 30602
Phone: 706.549.0988
FAX: 706.542.2183
E-Mail: raramimi@uga.edu

BRANDS IN ACTION: THE ROLE OF BRAND PLACEMENTS IN
BUILDING CONSUMER-BRAND IDENTIFICATION
Andrew T. Stephen, University of Queensland, Australia
Leonard V. Coote, University of Queensland, Australia
SUMMARY target consumers, but the placement also shows brands

being used or consumed in natural settings. This may be
Brand placement in mainstream media such as films, more believable, since media characters with which con-
television programs, computer and video games, and sumers might relate to and identify with use the placed
music videos has become a common practice, and a brands. Social identity theory is thus relevant in this
component of many integrated marketing communication context. A conceivable outcome of brand placements
strategies. There are now countless examples of brand putting brands in “action” in the eyes of consumers may
placements in media, particularly in films and television be the development of consumer-brand identification.
programs. Some recent placements in television programs Identification literature has considered image-, knowl-
include General Motors’ products in CBS’s Survivor and edge-, and social-related antecedents of identification
Bravo’s Queer Eye for the Straight Guy, Mitsubishi, (e.g., Bhattacharya and Sen 2003). Brand placements
American Express, and Coors in NBC’s The Restaurant, showing brands in apparently realistic (yet contrived)
Coca-Cola in Fox’s American Idol, MSNBC in NBC’s contexts may signal richer image, knowledge, and social
The West Wing, Dell Computer in Fox’s 24, and the information about brands to potential consumers than
Trump corporate brand as the focus of NBC’s The Ap- other forms of marketing communications due to the more
prentice. Most commercial studio films feature brand realistic medium. Examining brand placements and con-
placements, such as Fox News in 20th Century Fox’s Day sumer-brand identification then helps not only to advance
After Tomorrow, and the suite of brands placed through- knowledge of the effectiveness of this communication
out the recent James Bond films (including Omega, BMW, device, but also to improve our understanding of how to
Aston Martin, Heineken, and Visa). actualize deep-level bonds of identification between con-
sumers and brands.
Past research on brand placement has focussed on
measuring the effectiveness of placements with respect to A conceptual framework advanced in this paper
consumer memory and recall and attitudes towards brands, proposes a process through which image-, knowledge-,
practitioners’ views on brand placement, and consumers’ and social benefit-related characteristics of brand place-
attitudes towards the practice of brand placement with ments contribute to the development of identification-
respect to mostly ethical and moral considerations. Little based bonds between consumers and brands. Loyalty or
extant research has considered more enduring and funda- relational outcomes of identification are also considered
mental processes or outcomes of brand placements. This (i.e., brand loyalty, brand advocacy, and brand defence).
paper links brand placement with brand loyalty outcomes This process is thought to be moderated by characteristics
through a process of consumer-brand identification. The of a marketing communication strategy of which a given
integration of consumer-brand identification with a non- placement is part, and characteristics of the focal brand.
traditional marketing communications strategy such as Consumer characteristics with respect to (1) attitudes
brand placement is novel. Despite their prevalence in towards brand placement (e.g., acceptance of the device),
marketing, the processes for leveraging brands to create and (2) personal relevance of the focal brand are also
strategically valuable communication mechanisms and posited as moderators of this process.
highly loyal consumers is not yet fully understood. The
existing literature has built a detailed and thorough foun- Implications for theory and practice are considered.
dation for understanding brand placement and its poten- The main contribution of this paper is the integration of
tial benefits to advertisers and marketers, however it is yet social identity theory and consumer identification with
to consider the contributions brand placement makes to the brand placement literature. These theories provide
developing strong social and emotional bonds between compelling justifications for brand placement as a mar-
consumers and their brands. keting communication strategy because they explicate the
deeper connections that can form between consumers and
Brand placements are a unique form of marketing their brands, and the attitudinal and behavioral loyalty-
communication in that they allow consumers to view related outcomes of these connections that are possible. In
brands “in action.” When brands are placed in popular terms of practical implications, this paper provides mar-
mass media, this not only provides exposure to potential keters and brand managers with a viable approach to

fostering consumer-brand identification. Bonds of identi- with brands and the outcomes of this process. The ability
fication between consumers and their brands are difficult of a brand placement to show a brand in “action” and to
to break; a loyal customer base with a high level of personify a brand by associating it with media characters
consumer-brand identification is strategically valuable in a realistic and believable manner makes it a unique
and a potential source of competitive advantage. communication device, and one that is conducive to
fostering consumer identification with a brand. While
Although brand placement has been used in mass advertisers and marketers rightly perceive brand place-
media for many decades, its current growth and economic ment to be an effective communications and brand-build-
value across a variety of media warrants a deeper exami- ing strategy, this paper considers the possibility that the
nation of its effectiveness as an emerging “hybrid” com- benefits advertisers can potentially enjoy from investing
munications strategy (cf., Balasubramanian 1994). This in such a strategy may be more wide-reaching and valu-
paper considers placement from a different perspective by able than previously thought. References available upon
examining its role in the process of consumers identifying request.

Andrew T. Stephen
UQ Business School
University of Queensland
Brisbane, Qld 4072
Australia
Phone: +61.7.3365.9721
FAX: +61.7.3365.6988
E-Mail: a.stephen@business.uq.edu.au

THE MARKETING COCOMPETITION PROCESS AND STRATEGIC
ALLIANCE INSTABILITY: A SYSTEM DYNAMICS MODEL
Anna Shaojie Cui, Michigan State University, East Lansing
Roger J. Calantone, Michigan State University, East Lansing
SUMMARY willingness to continue the partnership is based on the

firm dependence on the partner, satisfaction for past
To gain speed to market and to enable firms to address cooperation, and the learning potential in the partnership.
markets otherwise inaccessible in the short run, marketing
alliances often form. The coexistance of cooperation and Embedded in the model are positive loops capturing
competition between partners makes alliances a process the force of cooperation and negative loops capturing the
of cocompetition. The force of cooperation based on force of competition. Through the positive loops the
common benefits holds an alliance together, while the willingness to continue the partnership builds up itself.
force of competition based on private benefits tears an For example, increasing with the willing to continue,
alliance apart. The two countervailing forces interact with investment further enhances the firm’s dependence and
each other during the partnering process and influence the willingness to continue. Similarly, willingness to con-
stability of an alliance. Previous research has recognized tinue increases the cooperation effort and correspond-
competition as a source of alliance instability, however, ingly the benefits generated from the partnership, which
largely due to the limitation of methodologies employed in turn increases the willingness to continue. Through the
in extant literature, few research has taken a dynamic view negative loops the willingness to continue the partnership
and studied the cocompetition process overtime. reduces itself. In respect to benefit sharing, disadvantage
in sharing benefits, increased by high dependence and
System dynamics modeling is a simulation method reduced bargaining power, decreases the firm’s satisfac-
that has been widely applied to model all kinds of complex tion for the relative outcome and reduces its willingness to
systems. This methodology provides a way to study continue. Related to partner availability, learning im-
complex interactions and feedbacks overtime. Having not proves a firm competency and increases the availability of
been widely used in marketing, more specifically in alternative partners, which in turn decreases the firm’s
interfirm partnerships, this methodology has the potential dependence and willingness to continue. Also, because of
to improve our understanding of the interaction between opportunism, a firm dependence on the partner firm
cooperation and competition and its influence on the decreases the partner firm cooperation effort, which re-
instability of an alliance. duces the focal firm learning and willingness to continue.
In this study we build a system dynamics model in the The model is run under a symmetric setting where all
context of horizontal learning alliances (Figure 1). The parameters are set the same for the two firms and an
dyadic model captures the two partner firms’ learning, asymmetric setting where some parameters are set differ-
increase in competitive advantage, dependence and will- ent for the two firms. The simulation results confirm that
ingness to continue the partnership, with “willingness to when the force of competition overweighs the force of
continue the partnership” directly related to the stability of cooperation, the propensity of instability is increased.
an alliance. The model is symmetric between firm A and Further, when two firms are perfectly symmetric in learn-
B. For both firms, private learning increases individual ing capability, partner knowledge and intent of appropria-
firm’s competitive advantage, while collective learning tion, the partnership is highly stable, while asymmetry in
increases the joint competitive advantage, which is shared these characteristics increases the propensity of instabil-
by the two firms; dependence increases with investment in ity.
the partnership and unavailability of alternative partners;

FIGURE 1
A System Dynamics Model of Cocompetition in Learning Alliances

Anna Shaojie Cui
Department of Marketing and Supply Chain Management
The Eli Broad College of Business
Michigan State University
N370 Business College Complex
East Lansing, MI 48824
Phone: 517.432.5535
E-Mail: cuishaoj@msu.edu

FOCAL SUPPLIER OPPORTUNISM IN RETAILER
CATEGORY MANAGEMENT
Neil A. Morgan, University of North Carolina at Chapel Hill
Anna Kaleka, Cardiff University, United Kingdom
Richard A. Gooner, Scott, Madden & Associates, Raleigh
SUMMARY complex issue required by retailers. For example, most

theory and empirical evidence regarding buyer-seller
Leveraging suppliers resources and capabilities via relationships adopts a dyadic perspective, either focusing
category management has become the focus of wide- on a single buyer-supplier relationship (e.g., Dyer and
spread attention in the supermarket industry, as retailers Singh 1998; Jap 1999), or treating each supplier relation-
seek competitive advantage in the face of industry con- ship with a common buyer separately (e.g., Subramani
solidation, globalization, and the rapid expansion of mass- and Venkatraman 2003). Yet, since retailers have more
merchandisers into the grocery market (ACNielsen 1998; than one supplier to almost all categories, and prescrip-
Gruen and Shah 2000; Hopkins 2003). Category manage- tions advocate giving a key supplier significant influence
ment concerns treating sets of complementary and/or over retailer category management, the retailer-focal sup-
competing brands in retail settings as strategic business plier relationship may affect relationships with other
units and allocating resources within these categories to suppliers to the category in ways that impact the retailer
maximize planned outcomes (e.g., Blattberg and Fox performance. Since these conditions have not been ad-
1995; Dhar, Hoch, and Kumar 2001; Zenor 1994). Cat- dressed in previous theory development and empirical
egory management therefore involves the analysis of research, the literature has little guidance to offer manag-
category-level data, setting goals for category perfor- ers facing such situations.
mance, and the formulation and execution of category
plans, all of which might be undertaken with varying This research contributes to theories of interfirm
degrees of involvement from suppliers (Basuroy, Mantrala, relations and competitive advantage and offers guidance
and Walters 2001; Dussart 1998). However, analysts for retailer and supplier managers by developing and
suggest that retailers can enjoy significant performance testing a model of opportunism among focal suppliers at
benefits if retailers allow a focal supplier to assume the the category-level, its antecedents, and associated perfor-
role of “category captain” where the supplier undertakes mance outcomes. Our research contributes to knowledge
or has significant input into the retailer’s category man- in three areas. First, synthesizing insights gained from
agement (e.g., Blattberg and Fox 1995; Cannondale 1999; qualitative fieldwork with those available in the literature,
Freedman, Reyner, and Tochtermann 1997). Despite this, we develop a conceptual model of important antecedents
fear and/or experience of supplier opportunism means and consequences of category-level focal supplier oppor-
that many retailers are either unconvinced or report hav- tunism. Second, we test our model using data from 73
ing failed to make such category management relation- category managers representing six grocery retailers across
ships with focal suppliers work (e.g., Brandweek 1999; a representative set of 35 different product-categories to
Supermarket Business 1999). provide new empirical insights into focal supplier oppor-
tunism and its impact on category-level performance. Our
The literature offers surprisingly little guidance to data suggest that focal supplier opportunism decreases
retailer and supplier managers regarding this important retailer category performance. Consistent with retailer
issue. There has been little empirical study of category fears, we find that focal suppliers with significant influ-
management generally, and no research focused on the ence in retailers category management efforts are more
key issue of focal supplier opportunism at the category- likely to engage in opportunistic behavior. However, our
level (Dhar et al. 2001; Gruen and Shah 2000). The data also reveal that retailer fears that being dependent on
literature reveals a number of different theories relevant to a focal supplier will lead to greater supplier opportunism
understanding buyer-supplier relationships (e.g., transac- are largely unfounded, while supplier dependence on the
tion cost analysis, relational exchange, resource depen- retailer is also unrelated to focal supplier opportunism.
dence, etc.) and their impact on firm performance (re- Finally, we find that retailers’ ability to monitor – but not
source-based view, industrial organization, etc.). How- to punish – its focal suppliers is negatively related to
ever, many of these theories offer different viewpoints on opportunistic behavior among focal suppliers. Third, our
various aspects of supplier-retailer relationships in cat- findings illuminate the important but largely ignored case
egory management, and none provide a comprehensive in which a buyer relationship with a focal supplier puts
framework that permits the broad understanding of this that supplier in a position to directly influence the buyer

relationships with other competing suppliers, who also the category but that these militant behaviors do not
continue to supply products to the buyer. From this significantly negatively impact retailer category-level
perspective, we find that focal supplier opportunism in- performance.
creases militant behaviors among non-focal suppliers to

Neil A. Morgan
Kenan-Flagler Business School
CB# 3490, McColl Building
University of North Carolina at Chapel Hill
Chapel Hill, NC 27599–3490
Phone: 919.962.9835
FAX: 919.962.7186
E-Mail: Neil_Morgan@unc.edu

“NO GREATER SATISFACTION THAN TO VINDICATE EXPECTATION”
HOW AFFECTIVE EXPECTATIONS SHAPE
CONSUMPTION EXPERIENCE
Andrew K.C. Wong, The Chinese University of Hong Kong, Hong Kong
SUMMARY result, disconfirmed experience tends to be assimilated to

the affective expectations, leading to satisfaction.
Consumption experience is subjective, dynamic and
subject to external influences. Consumers may uncon- P1: Ad-induced affective expectations will lead to an
sciously color their judgments through the tinted glass of assimilation bias such that negatively disconfirmed
a priori expectations. Imagine, when we take the first sip experience will be evaluated in the direction toward
of an award-winning Pauillac wine, or when the long the prior affective expectations, provided that the
awaited sequel of an international blockbuster finally discrepancy is not noticed.
opens, are we immune to the influence of expectations and
let the moment of truth tells the truth? Ambiguity Moderates the Effects of Affective Expec-
tations
According to the Disconfirmation Paradigm (Olson
and Dover 1979), satisfaction is determined by the dis- The assimilation bias of affective expectations is
crepancy between expectation and experience. Treating contingent on the noticeability of the discrepancy (Wilson
and measuring expectation and experience as two inde- et al. 1989). When facing an ambiguous experience,
pendent entities, one can compute the discrepancy by which is open to multiple interpretations (Herr et al. 1983;
simple mathematics and determine the dis/satisfaction. Hoch 2002), a consumer may have difficulty in recogniz-
Accordingly the prediction is intuitively appealing: the ing the discrepancy, if any. Past research suggested that
higher the expectation, the bigger the disappointment. But ambiguous stimuli tended to be assimilated to contextual
what if the experiential perception is vulnerable to the stimuli (Herr et al. 1983; Hoch and Ha 1986; Martin et al.
magnetic force of expectations? The focus of this paper is 1990; Schwarz and Bless 1992), providing tentative sup-
to examine how prior expectations shape subsequence port to the notion. In the discussion of product attributes,
consumption experience. Hoch and Deighton (1989) related ambiguity to the search,
experience and credence qualities. Information in an ad
Affective Expectation Theory highlighting the experience or credence qualities of the
focal product is by default more ambiguous and likely to
Olson and Dover (1979) defined expectations as induce greater assimilation to affective expectations.
pretrial beliefs about the product, and belief as a subjec-
tive probability of association between a product and an P2: With the presence of affective expectations, an ad
attribute. Subsequent studies mostly focused on attribute- highlighting experience or credence attributes of a
based evaluation (e.g., Deighton 1984; Smith 1993). product or service will lead to greater assimilation
While emphasis has long been given to the cognitive side, bias than does an ad highlighting search attributes.
the affective side of expectation is under explored.
Ambiguity can be reduced by the presence of a
Not until recently was the affective nature of expec- comparison standard. It is not uncommon that past expe-
tations formally conceptualized and empirically tested. rience is recalled as a standard to resolve ambiguity. The
Wilson and Klaaren (1992) defined affective expecta- retrieval of past experience strongly suggests self-refer-
tions as “people’s predictions about how they will feel in encing, which occurs “when information is processed by
a particular situation or toward a specific stimulus” (p. 3). relating it to aspects of oneself, e.g., one’s own experi-
In the context of consumption, advertising can induce not ence” (Burnkrant and Unnava 1995, p. 17). Phillips and
only attribute-related expectations, but also the affective Baumgartner (2002) proposed that affective expectations
expectations of how a consumer feels during and after could be formed based either on the remembrances of
consumption. In accordance with the affective expecta- retrospective consumption-related emotions, or anticipa-
tion theory, it is possible that even the actual consumption tory visions of future emotions. These two types of self-
experience objectively fails to meet a priori expectation, referencing attempt are in accordance with the retrospec-
consumers would still unconsciously surrender to their tive and anticipatory self-referencing proposed by
own affective expectations such that the discrepancy Krishnamurthy and Sujan (1999). They defined retro-
between expectation and experience is not noticed. As a spective self-referencing as anchoring the stimulus to

autobiographical experiences, and anticipatory self-refer- the actual experience should be more “open” and more
encing as referencing the stimulus to imagined future likely to be assimilated to the affective expectations. In
experience. summary, the temporal orientation of self-referencing
may either resolve (under retrospective self-referencing)
Under retrospective self-referencing, the retrieved or enhance (under anticipatory self-referencing) the am-
details of a past consumption activated by an ad may serve biguity of the experience, and thus moderates how affec-
as a background against which the actual consumption is tive expectations influence judgment.
evaluated. Discrepancies, particularly those negative ones
that fall short of the expectations, should be more notice- P3: The assimilation bias of affective expectations is
able and subsequently reduce the assimilation bias or even reduced under retrospective self-referencing, but en-
lead to a contrast effect (Schwarz and Bless 1992). In the hanced under anticipatory self-referencing.
case of anticipatory self-referencing, the interpretation of

Andrew K.C. Wong
The Chinese University of Hong Kong
Shatin, N.T.
Hong Kong
Phone: 852.2609.7808
FAX: 852.2603.5473
E-Mail: kcwong@baf.msmail.cuhk.edu.hk

CUSTOMER DELIGHT: AN ATTEMPT TO COMPREHEND THE
DIMENSIONS THAT COMPOSE THE CONSTRUCT AND ITS
BEHAVIORAL CONSEQUENCES
Stefânia Ordovás de Almeida, Pontifícia Universidade Católica do Rio Grande do Sul, Brazil
Walter Meucci Nique, Universidade Federal do Rio Grande do Sul, Brazil
ABSTRACT UNDERSTANDING CUSTOMER DELIGHT
The goal of this research is to propose and test a scale Customer delight can be defined as “an emotion,
that is suitable to measure the dimensions that compose characterized by high levels of joy and surprise, felt by a
the customer delight construct, as well as its behavioral customer towards a company or its offering (product/
consequences. Aiming that, three studies were used to service)” (Kumar 1996, p. 9). Thus, customer delight is
prove the existence of content and construct validity for defined as a rather positive emotional state towards the
the proposed dimensions. purchase/consumption experience, generally derived from
the surprisingly positive disconfirmation level of per-
INTRODUCTION ceived performance (Oliver et al. 1997; Rust and Oliver
2000). Delight would be characterized as an emotion
Regardless of the great number of academic re- made up of cognitive and affective aspects, including here
searches giving prominence to customer satisfaction, the surprise (Kumar 1996). In this sense, Izard (1997) clari-
understanding of what happens to the customer when he/ fies that even the cognitive concepts inherent in satisfac-
she experiences something beyond satisfaction during the tion and, consequently, in customer delight – such as need
post-consumption experience is still incipient. Going be- and desire –, and its comparative standards are considered
yond satisfaction engenders a deeply positive emotional affective by nature or, at least, as having an affective
state regarding the experience of buying or consuming, component.
which is known as customer delight. Considering this
scenario, this paper attempts to contribute to the Con- Notwithstanding the fact that a delighted customer
sumer Behavior literature by proposing a scale that is an should be at first satisfied, delight cannot be mistaken for
attempt to comprehend and measure the dimensions that mere satisfaction. The differentiation basically occurs at
compose the customer delight construct, as well as its an arousal level of the positive emotional response: at a
behavioral consequences, and which is also suited to low level there lies satisfaction; at a high level, delight
evaluate different cases of buying or consuming experi- (Oliver and Westbrook 1993). In this sense, previous
ences. studies also determined that pleasure and arousal are part
of intrinsic characteristics of either the goods or the
With this in mind an exploratory/qualitative research services experiences. Therefore, both dimensions are
was carried out, trying to understand the construct and viewed as complementary in the delight formation, con-
generate items for the construction of a measurement sidering that pleasure by itself is regarded as a positive
scale. Subsequently, a quantitative/descriptive research affect of moderate arousal (Mano and Oliver 1993).
was conducted through three studies: the first two aimed
at the fine-tuning of the measures by successively apply- Other evidence on the differentiation between satis-
ing the scale to distinctive samples; the third also aimed at faction and delight can be found in the studies of Oliver
the validation of the scale through the criteria of model and Westbrook (1993). These studies showed that cus-
adjustment, unidimensionality, reliability, convergent, and tomers who experience high levels of joy and surprise
discriminant validity properly attributed to the instru- presented greater indexes of repurchase intention and a
ment. The content validity was a permanent concern high degree of expectations disconfirmation. Therefore,
during the study. the authors empirically confirmed the existence of a
delight state separated from satisfaction. Kumar (1996)
Furthermore, this study consists of a theoretical re- also attributed discriminant validity to the constructs and
view of the aspects that involve customer delight and the concluded that the effects of delight on the repurchase
description of the methodological procedures used in intentions of individuals are higher and go beyond those
field research. The explanation regarding the analyses of satisfaction. In this sense, it can be observed that the
performed and results found has the purpose of describing differentiation between satisfaction and delight is not
the research findings as a whole, giving a basis for the present just at the level of surprisingly positive
discussion that encloses the study. disconfirmation experienced, but also at the level of the

positive affect felt by the consumer, including here arousal, different samples of the same population were used.
and in the post-consumption behavior generated. Undergraduates from business schools of three different
universities in Brazil formed the population. The first and
THE DELIGHT SCALE DEVELOPMENT – second samples, which were used only for scale fine-
EXPLORATORY PHASE tuning, had 146 and 124 respondents respectively; the
third, which should also be submitted to validation, had
The exploratory/qualitative phase of the scale devel- 240 respondents. Regarding the samplings, it was verified
opment process was responsible for the construct domain that a rate of 81 percent of the respondents in the first
delimitation and creation of items. First a bibliographic study, 92 percent and 71 percent in the second and third
investigation was carried out through the analyses of studies respectively, analyzed experiences with goods,
several studies in the marketing and psychology areas. while the others did so with services.
After this investigation, 30 consumers were questioned
concerning their highly satisfactory consumption experi- The procedures used for testing and fine-tuning the
ences. Subsequently, 20 consumers whose illustrations fit scale were based on the propositions made by DeVellis
into the theoretical formulation of the theme were selected (1991), in which the exploratory factorial analysis (EFA),
to participate in the in-depth interviews. the communalities of the items – accepted at 0.5 (Evrard
2002); the measure of the reliability through Cronbach’s
The highly positive disconfirmation of expectations Alpha – accepted at 0.6 (Evrad 2002); the item-item
was observed in all respondents’ descriptions as a funda- correlation – minimum of 0.5 (Evrard 2002) and the item-
mental point in delight formation. In this sense, surprising total correlation – maximum of 0.80 (Kline 1998) were
consumption was experienced by all respondents inter- taken into account, apart from the basic descriptive mea-
viewed and was always related to the positive affect. The surements. The software used here was SPSS 10.0. In
possibility of personalization of services as a way of addition, following the recommendations of Hair et al.
surprising consumers, as well as the product tailored to (1998), unvaried outliers were examined through the Z
one’s needs and desires, was a reason for the descriptions coefficient test and only in the third data collection the
of delightful experiences through personalization. Con- necessity of withdrawing eight outliers from the sample
cerning the price perception, six respondents mentioned arose.
this item directly and, often, the positive perception of
price was associated to the positive performance of the Data Collection and Measurement Scale
product or service. The intention to buy again and the
tendency towards positive word of mouth were observed Data collection was carried out throughout March
among all respondents. and May/2003 by the researcher in the classrooms, and
respondents filled out the questionnaires by themselves.
The measuring items for the scale construction come The recommendation given to the participants when an-
from different sources. The ones responsible for the swering the questionnaire was to take into account a
positive surprise measurement were taken from the Dif- highly satisfactory consumption experience that they had
ferential Emotions Scale – DESII by Izard (1977). The gone through, considering the study focus. As a measure-
items used to measure the other dimensions were mainly ment scale, a 7-point Liket scale was used, where point
adapted from the studies of Kumar (1996), Oliver et al. one represented the “totally disagree” assertive, and point
(1997) and Rust and Oliver (2000), with the grounding of 7 a “totally agree” assertive. Items indicating high agree-
the in-depth interviews. These items were submitted to a ment represented a higher tendency to delight.
Portuguese version process. Once the version was fin-
ished, two marketing academics evaluated its comprehen- THE VALIDATION OF THE DELIGHT
sion. Therefore, three marketing experts analyzed the MEASUREMENT SCALE
items defined for the scale and suggested improvements.
Thus, 40 items formed the first version of the built scale Aiming at the proposed scale validation, two proce-
through five broad dimensions: cognitive aspects, affec- dures were carried out: the content validity and construct
tive aspects, positive surprise, price perception and post- validity, both suggested by DeVellis (1991). The content
purchase evaluation. validity was achieved by logical analysis. The construct
validation was done using the AMOS 4.0. software.
THE DELIGHT SCALE DEVELOPMENT – Besides the validation process, the inexistence of multi-
DESCRIPTIVE PHASE variate outliers (measured by the Mahalanobis D² dis-
tance) was confirmed. Small problems of normality were
The descriptive phase of the delight scale develop- also verified and eliminated, according to the principles
ment began with a pre-test of the instrument done with 10 suggested by Kline (1998). The results of the scale devel-
people from the same study population. With the objec- opment process and its later validation are presented
tive of testing, purifying and validating the scale, three below.

RESULTS OF THE DELIGHT SCALE The items related to the price perception measure-
DEVELOPMENT PROCESS – FIRST STUDY ment were not widely supported in the literature and they
were inserted in the analysis for their constant appearance
The first scale version, applied to the first sample and throughout the in-depth interviews. The observation of its
submitted to EFA, indicated nine factors responsible for low adherence to the scale and, consequently, to the
representing 67.32 percent of total variance explained. studied construct, led us to the belief that price perception
The results of the EFA, after some gatherings were done was not correlated to the presence of delight. Conse-
to regroup variables that were isolated or misallocated quently, the necessity of rewriting the statements on price
based on their conceptual coherence and their factorial was brought up from the moment of the first data collec-
loadings (Evrard 2002), are presented in Table 1 through tion – along with the bad indexes that also indicated the
the factorial loadings of the items and their communali- minor correlation of price factors with the construct.
ties. Therefore, despite the fact that the final price factor
(variables V37 and V39) had presented in the latter study
With the reliability and correlation analyses done, a Cronbach’s Alpha of 0.6996 and an item-total correla-
considering the EFA results, it was decided that some tion of 0.5381, the elimination of the factor was applied
variables needed to be removed. The variables and the for the theoretical motives explained above.
factor they belong to are: Factor 1: V4, V5, V10. Factor 2:
V18, V26. Factor 5: V6, V9, V36. Factor 6: V27. Since the With the removal of the price perception factor, a new
standardization measured through item V18 was not be- EFA was run and the final factorial structure presented
ing well comprehended in its relation to delight, a new five factors. Items V20 and V18 were presented alone
item called “I received personalized service/good” was under one factor that was called personalization dimen-
inserted with the same variable code for the second study. sion. According to Kumar’s (1996) study personalization
The Cronbach’s Alpha for the factors after the modifica- is a feeling that can be associated both to satisfaction and
tions were: Factor 1: 0,9303; Factor 2: 0.8542; Factor 3: delight. Furthermore, he points out that this appraisal is
0.8125; Factor 4: 0.8343; Factors 5: no items left; Factor diagnostic and a significant predictor of delight. Person-
6: only one item left; Factor 7: 0.5164. After all these alization is, most of the time, related to unexpected extras
procedures, a new EFA was done and no items presented provided by the employees to tailor the offer to customer
values lower than 0.5. needs. In this sense, higher levels of personalization make
the customer feel that the company cares about him,
THE DELIGHT SCALE DEVELOPMENT leading to delight (Kumar 1996). The final factorial struc-
PROCESS – SECOND STUDY ture of the scale to be validated, covering a total variance
explained of 68.71 percent, followed by its communali-
The adjusted factorial composition of the second ties, means and standard deviations is presented in Ta-
study showed five dimensions and a total variance ex- ble 2.
plained of 67.19 percent was presented, with a KMO
value of 0.861. Only one factor, represented by items V37, When analyzing the descriptive measures here pre-
V38, and V39, presented improper solutions in the analy- sented, the existence of high standard deviations was
sis performed, with a Cronbach’s Alpha of 0.3848, and detected. The comprehension for those values came though
improper correlations. In this way, variable V37 was the employment of an ANOVA among consumers of
rewritten as “the price of this service/good spurred me to services and goods that corroborated the existence of
buy,” and variable V38 was excluded for being prone to statistically significant differences in the groups’ means
dubious interpretations. Apart from that, no other changes for 16 of the 23 analyzed variables. In these cases, the
were made in the scale related to the previous study. consumers of services always had the highest means.
THE DELIGHT SCALE DEVELOPMENT Thus, the scale observed in Table 2 is the one pre-
PROCESS – THIRD STUDY sented for validation, being composed of 23 items, di-
vided into five factors. The dimensions that compose
The third sampling pointed to a factorial structure customer delight and its behavioral consequences are
composed of six dimensions, responsible for a total vari- described below, along with their respective Cronbach’s
ance explained of 64.52 percent and a KMO measure Alpha values: Factor1 – Affective Aspects Dimension
value of 0.874. For the results analysis, the EFA was (0.9179); Factor 2 – Cognitive Aspects Dimension
initially conducted and its results were compared to those (0.8497); Factor 3 – Post-consumption Evaluation Di-
from the internal consistency and item-item/item-total mension (0.7735); Factor 4 – Positive Surprise Dimen-
correlation analysis, running all the procedures again each sion (0.7725) and Factor 5 – Personalization Dimension
time a variable was removed. Thus, at the end of all (0.7830). As a result, the validation process of the pro-
analyses employed, the rejection of the following items posed scale is demonstrated below.
was performed: V40, V35, V17, V34, V13, and V16.

TABLE 1
First Study Initial Factorial Structure
Factor 1 – Items Loa. Com.

V11 This experience brought me happiness .803 .740
V3 This service/good transmitted positive feelings to me .698 .736
V7 This was a stimulating experience .679 .698
V15 This experience brought personal satisfaction to me .670 .662
V16 This was an irresistible experience to me .673 .736
V2 This experience brought me pleasure .666 .743
V1 This experience brought me joy .654 .758
V8 This experience was fascinating .652 .710
V14 I felt fulfilled with this consumption experience .647 .790
V17 This experience was attractive to me .608 .707
V10 This service/good made a dream come true .600 .602
V13 This experience made me feel special .597 .754
V5 I personally I identify myself with this service/good .592 .676
V12 This experience made me feel important .564 .723
V4 This consumption experience was one of the most important I have ever had .356 .600
V23 This service/good had an exceptional performance .794 .744
V24 This consumption experience completely satisfied my desires .755 .728
V19 This experience surpassed all my expectations .680 .670
V20 I received personalized assistance .678 .719
V22 The quality of the service/goods is superior to that of others .677 .643
V21 In this experience all the attributes that could be satisfactory were more than satisfactory .655 .691
V25 This service/good had great value to me .632 .562
V18 I received standard service/goods .481 .691
V26 I received more than is expected in this consumption experience .424 .608
V30 I was positively amazed with this experience of consumption .749 .651
V29 I was positively astounded with this consumption experience .692 .799
V28 I was positively surprised with this consumption experience .537 .711
V32 I intend to repurchase this service/good in the future .795 .723
V33 I would recommend this service/good to a friend .735 .803
V34 After this consumption experience I felt like complimenting the company/ employee .672 .726
V31 I am likely to make positive remarks about this service/good to other people .605 .609
V35 I am the most loyal customer after this experience .511 .662
V6 If I had not bought this service/good, I would have felt frustrated .694 .679
V36 I consider having a connection with the company whose service/goods I bought .653 .585
V9 This service/good was made to order for me .535 .570
V37 The price was a decisive variable within the choice of this service/good .834 .758
V27 There was flexibility so as to assist my necessities in this consumption experience .543 .530
V38 Regarding my purchase potential, the price of this service/good is too high .741 .727
V39 The price of this service/good influenced my level of satisfaction towards the
consumption experience .698 .577
V40 I believe I have paid a fair price for this service/good. .597 .573
Extraction Method: Principal Component Analysis with VARIMAX Rotation Source: Data Collection
Kaiser-Meyer-Olkin Measure (KMO): 0.862 Bartlett’s Test of Sphericity 1911.087 – sig. = 0.000

TABLE 2
Third Study Final Factorial Structure
Factor 1 – Affective Aspects Loa. Com. Means/Std.Dv.

V1 This experience brought me joy .824 .763 5.6/1.3
V7 This was a stimulating experience .809 .740 5.3/1.4
V3 This service/good transmitted positive feelings to me .806 .740 5.8/1.2
V2 This experience brought me pleasure .796 .677 5.6/1.3
V11 This experience brought me happiness .793 .653 5.3/1.3
V14 I felt fulfilled with this consumption experience .669 .629 5.2/1.4
V15 This experience brought me personal satisfaction .667 .622 5.5/1.4
V25 This service/good had great value to me .653 .596 4.4/1.7
V8 This experience was fascinating .638 .602 4.9/1.5
V12 This experience made me feel important .612 .572 4.6/1.7
Factor 2 – Cognitive Aspects Loa. Com. Means/Std.Dv.
V23 This service/good had an exceptional performance .812 .772 5.4/1.3
V22 The quality of the service/goods is superior to that of others .742 .595 5.2/1.5
V24 This consumption experience completely satisfied my desires .738 .689 6.0/1.2
V21 In this experience all the attributes that could be satisfactory
were more than satisfactory .644 .681 5.0/1.4
V19 This experience surpassed all my expectations .605 .597 5.3/1.3
Factor 3 – Post-Consumption Evaluation Loa. Com. Means/Std.Dv.
V33 I would recommend this service/good to a friend .832 .792 6.2/1.1
V32 I intend to repurchase this service/good in the future .820 .719 5.6/1.7
V31 I am likely to make positive remarks about this service/good
to other people .603 .605 5.8/1.4
Factor 4 – Positive Surprise Loa. Com. Means/Std.Dv.
V30 I was positively amazed with this experience of consumption .770 .672 4.1/1.6
V29 I was positively astounded with this consumption experience .702 .746 5.3/1.5
V28 I was positively surprised with this consumption experience .546 .711 5.6/1.2
Factor 5 – Personalization Loa. Com. Means/Std.Dv.
V20 I received personalized assistance .875 .818 4.9/2.0
V18 I received personalized service/good .868 .812 4.5/2.0
Extraction Method: Principal Component Analysis with VARIMAX Rotation Source: Data Collection
Kaiser-Meyer-Olkin Measure (KMO): 0.887 Bartlett’s Test of Sphericity: 3105.668 – sig.= 0.000
Note: Although the load of V28 is rather low, it was decided that this variable should be maintained, since otherwise
we would have a factor with only two variables, which is not recommended in the validation process. Furthermore,
this three-item factor had already been proposed and validated by Izard (1977), as well as used in many studies before
this one, proving its reliability.
RESULTS OF THE DELIGHT SCALE 1991; Garver and Mentzer 1999; Jöreskog and Sörbom
VALIDITION PROCESS 1982; Steenkamp and Trijp 1991). In these circumstances,
these authors recommend the necessary procedures to
Since the content validity was already achieved achieve construct validity as unidimensionality, reliabil-
through the methodological accuracy employed in the ity, convergent and discriminant validity, besides the
development phases and in the scale fine-tuning, clearly measurement models adjustment criterion. Table 3 shows
specified in the study method, with the aim of verifying the CFA results for the five-scale dimensions.
the construct validity, the five-scale dimensions were
submitted as five different measurement models to the Analyzing the fit-indexes, one can detect that most of
confirmatory factor analysis (CFA). This method is widely the measures are satisfactory. The χ2/df for the personal-
applied in marketing studies, as well as correlated fields, ization dimension is higher than expected, but the prob-
in the search for the construct validity (Bagozzi et al. ability of the qui-square being influenced by the sample

TABLE 3
Results of CFA for the Five-Scale Dimensions
Fit-Indexes Affect Cognition Post-Purchase Surprise Personalization
χ2)
Chi-square (χ 59.145 8.534 4.745 6.285 6.980
Degrees of freedom (GL) 33 4 1 2 1
χ2/GL)
Chi-square/degrees of freedom (χ 1.792 2.134 4.745 3.124 6.980
Goodness-of-fit Index (GFI) 0.952 0.985 0.986 0.983 0.971
Adjusted Goodness-of-fit Index (AGFI) 0.919 0.944 0.917 0.950 0.913
Tucker-Lewis Index (TLI) 0.976 0.975 0.964 0.976 0.945
Comparative Fit Index (CFI) 0.983 0.990 0.988 0.984 0.945
Root Mean Square Error of
Approximation (RMSEA) 0.059 0.070 0.127 0.096 0.161
Estimation Method: Maximum Likelihood Source: Data Collection
size is well known and discussed in the literature (Hair an average variance extracted (AVE) over 0.5. By analyz-
et al. 1998; Kline 1998). Therefore, Raykov and ing the reliability of the five dimensions, we could affirm
Marcoulides (2000) recommend that other fit-indexes that all the measurement models are reliable. Thus, the
should be analyzed, so that a real idea on the model affective dimension presented CR = 0.9749 and AVE =
adjustment can be reached. 0.6880, the cognitive dimension presented CR = 0.9609
and AVE = 0.6296, the post-consumption evaluation
When analyzing the measures of the indexes GFI, dimension presented CR = 0.9672 and AVE = 0.6887, the
AGFI, TLI, and CFI, it can be noted that their values were positive surprise dimension presented CR = 0.9695 and
always higher than 0.90, which is in agreement with the AVE = 0.7121, and the personalization dimension pre-
recommendations of Hair et al. (1998) and Kline (1998). sented CR = 0.9648 and AVE = 0.6406.
The adjustment measure RSMEA possesses acceptable
values ranging between 0.05 and 0.08 (Hair et al. 1998). For Garver and Mentzer (1999) and Bagozzi et al.
Analyzing the data mentioned in Table 3, one can observe (1991), the convergent validity of a scale is given by the
that the dimensions of post-purchase evaluation, positive quality of its fit-indexes. Besides all the scale dimensions
surprise and personalization presented values that ex- having been validated through the examination of their
ceeded the acceptability for the referred index. According fit-indexes, the factor regression coefficients were also
to Raykov and Marcoudiles (2000), this index shares the verified to certify the convergent validity. All the factor
same theoretical nature of the CFI. Thus, due to the high regression coefficients were statistically significant, which,
values of the CFI obtained in the analysed dimensions, according to Bagozzi et al. (1991), means that t-values are
these measures were considered validated as well. higher than |2,00| for p < 0.05 (7.86 was the lowest t-value
found in the study). As a strong condition for convergent
After presenting the criteria that led to the conclusion validity, Steenkamp and Trijp (1991) suggest that all the
that all the dimensions of the proposed scale are fit and factor regression coefficients should exceed 0.5. This
validated, unidimensionality was pursued and achieved condition was also achieved, since 0.591 was the lowest
through the examination of standardized residues pro- value found.
posed by Garver and Mentzer (1999), Jöreskog and Sörbom
(1988), and Steenkamp and Trijp (1991). According to the As a final procedure, so as to infer construct validity,
authors, in order to present unidimensionality, a dimen- the discriminant validity was verified through the method
sion should have all its standardized residuals lower than proposed by Fornell and Larcker (1981). This method
2.58. The highest standardized residual score found in the proposes that the average variance extracted for each
study was 2.049 in the personalization dimension, there- dimension should be higher than the squared correlation
fore, all dimensions may be considered unidimensional. between this dimension and any other dimension, which
characterizes the shared variance. For the studied con-
According to Garver and Mentzer (1999) a reliable struct – customer delight and its behavioral consequences –
measurement model – in this case a reliable dimension – discriminant validity was detected among the dimensions.
should present a composite reliability (CR) over 0.7 and This can be observed by the values of the average variance

extracted for these dimensions: 0.6880 for the affective found. However, the ANOVA realized in the third study
aspects, 0.6296 for the cognitive aspects, 0.6880 for the denotes the existence of statistically significant differ-
post-purchase evaluation, 0.7122 for the positive surprise ences between the means of consumers of goods and
and 0.6406 for the personalization dimension. These services: the latter group’s means are always statistically
values were by far higher than the highest value found for higher. In these circumstances, the importance of the
the shared variances among dimensions (0.3600). The insertion of a variety of services in goods that are gener-
following discussion comments all these results. ally standardized must be emphasized, considering the
necessity of variety when one is seeking to deliver an
DISCUSSION unexpected pleasant performance (Oliver et al. 1997).
What could be observed from the consumption of goods
The scale construction and test practiced in this study is that most of them were symbolic ones, so the need of
meet the demand of developing a greater theoretical personalization was replaced, in these cases, by the con-
comprehension on customer delight. Within this context, gruency of the good with consumer self-concept. Besides
the first evidence on the study can be described as a that, these differences are not relevant to the process of
measure purification relevance, through a successive scale scale development; neither do they show the need for
reapplication; for only the empirical evidence can prove separated scales to evaluate the consumption of goods and
that some items, or even entire dimensions, such as price services. The difference between these consumers lies in
perception, did not belong to the construct. Hence, an the kind of consumption experience and not in the dimen-
implication for the marketing academy is demonstrated in sions analyzed, which pertain to both of them.
the confirmation of the dimensions that compose the
customer delight construct and its behavioral conse- Another question to be regarded is the conclusion that
quences. This evidence was achieved through the favor- the price perception variables are not a customer delight
able fit-indexes of the models and the confirmation of dimension. Thus, the perception of a low or fair price is
unidimensionality, reliability and validity for the pro- only part of customer delight when accompanied by other
posed scale in all analyzed dimensions, which confirms its attributes that also have the capacity of delighting (e.g.,
construct validity. “these pants fit me perfectly and they were also cheap” –
respondents answered in one in-depth interview). The fact
The content validity was also achieved, by logical of being only cheap is not enough to give the pants the
analysis, because of the quality of the scale measurement power of delight, it solely satisfies the bargain or equity
items, accessed by the exploratory/qualitative research, item. The presence of delight presumes that the experi-
which was done before the scale construction, and also by ence possesses other components that might effectively
the successive fine-tunings. Yet, a scale application with generate a positive affect and surprise. Therefore, still
new contexts and another population is encouraged, sup- maintaining the consideration of the need for new studies
plying the deficiencies that might have come up in this so as to confirm the real accessory role of price in delight
study, such as the employment of the students’ sample, formation, care must be taken when intending to use it as
concerning the limitations of external validity that result a way of delighting customers. This observation is impor-
with this type of sampling. tant for organizations to perceive that delight should occur
through their differentiated attributes rather than strate-
It’s important to note that although discriminant gies that can be easily copied by their competitors, such as
validity was established between the dimensions of the price promotion (Rust and Oliver 2000).
scale proposed, this was not done between delight and its
close relatives, such as joy, as proposed by Oliver et al. All things considered, the inclusion of delight, through
(1997), which suggested a more psychometric specifica- the proposed scale, in theoretical models on the post-
tion and differentiation between the called “positive af- consumption evaluation process will allow a deeper un-
fects.” Therefore, this can be a suggestion for future derstanding of the distinct impacts between satisfaction
studies. The discriminant validity between satisfaction and customer delight in following behavioral results. The
and delight was not the goal of this study either, as already complete theoretical model test is another step to be taken
established by Kumar (1996). The peditive and nomologi- towards the comprehension of the existent relations among
cal validity were not established either, since, to infer this, the construct dimensions and the verification of which
it would be necessary to test the whole structural model. dimensions have a higher impact on subsequent behav-
Thus, it remains as a suggestion for the future. iors. Furthermore, other relations that search for a greater
comprehension of the role of customer delight in the post-
When evaluating the research object used by the consumption evaluation should be empirically verified,
respondents in the three studies, it is observed that the pursuing a wider vision of customer delight as a highly
citation of goods shows supremacy when compared to the positive emotional result to the experience of consump-
ones of services. Causes for this phenomenon were not yet tion.

REFERENCES Kline, Rex B. (1998), Principles and Practice of Struc-
tural Equation Modeling. New York: The Guilford
Bagozzi, Richard P., Youja Yi, and Lynn W. Philips Press.
(1991), “Assessing Construct Validity in Organiza- Kumar, Anand (1996), Customer Delight: Creating and
tional Research,” Administrative Science Quarterly, Maintaining Competitive Advantage. Doctoral dis-
36, 421–58. sertation, Graduate Faculty, Indiana University.
Devellis, R.F. (1991), Scale Development: Theory and Mano, Haim and Richard L. Oliver (1993), “Assessing the
Aplications. Newbury Park: Sage. Dimensionality and Structure of Consumption Expe-
Evrard, Yves (2002), “Instrumentos de Pesquisa, Coleta rience: Evaluation, Feeling and Satisfaction,” Jour-
e Análise de Dados,” class material, Business School, nal of Consumer Research, 20 (December), 451–66.
Federal University of Rio Grande do Sul/Brazil. Oliver, Richard L. and Robert A. Westbrook (1993),
Fornell, Claes and David F. Larcker (1981), “Evaluating “Profiles of Consumer Emotions and Satisfaction in
Structural Equation Models with Unobservable Vari- Ownership and Usage,” Journal of Consumer Satis-
ables and Measurement Error,” Journal of Marketing faction, Dissatisfaction, and Complaining Behavior,
Research, 18 (February), 39–50. 6, 12–27.
Garver, Michael S. and John T. Mentzer (1999), “Logis- ____________, Roland T. Rust, and Sajeev Varki (1997),
tics Research Methods: Employing Structural Equa- “Customer Delight: Foundations, Findings, and Mana-
tion Modeling to Test for Construct Validity,” Jour- gerial Insight,” Journal of Retailing, 73 (3), 311–36.
nal of Business Logistics, 20 (1), 33–57. Raykov, Tenko and George A. Marcoulides (2000), A
Hair, J.F.J., R.E. Anderson, R.L. Tatham, and W.C. Black First Course in Structural Equation Modeling. New
(1988), Multivariate Data Analysis. New Jersey, Jersey: Lawrence Erlbaum Associates.
EUA: Prentice Hall. Rust, Roland T. and Richard L. Oliver (2000), “Should
Izard, Carroll E. (1977), Human Emotions. New York: We Delight the Customer?” Journal of the Academy
Plenum. of Marketing Science, 28 (1), 86–94.
Jöreskog, Karl G. and Dag Sörbom (1982), “Recent Steenkamp, Jan-Benedict E.M. and Hans C.M. Trijp
Developments in Structural Equation Modeling,” (1991), “The Use of LISREL in Validating Market-
Journal of Marketing Research, 19 (November), ing Constructs,” International Journal of Research
404–16. in Marketing, 8, 283–99.

Stefânia Ordovás de Almeida
Faculdade dos Imigrantes/Pontifícia Universidade Católica do Rio Grande do Sul
Afonso Taunay 180/605
Porto Alegre/RS
Brazil 90520–540
Phone: 00.55.54.2271433
FAX: 00.55.54.2271433
E-Mail: stefania.almeida@terra.com.br

TOWARDS A CONCEPTUAL FRAMEWORK OF E-CONFUSION
Vincent-Wayne Mitchell, City University of London, United Kingdom
Gianfranco Walsh, University of Strathclyde in Glasgow, United Kingdom
SUMMARY ability to better organize visual stimuli makes field inde-

pendent consumers less likely to experience e-confusion
In 2004, the worldwide Internet population amounted from overload and unclarity antecedents. Situational vari-
to 945 million and projections for 2005 put this at 1.1 ables, such as using the Internet under time constraints,
billion (Clickz.com 2004a). However, this growing num- can lead to rushed decision-making, shortened informa-
ber of consumers is presented with overwhelming amounts tion-processing and inference-making time which is ex-
of online information and websites and the pace of tech- pected to increase e-confusion because of decreased pro-
nological development is faster than some consumers can cessing time.
adapt to. Recent authors have suggested that hyper-choice
confuses people, increases regret, is initially attractive but When e-confused, consumers attempt to understand
ultimately unsatisfying and is psychologically draining who they will attribute responsibility for the confusion
(Mick, Broniarczyk, and Haidt 2004). Such confusion can and can either blame themselves or others for the confu-
affect attitudes toward online-shopping and make con- sion. We suggest that attribution serves as a moderator of
sumers; dissatisfied, less brand loyal, less trusting, post- the confusion-outcome link. We propose that the immedi-
pone purchases, and shop offline (Walsh et al. 2002). It is ate effect of all e-confusion is indecisiveness and hesita-
surprising then that so little research has addressed the tion resulting in the consumer either doing nothing or
specific issue of consumer confusion on the Internet and postponing their web activity. We conceptualize e-confu-
its effects on consumers. The aim of this research then is sion as a temporary state of disorientation which has
to develop a conceptual model of e-confusion and to feelings associated with it. These affective consequences
explore some of its antecedents and consequences using a represent an individual’s feelings and emotions in relation
cognitive, affective and behavioral framework. to their evaluation of the website such as frustration,
irritation, anxiety, or even anger. Cognitive consequences
Our framework extends that of Walsh et al. (2004) refer to consumers’ knowledge and beliefs about a website.
who suggest confusion can be caused by too similar, too Consumers may form erroneous beliefs about the website
many or unclear stimuli. Websites that look and feel which can affect trust, satisfaction and loyalty. Behavioral
similar and give visual cues and/or brand appearance of a consequences represent consumers’ behavioral intentions
“look-a-like” or “fake” website can cause consumers to or tendencies, or actual overt actions toward the website.
believe mistakenly, for example, that website A is the We divide behavioral consequences into two types; (1)
same as website B. The website, www.pirated-sites.com, Generic, i.e., those which are generic to e-confusion and
offers many cases of Web design piracy. Consumers can (2) Specific, i.e., those which are specific to the anteced-
feel unclarity e-confusion when they encounter ambigu- ents causing e-confusion. E-companies need to learn how
ous, incongruous, misleading or inadequate Internet in- to identify and eliminate sources of confusion on their
formation. For example, Mobile phone sites (e.g., websites and could reduce confusion sources by doing a
www.dialaphone.com) can make prices difficult to under- confusion audit of their websites. From a regulation
stand and workout. Overload e-confusion antecedents can perspective, more governmental and private monitoring
occur when consumers are confronted with more product- of the web might help to reduce the numbers of copycat
related e-information than can be processed in the short- website and cyber squatting which will reduce sources of
term memory. A good example of e-information overload similarity-induced e-confusion. Better e-mail filtering
is the often uncategorized results of search engines. Indi- systems will obviously help, but it may be time for the
vidual characteristics exert an influence on e-confusion establishment of a government sponsored e-mail prefer-
because they are often linked to the consumer’s ability to ence service (EPS), to which all organizations should
rationalize and process stimuli. For example, age may subscribe. Our main contribution to theory lies in devel-
reduce e-confusion through an Internet experience frame- oping a model of consumer e-confusion. Future research
work or may increase e-confusion as processing compe- should include the development of a measurement instru-
tence decreases with the ageing process. Field indepen- ment and subsequent testing of the conceptual model and
dent individuals impose organization upon visual stimuli, hypothesized relationships. References are available upon
and are able to locate a sought-after component. The request.

Vincent-Wayne Mitchell
Cass Business School
City University of London
106 Bunhill Row
London
EC1Y 8TZ
Phone: +44.207.040.5108
FAX: +44.207.040.8328
E-Mail: V.Mitchell@city.ac.uk

DIFFERENTIAL IMPACTS OF CRM ON CONSUMER RESPONSES:
TOWARD AN INTEGRATIVE FRAMEWORK
Frederick Hong-Kit Yim, Drexel University, Philadelphia
SUMMARY ers” is more strongly related to value perceived by con-

sumers.
As a new paradigm in marketing, CRM is in need of
theoretical assistance (Gummesson 2002). Much, if not Moreover, drawing from the diverse research on
most, of the research on CRM has tended to focus on its relational exchanges (e.g., Garbarino and Johnson 1999;
favorable performance metrics such as sales growth. Nijssen et al. 2003; Sirdeshmukh, Singh, and Sabol 2002),
There is scant research effort directed at examining the a comprehensive theoretical framework for understand-
differential impacts of CRM dimensions on delivering ing how consumer loyalty can be achieved through value,
consumer value. In particular, there is no attempt to look satisfaction, and trust in the CRM context is developed.
into the mechanism of how customer loyalty can be Specifically, we postulate that value perceived by con-
nurtured via value, satisfaction, and trust in CRM projects. sumers is positively related to satisfaction with the service
firm, which, in turn, is positively related to trust in the
Notwithstanding some disappointing CRM results, service firm. We further posit that trust in the service firm
we believe that the early failures of CRM are not perennial is positively related to loyalty to the service firm.
and CRM movements are not expected to terminate (Sheth
2002). A sharper theoretical understanding of CRM can In conclusion, in response to the call for more theo-
effectively advance future CRM endeavors and success- retical assistance for CRM (Gummesson 2002), our re-
fully breed outstanding firm performance. To avoid the search contributes to the marketing discipline both theo-
premature death of the CRM discipline (Fournier, Dobscha, retically and managerially. Theoretically, it advances the
and Mick 1998), this study seeks to advance the CRM CRM literature by explicating how a number of ex-
literature by examining how the CRM dimensions differ- change-relevant constructs are related, and sheds light on
entially impact consumer value. In particular, CRM is how organizational CRM efforts differentially impact
conceptualized to have four distinct behavioral dimen- value perceived by consumers in a service setting. As far
sions, namely, (1) focusing on key customers, (2) organiz- as managerial implications are concerned, managers are
ing around CRM, (3) managing knowledge, and (4) lever- afforded the broad picture of how loyalty can be grown via
aging information technology (Yim 2002). We posit that a number of consumer response constructs, and therefore
when compared with other CRM dimensions of “organiz- inculcated the effective way to direct their CRM efforts in
ing around CRM,” “managing knowledge,” and “leverag- terms of delivering value to their key consumers.
ing information technology,” “focusing on key custom-

Frederick Hong-Kit Yim
Bennett S. LeBow College of Business
Drexel University
32nd and Chestnut Streets
Philadelphia, PA 19104
Phone: 215.990.9129 or 215.895.2145
FAX: 215.895.6975
E-Mail: hky23@drexel.edu

EXPLORING THE PHENOMENON OF BUYER-SELLER MISMATCHES
IN BUSINESS-TO-BUSINESS RELATIONSHIPS
Christopher P. Blocker, University of Tennessee, Knoxville1
ABSTRACT es relative to the situation (Coviello et al. 2002). However,

research is still seeking to build theory that can prescribe
Evidence shows relational strategies can prove suc- targeted relationship strategies to firms as they interact
cessful in some cases but are costly and ineffective in with diverse sets of customers in various situations
others. Yet, understanding favorable conditions for a (Siguaw, Baker, and Simpson 2003). Existing efforts to
relational over an arm’s length approach is still emerging. address this issue take several approaches, including:
One area that holds promise for shedding light on this identifying relationship-relevant dimensions on a contin-
issue is buyer-seller mismatches, as they occur when firms uum (Anderson and Narus 1991; Webster 1992), map-
strive for dissimilar goals and relate in incompatible ways. ping product market types to relational strategies (Coviel-
This paper draws focus to this concept by integrating the lo et al. 2002), analyzing relationships with ROI (Gum-
literature and exploring the complexity of buyer-seller messon 2004), and identifying profitable customers (Re-
incompatibility. inartz and Kumar 2003). For the most part, these ap-
proaches take a seller’s one-sided viewpoint. Unfortu-
INTRODUCTION nately, key questions of whether the other party desires a
relationship – or perhaps more important – whether a
Building strategic buyer-seller relationships that are “close relationship” means the same thing to both parties
marked by long-term, close dealings has captured the are largely left unaddressed.
attention of firms and researchers alike (Day 2000). Given
the substantial resources firms invest to make relation- Additionally, some have used the concept of buyer-
ships work, it is unsettling to find a significant number of seller mismatches – specified as a mismatch in relation-
relationships end in failure (Das and Teng 2000; Wathne, ship goals – to provide greater explanation into why
Biong, and Heide 2001). One leading scholar posits that relationships might succeed or fail (Gronroos 1997; Pels,
“some firms lose half or more of their customers every Coviello, and Brodie 2000). This approach has at least two
three years, and the worst is still to come” (Day 2000, key advantages. First, assessing mismatches requires a
p. 25). These are difficult realities considering the abun- dyadic perspective. Knowing only one party’s intentions
dance of research performed in relationship marketing and goals is not enough. Second, mismatch consider-
over the past two decades. ations give priority not only to stable characteristics of
each party but also to situations that might significantly
In view of these realities, research in business-to- influence relationship outcomes. If firms can understand
business relationships shows a noticeable movement from possible mismatches up front, before investing significant
enthusiasm that included declarations of a new paradigm resources, there is potential to drive down the rate of
(Gronroos 1994) and a fundamental re-shaping of the relationship failure, and thus, risks associated with sunk
field (Webster 1992), to a dampened realization that costs and missed opportunities elsewhere.
relationship strategies are not universally applicable or
painless to implement (Day 2000). For example, many Use of buyer-seller mismatches as a specified con-
firms make significant investments in technologies (e.g., struct in relationship research is infrequent. However, the
CRM) promising greater customer retention and loyalty, concept is inherent in much of the literature that, for
but the large majority fail to realize the anticipated return example, discusses mismatches in relationship values,
(Newell 2003). Additionally, researchers historically have social mismatches leading to relationship failure, and
argued for suppliers to shift customer interaction from a relationship dysfunctions, to name a few. In general,
transactional mode to a relationship building mode (Kot- research questioning the appropriateness of relationships
ler 1991; Webster 1992). Regrettably, this approach has in light of potential incompatibilities supports the idea of
influenced some firms to attempt partnering initiatives mismatches (Colgate and Danaher 2000). However, mis-
without respect to the other party’s fit or desire (Day 2000; match research thus far oversimplifies the phenomenon.
Garbarino and Johnson 1999). Existing research points solely to a single match or mis-
match from a global view of the relationship. At this
Recently, researchers have urged firms to pull back aggregate level, there is little appreciation for the com-
the reins on relationship strategies, in favor of plural plexity of business exchanges, wherein a variety of mis-
strategies blending relational and transactional approach- matched goals might occur at different levels of the

relationship. This leads to several questions such as: What 1987), satisfaction (Berry and Parasuraman 1991), and
types of mismatches are common across various ongoing profitability (Kalwani and Narayandas 1995). The main
exchanges? Which ones significantly impact the relation- thrust of research to date generally assumes that firms
ship as a whole? When should a mismatch be overlooked should be in the process of shifting from arm’s length
versus deemed insurmountable? Given these questions dealings to building close relationships (Colgate and
and current difficulties that business relationships face, a Danaher 2000). Finally, research generally takes a seller’s
deeper look is warranted. “push” perspective in attempting to retain customers,
erect switching barriers, and build profitability.
A more granular understanding of the buyer-seller
mismatch phenomenon could provide additional insight Despite substantial attention directed towards devel-
into understanding why relationships succeed or fail. oping buyer-seller relationships, current signs point to
Thus, a primary question needing further development is instability in relationships, continued customer attrition,
this: how might a deeper look into the nature of mis- and termination of partnerships (Wathne, Biong, and
matches expand upon previous conceptions of relation- Heide 2001). Blatant drawbacks to the relationship ap-
ship incompatibility and provide a clearer window into proach have drawn criticism including potential for: hid-
the true complexity of buyer-seller relationships? den costs, negative return on investment (Gummesson
1997), and other negative outcomes. Further, despite the
BUYER-SELLER RELATIONSHIP popularity of relationship strategies, many firms continue
PERSPECTIVE to rely on competitive, transactional dealings (Cannon
and Perreault 1999). One key learning from this discus-
Developing buyer-seller relationships has been rec- sion is the realization that, prior to embarking upon
ognized by firms as an integral component to overall relationship strategies, greater consideration should be
strategy for many years (Cannon and Perreault 1999). given to the resources needed and potential risks. Doing
Specifically, firms now see customer retention and man- so might prevent wasted resources, and ultimately, messy
agement of relationships as a key to fending off rivals, relationship failures. So, while the potential benefits of
who constantly seek to attract customers away (Day relationships are clear, the net recommendation is for
2000). From a financial perspective, managers covet the firms to pursue plural strategies that employ both transac-
performance gains that research suggests can be achieved tional and relational approaches relative to varying situa-
through retention (Noordewier, John, and Nevin 1990). tions, customers, and environments (Frazier 1999;
Furthermore, market drivers such as elimination of middle- Garbarino and Johnson 1999).
men, global competition, technological innovation, and
supply-chain coordination provide both the motive and However, this shift is not easily accomplished. Many
means for fostering close relations (Sheth and Parvatiyar agree that existing theory is not sensitive to situational
2000). factors (Wilson 2000). Beyond this, it rarely allows for a
much needed dyadic perspective. Though a supplier might
Alongside the flurry of relationship activity in the see a particular customer as ripe for relationship building,
business community, a rich tradition of research has the customer may not be responsive. Even more complex,
supported the surge of interest in buyer-seller relation- both parties might agree in principle but simply view
ships. Early works observed business relationships di- relationships in different ways. A supplier may believe a
verging from transactional models towards long-term strong relationship is one that is exclusive. Yet, the re-
associations (Arndt 1979). Research streams have since spective customer might feel it is simply good business
examined traits like trust and commitment (Gundlach, practice to keep suppliers “on their toes” by engaging
Achrol, and Mentzer 1995), dependence (Heide and John competitors. The next evolution, then, of buyer-seller
1990), and potential performance outcomes (Kalwani and theory should adequately address these variations and
Narayandas 1995). Generally, one could argue that the develop theory around the issue of with whom firms
most significant progress thus far has been exploring a should build relationships in given situations.
wide selection of relationship-relevant constructs. These
constructs serve as key variables in the process of devel- BUYER-SELLER MISMATCHES
oping and maintaining relationships.
The term mismatch in business relationships was first
Upon examining the development of buyer-seller used by Pels, Coviello, and Brodie (2000), who sought to
relationships, at least three themes emerge. First, an over- integrate transactional and relational strategies into a
whelming bulk of research has sought to explain how plural approach. They posit that buyers and sellers often
good relationships are built. Second, researchers have operate under mismatched exchange orientations. Ex-
given a great deal of attention to linking relationships to change orientations, they theorize, are overarching ways
positive outcomes such as lower costs (Noordewier, John, of dealing with another party and arise from specific
and Nevin 1990), greater control (Dwyer, Schurr, and Oh relationship needs. In other words, relationship needs

give rise to a party’s pursuit of either transactional or specific discrepancies that occur across a variety of ongo-
relational dealings. Thus, a mismatch of exchange orien- ing exchanges.
tation occurs when one party elects a transactional mode
and the other operates under a relational mode. In addition to these examples, the concept of buyer-
seller mismatches can be traced from a variety of sources
Similarly, Gronroos (1997) asserted that customers such as mismatches in relationship values, mismatches of
operate in one of three modes: a transactional mode, a relationship desire, behavioral and social mismatches
passive relational mode, or an active relational mode. leading to relationship failures, and relationship dysfunc-
Mapping these customer modes to a seller’s relational or tions, to name a few (Table 1). In reviewing this literature,
transactional approach, he argues that dollars spent by several concepts emerge. For one, the discussion on
sellers in relationships where modes do not match are mismatches and related topics has occurred as an after-
wasted efforts. These two perspectives provide a baseline thought to the excitement around building relationships.
understanding of buyer-seller mismatches and are illus- This might be attributed to a realization in the last decade
tratively condensed into Figure 1. This figure serves as an that many relationships fail because of relationship mis-
overarching picture of the phenomenon, yet does not matches. Additionally, up to this point, research address-
distinguish the reality that many mismatches can occur at ing mismatches has limited itself to high-level descrip-
different levels in a relationship. One other comparable tions of the phenomenon and has not explored integrative
example to note is the work of Gassenheimer, Houston, connections to the extant literature on relationships.
and Davis (1998), who introduce the notion of relative
relational distance. They theorize that social and eco- MISMATCHES DEFINED
nomic value discrepancies between parties create rela-
tional distance. The result of this distance is that parties Although the concept of mismatches has been dis-
evaluate social and economic value discrepancies by a cussed both explicitly and implicitly in relationship litera-
comparison of alternatives and subsequently decide to ture, no formal definition has been offered. This work
tolerate the differences or exit the relationship. While this seeks to expand upon previous perspectives (Gronroos
perspective is dyadic, relational distance is modeled as a 1997; Pels, Coviello, and Brodie 2000) of buyer-seller
“summary disposition” (p. 335) and does not attend to the mismatches and integrate ideas from literature that speaks
FIGURE 1
Buyer-Seller Mismatches
Buyer Exchange Paradigm

Relational Transactional
Relational
Seller Exchange
Paradigm
Transactional

American Marketing Association / Winter 2005
TABLE 1
The Concept of Mismatches in Relationship Literature
Topic As It Relates to Mismatch What Is Specifically Mismatched? Illustrative Author(s)
1. Mismatch of Exchange Orientation – Seller’s Exchange Orientation vs. Buyer's Exchange Orientation – Gronoos 1997; Pels, Coviello, and Brodie 2000
– Seller’s Offer and Buyer's Need
2. Mismatch of Desire for Relationship – Seller’s Motivation for Relationship vs. Buyer's Motivation for Relationship – Barnes 1997; Day 1994;Garbarino and Johnson 1999;
Sollner 1999
3. Mismatch of Relationship Values – Seller’s Values vs. Buyer’s Values – Gassenheimer, Houston, and DAvis 1998
4. Behavioral or Social Mismatch – Seller’s Relationship Investment vs. Buyer’s Relationship Investment – Krapfel, Salmond, and Spekman 1991
Leading to Relationship Failure – Seller’s Time Orientation vs. Buyer's Time Orientation – Garbarino and Johnson 1999
– Seller’s Collaborative Orientation vs. Buyer’s Collaborative Orientation – Garbarino and Johnson 1999
– Seller’s Level of Dependence vs. Buyer’s Level of Dependence – Gundlach and Cadotte 1994; Keep, Hollander, and
Dickinson 1998; Kumar, Scheer, and Steenkamp 1995
– Seller’s Level of Trust vs. Buyer’s Level of Trust – Kumar, Scheer, and Steenkamp 1996; Sirdeshmulch,
Sinah, and Sabol 2002
– Seller’s Levelof Information Sharing vs. Buyer’s Level of Information Sharing – Keep, Hollander, and Dickinson 1998
5. Mismatch of Marketing Tools – Seller’s Marketing Tools vs. Buyer’s Exchange Orientation – Jackson 1985
6. Unexpected Costs Leading to – Seller’s Expectations vs. Relationships Ability to Meet Financial Goals – Cannon and Homburg 2001; Cannon and Narayandas
Relationship Failure 2000; Dwyer, Schurr, and Oh 1987; Gummeon 1997;
Jackson 1985
7. Relationship Disfunctions – Buyer Seller Expectations vs. Inherent Drawbacks of relationships – Wilkinson and Young 1998; Hakansson and
Senhota 1995; Colgate and Danaher 2000; Walter,
Ritter, and Gurmenden 2001
8. Prerequisites for Relational Strategy – Seller Expectations vs. Actual Readiness for Relationship Strategy – Berry 1995; Bitner 1995; Colgate and Danaher 2000;
Gordon McKeage, and Fox 1998; Gronroos 1995;
Payne and Ballantyne 1991; Pine, Peppers, and
Rogers 1995; Sheth and Parvatiyar 1995
50
to mismatches (Table 1). As such, the following serves as (Sheth and Parvatiyar 2000). These goals play an active
an initial definition for buyer-seller mismatches: role in shaping relationship needs. When buyers and
sellers have incompatible relationship goals, they result in
Buyer-Seller mismatches are composed of active or mismatches.
latent differences in parties’ perceived needs and wants
from a business relationship, as shaped by situational Reflected in Incompatible Interfirm Relations
factors and dissimilar goals, that are ultimately reflected
in incompatible interfirm relations. Borrowing from the Nordic school of relationship
theory, mismatches are reflected in ongoing interaction
Variety of Mismatches: Inherent in this conceptual- between buyers and sellers (Gronroos 1997). Mismatches
ization is the idea that a variety of mismatches can occur become manifest in actions that buyers and sellers take or
in relationships. Business relationships can be complex fail to take towards each other, resulting in an unsatisfied
because a range of exchanges may take place. This com- need or want across a particular relationship dimension.
plexity allows for multiple layers of interaction across These differences can be assessed at a high level (like in
product/service exchanges, partnership investments (Rok- previous research) as one firm’s general desire for rela-
kan, Heide, and Wathne 2003), information exchange tional dealings versus the other party’s general desire for
(Anderson and Weitz 1992), social norms (Cannon, Ach- transactional dealings (see Figure 1). It is suggested here,
rol, and Gundlach 2000), operational linkages, and many however, that desired needs and wants from the relation-
others. So, while an overall mismatch of transactional ship drive the exchange orientation. Thus, mismatched
versus relational exchange orientations might exist, a exchange orientations serve as a proxy for a variety of
variety of mismatches can occur that may or may not underlying mismatched needs and wants. Ultimately,
significantly impact the relationship as a whole. This mismatches become salient, take shape, and result in
represents a departure from previous notions of a solitary incompatible dealings.
overarching match or mismatch.
To illustrate, buyers can be confronted with a seller
Perceived Needs and Wants who constantly attempts to raise the bar on the commit-
ment of the relationship, when in fact, the buyer senses no
Consistent with theory applied to relationships (e.g. return on this investment. While this seller might see
resource dependence, TCA, social exchange), parties potential growth in getting “deeper” into the customer
enter relationships with the expectation that doing so will relationship, they are met with buyer apathy. Alterna-
be rewarding (Berry and Parasuraman 1991). Buyers and tively, buyers can be frustrated in relationships where they
sellers pursue a range of needs and wants from the other continually vie for attention from a seller who is “busy
party. Within the context of mismatches, buyers and with other opportunities.” These differences in perceived
sellers can strive for needs that the other party is ultimately needs and wants are played out and result in mismatches.
unable or unwilling to provide. Mismatched needs can To further explore the connection between mismatched
occur across (1) homogenous relationship dimensions needs and wants within exchange orientations, the follow-
(i.e., common to both parties) or (2) heterogeneous rela- ing discusses buyer-seller tradeoffs associated with trans-
tionship dimensions, such as supplier dependability in actional and relational dealings.
exchange for an exclusive relationship. Finally, research
suggests that desired needs and wants can be latent or BUYER-SELLER EXCHANGE TRADEOFFS
active as the situations change (Gronroos 1997).
Organizations, for the most part, make intentional
Shaped by Situations and Goals decisions to establish relationships and have particular
goals in mind (Oliver 1990). Furthermore, they do so in
Several authors acknowledge that relationship theory ways that maximize rewards from these relationships
has been generally insensitive to situational factors that (Frazier 1983). Many have explored various goals and
shape relationships (Coviello et al. 2002). Mismatches tradeoffs for suppliers and customers arising from a
provide a framework for including them. For example, a transactional or relational exchange. For example, many
multitude of situational conditions in the macro and task posit that relationship strategies can lead to competitive
environment and firm characteristics can shape buyer- advantage (Day 2000), yet others suggest maintaining
seller mismatches such as: environmental factors (Keep, close relationships can also prove very costly (Sharma
Hollander, and Dickinson 1998), product-market charac- and Tzokas 1999). An integrated review of these benefits
teristics (Crosby and Evans 1990), firm strategy (Wilson and sacrifices as they map to both a transactional and
2000), and risk orientation. Relationship goals also shape relational orientation is presented in Table 2. It is sug-
buyer-seller needs and wants. Firms enter relationships gested that understanding these factors presents insight
for the purpose of fulfilling strategic goals like entering into why buyers and sellers might pursue incompatible
new markets and serving customer needs to name a few needs and wants from relationships.

TABLE 2
Buyer-Seller Exchange Tradeoffs
Buyer – Transactional Orientation Buyer – Relational Orientation

Benefits Gained / Sacrifices Avoided Associated with Transactional Orientation Benefits Gained / Sacrifices Avoided Associated with Relational Orientation
* Avoidance of Switching Costs, Sunk Costs e,f,m,p Access to Technology g,p,ag

High Avilability of Alternatives a,g,p‘ Access to Value-Added Services h,d
* High Leverage (Low Vulnerability) with Suppliers, l,p,ac,ah,ak Conflict Resolution More Manageable p
Less Hassle / Low Commitment j,aa * Cost Reduction through Synergies b,g,p,r,z,ag,ak
* Low Price through Adversarial Bidding a,g,p,aj Efficiency / Effectiveness for Value Chain p,ak
Greater Customization of Products/Services e,d,ad,ag
Higher Control of Direction of the Relationship ag
* Higher Relationship Satisfaction c,k,n,x,ab,ab,ac,ah
Greater, Specialized Service Levels h,w,ad
Leads to Economies of Scale s
Leads to Opportunities for Competitive Advantage ad
Outsourcing Benefits, Ability to Focus Energies Elsewhere d
Reducation of Choices / Efficient Decision Making d,ad
Risk Reduction in Crisis Situations s
Shortened New Product Development Time g,p
Seller – Transactional Orientation Seller – Relational Orientation

Benefits Gained / Sacrifices Avoided Associated with Transactional Orientation Benefits Gained / Sacrifices Avoided Associated with Relational Orientation
* Avoidance of High Service Costs, Investments, Adaptation n,o,t,v,aa,ac Cost Reduction through Synergies p,r,u,ac,ag
Avoidance of Overdependence, Relationship Burdens p,r,ae Customer Learning p,ag
Greater Ability to Focus on Acquiring New Customers j,i * Customer Retention, Satisfaction, Loyalty c,k,n,x,y,ac,ah
Less Hassle / Low Commitment j,o,aa Optimized Focus of Resources, Retention j,i.ag
Leads to Economies of Scale s
* Leads to Opportunities for Competitive Advantage j,u,ad,ai
Greater Price Stability p,ac
Risk Reduction / Higher Leverage with Buyers ae
Sales / Profitability Outcomes j,ac,af
* Items have received greated attention in Relationship Literature
a amihud 1976 k Garbarino and Johnson 1999 u Krapfel, Salmond, and Spekman 1991 ne Sollner 1999
b Anderson 1995 l Gasseheimer, Houston, and Davis 1998 v LaBahn and Krapfel 2000 af Spekman 1988
c Berry and Parasuraman 1991 m Billiland and BEllo 2002 w Lambert, Emmelhainz, and Gardner 2000 ag Sriram, Krapfel, and Spekman 1992
d Bitner, Gwinner, and Gremler 1998 n Golicic, Foggin, and Mentzner 2003 x Leuthesser and Kohli 1995 ah Webster 1992
e Brennan and Turnbull 1999 o Gronroos 1997 y Morgan and Hunt 1994 ai Weitz and Jap 2000
f Burnham, Frels, and Mahajan 2003 p Han and Wilson 1993 z Noordeweier, John, and Nevin 1990 aj Williamson 1985
g Cannon and Homburg 2001 q Helper and Levine 1992 aa Pels, Coviello, and Brodie 2000
h Cannon and Narayandas 2000 r Kalwani and Narayandas 1995 ab Sharma and Grewal 1995
i Child and Dennis 1995 s Keep, Hollander, Dickinson 1998 ac Sharma and Tzokas 1999
j Day 2000 t Krapfel 1991 ad Sheth and Sharma 1997
52
A MODEL OF BUYER-SELLER MISMATCHES An economic mismatch could include conflict over
preferred discounts or length of contract. A social
Given the newness of the concepts presented in this mismatch might arise from asymmetrical levels of
paper, the model (Figure 2) is a framework to guide theory trust. Form also explains whether the mismatch oc-
development and testing of buyer-seller mismatches. curs across a homogenous relationship dimension
(e.g., frequency of interaction) or a heterogeneous
Situational Conditions one (e.g., buyer’s need for reliability vs. seller’s push
for exclusivity).
The model assumes a context shaped by situational
conditions that emanate from the macro and task environ- 2. Salience: This refers to whether the mismatch is
ment, as well as both party’s characteristics. This eleva- latent or active for either party.
tion of situational conditions lends greater understanding
to the types of relationships that can feasibly form in a 3. Priority: Mismatches relative to goals, needs, and
given context. Conditions act to motivate, as well con- wants can be assigned a certain priority, which could
strain relationship development due to incentives and be perceived differently by the parties. A high prior-
deterrents arising from the context. In the model, situa- ity mismatch could be a shift in strategic direction that
tions shape buyer-seller goals, needs and wants from significantly alters the relationship structure. A low
relationships; mismatches that take place; comparison priority mismatch might be a personality conflict
level of alternatives; and resulting outcomes. between two boundary personnel.
Buyer and Seller Relationship Goals, Needs, and Wants 4. Extensiveness: The extensiveness of a mismatch
describes the degree to which incompatibilities exist
These components shape what is sought in a relation- at an individual level or organization. A seller may
ship and are where mismatches of needs and wants occur. determine that a blocked need for closer interaction
A disproportionate focus in the literature has discussed within a customer’s firm is localized around a par-
buyer needs and wants from a relationship, however, it is ticular gatekeeper. Buyers might see the overall seller’s
asserted that inclusion of seller’s needs and wants pro- offer (organization) falling short of its needs.
vides a more holistic picture.
5. Magnitude: Mismatches take on relative magnitudes
Buyer and Seller Exchange Orientations based on the perceived gap size.
The assumption here is that exchange orientations 6. Temporal Nature: Parties to a relationship may learn
arise out of need structures (Pels, Coviello, and Brodie to cope with certain mismatches over a long period of
2000). Changes in need structure might drive changes in time, while others require immediate action.
exchange orientation. If a seller sees a need to retain a
customer whose past interaction has remained arm’s length, Perceived Mismatches and Comparison Level of Al-
that seller might begin utilizing relationship tactics, thus ternatives (CL alt)
altering the exchange orientation. A buyer who desires
“less hassle” in a relationship might enact a transactional The model proposes two basic outcomes of mis-
orientation to ward off unnecessary use of time. Thus, it is matches. First, it suggests that, parties perceive mis-
suggested firms do not adopt some degree of transactional matches relative to their situation, goals, needs, and wants.
or relational orientation for its own sake. Rather, orienta- Second, they judge these mismatches against a compari-
tions arise as a perceived means for achieving desired son level of alternatives, hereafter CL alt (Thibaut and
needs and wants from the relationship. Kelly 1959). A significant stream of research has used the
concept of CL alt to study relationships (Wilson 2000),
Buyer-Seller Mismatches i.e., parties continually evaluate outcomes against desired
needs and wants. CL alt is the lowest level of outcomes
When considering the host of mismatches that can that a party will accept in light of other available alterna-
occur within a relationship, understanding their traits adds tives. This standard serves as the deciding factor for
insight towards classification of the phenomenon. The whether the party will cope with the outcomes or seek
following traits within the model are considered an initial alternatives (Thibaut and Kelley 1959).
set of aspects to define a mismatch. It is not assumed that
this list is exhaustive, as further inquiry will likely refine CONCLUSIONS
and/or add to this set.
Given the instability of many relationships, research
1. Form: Mismatches can take different forms. A simple recommends buyers and sellers move toward plural ap-
typology might include economic and social needs. proaches that blend relational and transactional strategies.

FIGURE 2
Buyer-Seller Mismatch and Mismatch Outcomes
Situational Conditions (Macro and Task Environment, Firm Characteristics)
Set of Buyer-Seller
Mismatches
Form
Salience
Buyer Seller
Priority
Goals, Needs, and Wants Goals, Needs, and Wants
Extensiveness
from Relationship Magnitude from Relationship
Temporal Nature
Buyer Exchange Orientation Seller Exchange Orientation

Perceived Mismatches
No No
Coping/Maintain Coping/Maintain
Below Comparison Below Comparison

Level of Alternatives Level of Alternatives
(CL alt)? (CL alt)?
Yes
Seek Alternatives Yes Seek Alternatives
54
Yet, making this move is not easy. Firms struggle with targeting particular customers for Key Account Manage-
determining strategies for particular situations and an- ment (KAM) programs and general stratification for sales
swering the question of with whom close or arm’s length support, and other services. Sellers might also use an
relationships make sense. On one side, firms might miss awareness of mismatches to pick up on customers’ chang-
the benefits of a close relationship. Conversely, some ing needs and wants (Flint, Woodruff, and Gardial 2002).
firms dive into relational strategies and waste significant Buyers might utilize mismatch information as input into
resources when relationships do not pan out. Further, balanced score cards for rating suppliers, or attempt to
many agree that existing research does not adequately build expertise within their processes to address mis-
account for diverse situations and customers (Cannon and matches. For relationships on the brink of dissolution, a
Perreault 1999; Coviello et al. 2002; Wilson 2000). granular understanding of the mismatches involved could
aid parties in understanding whether the relationship can
Understanding relationship incompatibility from the be reconciled or is beyond repair. Finally, by understand-
perspective of mismatches, casts a wider net that allows ing mismatches in the wake of a failed relationship, firms
for a more comprehensive view of both parties’ desires can identify issues to be addressed in order to prevent
and a deeper look at situations that influence success or future occurrences.
failure. Key contributions offered from this discussion are
threefold. First, a brief review of relationship literature Knowledge about how firms react to mismatches is
identifies the need for theory to offer a dyadic perspective an area where little is known. How might a firm persuade
and greater sensitivity to situational factors. Current re- the other party to close the gap on a clear mismatch? When
search is largely limited to a single firm perspective and might firms adjust their own goals to accommodate mis-
oversimplifies the concept of incompatibility. This paper matched needs? How might mismatches relate to con-
puts forth the first known attempt to integrate the mis- cepts of relational equity and return on relationship mea-
match concept to directly address these issues. Second, a surements? Research addressing costs and benefits for
detailed definition, with discussion of relevant variables relationships with high versus low levels of mismatch
and refinement of existing discussion on mismatches is might provide interesting insights for relational strategy
offered. This work builds on previous studies to theorize decisions. This paper began by asking if a deeper look into
that incompatible needs and wants lie at the core of the nature of mismatches could expand upon ideas of
mismatches. Finally, a model of buyer-seller mismatches relationship incompatibility and offer a clearer window
is proposed, including antecedents, aspects of the phe- into the true complexity of buyer-seller relationships. The
nomenon, and potential consequences around mismatch integrated literature on relationships and mismatches sug-
outcomes. gests that it does, but further work is needed to answer this
question more comprehensively. Prior to addressing con-
Managers are urged to appreciate the concept of struct measurement, initial efforts should focus on further
mismatches as they occur in relationships. The concept of development of the mismatch construct, ensuring concep-
mismatches forces managers to look deeper at both sides tual thoroughness and refinement of the variables dis-
of the equation. This might help firms both avoid wasted cussed herein. Qualitative inquiry seems the logical next
resources and optimize relationship investments. Within step for development.
existing relationships, mismatches could guide sellers in
ENDNOTE Arndt, J. (1979), “Toward a Concept of Domesticated

Markets,” Journal of Marketing, 43 (4), 69–75.
1
The author thanks Dr. John T. Mentzer for his valuable Berry, Leonard L. and A. Parasuraman (1991), Marketing
comments on a previous version of this paper. Services – Competing Through Quality. New York:
Free Press.
REFERENCES Cannon, J.P. and W.D. Perreault (1999), “Buyer-Seller
Relationships in Business Markets,” Journal of Mar-
Anderson, E. and B. Weitz (1992), “The Use of Pledges keting Research, 36 (4), 439–60.
to Build and Sustain Commitment in Distribution Cannon, Joseph P., Ravi S. Achrol, and Gregory T.
Channels,” Journal of Marketing Research, 29 (1), Gundlach (2000), “Contracts, Norms, and Plural
18–34. Form Governance,” Journal of the Academy of Mar-
Anderson, James C. and James A. Narus (1991), keting Science, 28 (2), 180–95.
“Partnering as a Focused Market Strategy,” Califor- Colgate, Mark R. and Peter J. Danaher (2000), “Imple-
nia Management Review, 33 (3), 95–113. menting a Customer Relationship Strategy: The Asym-

metric Impact of Poor Versus Excellent Execution,” (1995), “The Structure of Commitment in Exchange,”
Journal of the Academy of Marketing Science, 28 (3), Journal of Marketing, 59 (1), 78–92.
375–87. Heide, Jan B. and George John (1990), “Alliances in
Coviello, Nicole E., Roderick J. Brodie, Peter J. Danaher, Industrial Purchasing: The Determinants of Joint
and Wesley J. Johnston (2002), “How Firms Relate to Action in Buyer-Seller Relationships” Journal of
Their Markets: An Empirical Examination of Con- Marketing Research (JMR), 27 (1), 24–36.
temporary Marketing Practices,” Journal of Market- Kalwani, M.U. and N. Narayandas (1995), “Long-Term
ing, 66 (3), 33–47. Manufacturer Supplier Relationships – Do They Pay
Crosby, L.A. and K.R. Evans (1990), “Relationship Qual- Off for Supplier Firms?” Journal of Marketing, 59
ity in Services Selling: An Interpersonal Influence (1), 1–16.
Perspective,” Journal of Marketing, 54 (3), 68–82. Keep, William W., Stanley C. Hollander, and Roger
Das, T.K. and Bing-Sheng Teng (2000), “Instabilities of Dickinson (1998), “Forces Impinging on Long-Term
Strategic Alliances: An Internal Tensions Perspec- Business-to-Business Relationships in the United
tive,” Organization Science: A Journal of the Insti- States: An Historical Perspective” Journal of Mar-
tute of Management Sciences, 11 (1), 77–101. keting, 62 (2), 31–45.
Day, George S. (2000), “Managing Market Relation- Kotler, Philip (1991), “Phillip Kotler Explores the New
ships,” Journal of the Academy of Marketing Sci- Marketing Paradigm,” Review Marketing Science
ence, 28 (1), 24–30. Newsletter, 1 (Spring), 4–5.
Dwyer, F.R., P.H. Schurr, and S. Oh (1987), “Developing Newell, Frederick (2003), Why CRM Doesn’t Work.
Buyer-Seller Relationships,” Journal of Marketing, Princeton: Bloomberg Press.
51 (2), 11–27. Noordewier, Thomas G., George John, and John R. Nevin
Flint, Daniel J., Robert B. Woodruff, and Sarah Fisher (1990), “Performance Outcomes of Purchasing Ar-
Gardial (2002), “Exploring the Phenomenon of Cus- rangements in Industrial Buyer-Vendor Relation-
tomers’ Desired Value Change in a Business-to- ships,” Journal of Marketing, 54 (4), 80.
Business Context,” Journal of Marketing, 66 (4), Oliver, Christine (1990), “Determinants of Interorganiza-
102–17. tional Relationships: Integration and Future Direc-
Frazier, Gary L. (1983), “Interorganizational Exchange tions,” Academy of Management Review, 15 (2),
Behavior in Marketing Channels – A Broadened 241–66.
Perspective,” Journal of Marketing, 47 (4), 68–78. Pels, Jaqueline, Nicole E. Coviello, and Roderick J. Brodie
Frazier, Gary L. (1999), “Organizing and Managing Chan- (2000), “Integrating Transactional and Relational
nels of Distribution,” Journal of the Academy of Marketing Exchange,” Journal of Marketing Theory
Marketing Science, 27 (2), 226–41. & Practice, 8 (3), 11–20.
Garbarino, Ellen and Mark S. Johnson (1999), “The Reinartz, Werner J and V. Kumar (2003), “The Impact of
Different Roles of Satisfaction, Trust, and Commit- Customer Relationship Characteristics on Profitable
ment in Customer Relationships,” Journal of Mar- Lifetime Duration,” Journal of Marketing, 67 (1),
keting, 63 (2), 70–87. 77–99.
Gassenheimer, Jule B., Franklin S. Houston, and Charlene Rokkan, Aksel I., Jan B. Heide, and Kenneth H. Wathne
J. Davis (1998), “The Role of Economic Value, (2003), “Specific Investments in Marketing Rela-
Social Value, and Perceptions of Fairness in tionships: Expropriation and Bonding Effects,” Jour-
Interorganizational Relationship Retention Deci- nal of Marketing Research (JMR), 40 (2), 210–14.
sions,” Journal of the Academy of Marketing Sci- Sharma, Arun and Nikolas Tzokas (1999), “Antecedents
ence, 26 (4), 322–37. and Consequences of Relationship Marketing,” In-
Gronroos, Christian (1994), “From Marketing Mix to dustrial Marketing Management, 28 (6), 601–11.
Relationship Marketing: Towards a Paradigm Shift Sheth, Jagdish N. and Atul Parvatiyar (2000), “The Evo-
in Marketing,” Management Decision, 32 (2), 4–20. lution of Relationship Marketing,” in Handbook of
____________ (1997), “Value-Driven Relational Mar- Relationship Marketing, Jagdish N. Sheth and Atul
keting: from Products to Resources and Competen- Parvatiyar, eds. Thousand Oaks, CA: Sage Publica-
cies,” Journal of Marketing Management, 13 (5), tions Inc.
407–19. Siguaw, Judy A., Thomas L. Baker, and Penny M. Simpson
Gummesson, Evert (1997), “Return on Relationships,” in (2003), “Preliminary Evidence on the Composition
5th International Colloqium on Relationship Market- of Relational Exchange and its Outcomes: The Dis-
ing. Cranfield, U.K. tributor Perspective,” Journal of Business Research,
____________ (2004), “Return on Relationships (ROR): 56 (4), 311–22.
The Value of Relationship Marketing and CRM in Thibaut, John W. and Harold H. Kelley (1959), The Social
Business-to Business Contexts,” The Journal of Busi- Psychology of Groups. New York: John Wiley &
ness & Industrial Marketing, 19 (2), 136–48. Sons, Inc.
Gundlach, Gregory, Ravi Achrol, and John T. Mentzer Wathne, Kenneth H., Harald Biong, and Jan B. Heide

(2001), “Choice of Supplier in Embedded Markets: 17.
Relationship and Marketing Program Effects,” Jour- Wilson, David T. (2000), “An Integrated Model of Buyer-
nal of Marketing, 65 (2), 54–66. Seller Relationships,” in Handbook of Relationship
Webster, F.E. (1992), “The Changing-Role of Marketing Marketing, Jagdish N. Sheth and Atul Parvatiyar,
in the Corporation,” Journal of Marketing, 56 (4), 1– eds. Thousand Oaks, CA: Sage, Inc.

Christopher P. Blocker
University of Tennessee
304 Stokely Management Center
Knoxville, TN 37996–0530
E-Mail: cblocker@utk.edu

CONSUMER TRUST NORMS IN MULTI-CHANNEL FIRMS: THE ROLE
OF TRUST IN TECHNOLOGY AND FIRM COMMITMENT TO PRIVACY
PROTECTION IN TECHNOLOGY-BASED SERVICE DELIVERY
Devon Johnson, Northeastern University, Boston
SUMMARY protection. Trust in the firm, frequency of transactions,

and customer-company identity are examined as conse-
Although web-based self-service technology deliv- quences of the two trust norms. The model is evaluated
ers convenience benefits to customers and transaction using combined survey and transaction data from 834
cost savings to firms, it also heightens consumers’ con- personal computer (PC) banking customers of credit
cerns about the privacy of personal identity and contact union.
information. These concerns are perhaps salient due to the
absence of tangible and verifiable cues regarding a seller’s The results reveal that trust in technology and per-
intentions and capabilities (Balasubramanian, Konana, ceived commitment to privacy protection play an essen-
and Menon 2003). Information security and technology tial mediating role between the aforementioned anteced-
concerns have been recognized as essential to consumer ent and consequence variables. Interestingly, overall trust
evaluations of a Web a site’s trustworthiness. Trust in a in the firm does not mediate the impact of these trust
web site is therefore regarded as critical to web site norms on the consequence variables, i.e., commitment to
success and a key focus of competitive differentiation privacy and trust in technology directly impact customer-
(Urban, Sultan, and Qualls 2000). company identity and frequency of transaction, respec-
tively. The findings indicate that managers and research-
This article addresses a gap in the literature by pro- ers need to consider cross-channel effects of consumer
posing two new facets of consumer trust in a technology- evaluation of other firm channels in evaluating the trust-
based service firm: trust in technology and perceived worthiness of a web site. Managers also need to explore
commitment of the firm to privacy protection. The study ways of demonstrating and reinforcing firm identity ide-
proposes a model that demonstrates the importance and als in technology-based service delivery. Finally, the
role of these two constructs in the wider nomological study suggests that rather than regard service delivery
network of consumer-multi-channel service firm rela- technology as a generic invisible black box, managers
tionship. Affective commitment, operational benevolence, should communicate the capabilities and integrity of its
and problem solving orientation are modeled as anteced- service delivery technology to customers.
ents of trust in technology and commitment to privacy

Devon Johnson
Northeastern University
Boston, MA 02115–5000
Phone: 617.373.3549
FAX: 617.373.8366
E-Mail: d.johnson@neu.edu

VALIDATION AND APPLICATION OF A BI-DIMENSIONAL
LONG-TERM ORIENTATION SCALE
William O. Bearden, University of South Carolina, Columbia
R. Bruce Money, Brigham Young University, Provo
Jennifer L. Nevins, University of South Carolina, Columbia
SUMMARY to address the shortcomings of extant measures, Bearden

et al. (2003) developed a measure of long-term orientation
Time orientation exists as a ubiquitous influence that composed of eight items in two subscales of four items
permeates many aspects of life for every individual, each designed to address tradition and planning aspects of
whether in a personal or business-related context. As LTO. In the current research we evaluate this two-factor
such, marketing researchers have long been interested in LTO scale with data collected from two samples of U.S.
how time is valued by consumers and decision makers in (n = 339) and Japanese (n = 253) adults (study 1).
business settings. Measures of long- versus short-term
time orientations have been the subject of recent inquiry Validation and Equivalence
regarding their conceptualization, measurement and lev-
els of analysis. The research reported here involves an From the U.S. and Japanese data, a uni-dimensional
assessment and validation of a multi-factor set of items model in which all eight items loaded on one factor, a two-
designed to measure long-term orientation at the indi- factor uncorrelated model, and a two-factor correlated
vidual level using data from a study involving 592 Ameri- model, in which the two sets of four items loaded on their
can and Japanese adult respondents. In addition, two respective tradition and planning factors, were estimated.
application studies investigate the predictive validity of Comparisons of these models (cf., Anderson and Gerbing
the long-term orientation measures in consumer and mana- 1988) across both country analyses revealed that the two-
gerial decision contexts. factor correlated model provided the best representation
of the data. Overall, these results provide evidence of
Time Orientation Use and Measurement discriminant validity for the separate factor LTO subscales
(tradition and planning) and are consistent with the previ-
Hofstede’s (1980) work on national cultural values ous results reported by Bearden et al. (2003). Addition-
has generated numerous citations and discussion, as well ally, we examined the measurement equivalence of the
as criticism (Oyserman et al. 2002; Spector et al. 2001, data following the procedures outlined by Singh (1995)
2002). Specifically, Long-Term Orientation (LTO), or and Steenkamp and Baumgartner (1998), who recom-
Confucian Dynamism, as Hofstede’s fifth dimension was mend tests of configural equivalence and metric equiva-
originally called, has been the subject of recent inquiry lence, both of which exhibited satisfactory levels of fit.
regarding the conceptualization, measurement, and levels These results support the cross-cultural applicability of
of analysis this dimension was designed to capture (Fang the LTO scale.
2003). The lack of clarity regarding the underlying values,
as well as confusion surrounding the dimensionality and Additional Evidence
polarity of the LTO construct has contributed to increas-
ing scrutiny of this cultural dimension. Responses from the U.S. and Japan were also col-
lected for the eight-item consumer frugality scale of
Hofstede’s fifth dimension has come to be used as a Lastovicka et al. (1999), a four-item measure of personal
measure of time orientation. Indeed, attitudes toward time ethics (Vitell, Rallapalli, and Singhapakdi 1993), and a
have been shown to vary systematically among consum- twelve-item consumer compulsive buying scale (Valence,
ers (Prelec and Loewenstein 1998), business/marketing d’Astous, and Fortier 1988). It was hypothesized that
managers (Kim and Oh 2002) and national cultures (Gra- individuals scoring high in the tradition and planning
ham 1981). The concept of LTO has been used in a facets of long-term orientation would score higher in
number of international marketing studies (Nakata and terms of consumer frugality, and ethics, and lower in
Sivakumar 1996), with many indicating the need for compulsive buying tendencies. In addition to these other
improved measurement. Some of these studies either constructs, single item convergent validity items for both
empirically or conceptually correlated Hofstede index of the proposed LTO scales were also assessed using the
scores for the countries involved, but could have benefit- procedures recommended by Bagozzi (1993). A series of
ted from better individual-level measurement. In an effort consistently significant (p < .01) and generally strong

estimates between the two tradition and planning subscales term outcomes related to firm profits and market share.
and these additional measures provided supportive evi- Following completion of the choice task, respondents
dence of validity for the proposed LTO measures. then replied to an item designed to elicit their confidence
in their decision, as well as the tradition and planning LTO
Application Studies measures. While the correlations between the tradition
and planning measures and the dichotomous choice vari-
In addition to study 1 assessments of equivalence and able were modest, the estimates are positive and signifi-
validity, two additional studies were conducted to inves- cant as expected and provide additional support for the
tigate the extent to which the proposed LTO measures are proposed scales in terms of predictive validity.
correlated as expected with important consumer out-
comes (study 2) as well as managerial decision making Conclusions and Implications
(study 3). In study 2, data were collected from 54 under-
graduate business students under the ruse of a study of The results reported in this article are generally
consumer credit use among college students. Two mea- supportive of the bi-dimensional measure of long-term
sures of credit involvement were obtained: total average orientation proposed by Bearden et al. (2003). A replica-
monthly credit card balance and the total number of credit tion and application of the proposed LTO measure yielded
cards owned. Although the correlations between the credit evidence of convergent, discriminant, and predictive va-
measures and planning and tradition were mixed, both the lidity, dimensionality, and nomological validity. The im-
simple correlations, as well partial correlation estimates, plications of our research are important to business aca-
for the total eight-item LTO scale and both of the credit demics for several reasons. Our study adds to the body of
measures were significant (p < .10). We also found knowledge regarding culture, and LTO in particular, and
evidence of the ability of our measures to moderate the supports a valuable tool for studying an important aspect
negative relationship between Lastovicka et al.’s (1999) of culture. Rather than merely assign index scores as
consumer frugality scale and the two credit outcome reflective of whole groups, researchers using new scales
variables described above. such as the LTO scale described in the current research,
will be able to investigate outcome phenomena dependent
In study 3, seventy-three MBA students first reacted directly on the individual values of subjects in various
to a choice scenario involving selection of a pricing countries. Citation information available upon request.
strategy that varied in terms of long-term versus short-

William O. Bearden
The Moore School of Business
University of South Carolina
Columbia, SC 29208
FAX: 803.777.6876
E-Mail: bbearden@moore.sc.edu

CONSUMER SOCIALIZATION OF THIRD CULTURE KIDS IN
A COSMOPOLITAN CITY
Alfred Y. Sit, Chinese University of Hong Kong, Hong Kong
Haksin Chan, Chinese University of Hong Kong, Hong Kong
SUMMARY have largely maintained their socializing influence on our

ATCK participants. Often deprived of an extensive net-
work of social support as they frequently relocate, many
Third culture kids (TCKs, children raised outside ATCKs have developed an unusually strong bond with
their passport countries) are a growing segment of global the immediate family.
significance. We conducted an in-depth, qualitative study
on consumer socialization of adolescent TCKs (ATCKs), The consensus in the literature is that peer influence
and the data suggest that the unique cross-cultural expe- is negatively correlated with parental influence. Given
riences of ATCKs are conducive to self-socialization, a ATCKs’ strong family ties, one would expect their peers
phenomenon that has largely been neglected in the mar- to assume a smaller role. Indeed, the strength of peer
keting literature. Our findings highlight ATCKs’ flexible group influence does appear weaker for ATCKs. How-
use of socialization agents, as well as the influence of ever, the diminished peer group influence may also be
ATCKs’ own motives and values. attributable to the mobility of ATCKs. Their nomadic
lifestyle apparently attenuates peer group influence.
Contrary to much of previous research, the adoles-
cent consumer is not a passive recipient but rather an School is the place where adolescents spend most of
active agent in the consumer socialization process. Strad- their time awake. Interestingly, extant consumer social-
dling the home (i.e., first) and host (i.e., second) cultures, ization research has produced little evidence of school
ATCKs are naturally exposed to a wide range of market influence. Note that school life extends far beyond the
information, but they are at the same time quite purposive classroom, and extracurricular activities are an important
as they “pick and choose” from a variety of potential part of school life for many ATCKs. One interesting
consumer socialization agents. Notably, ATCKs’ motives finding is that extracurricular activities afford special
and values play a significant role in the consumer social- shopping and consumption opportunities that otherwise
ization process, as they negotiate the multi-cultural terrain would not be available to the adolescent consumer.
of global markets. From a self-development perspective,
ATCKs strategically engage in selective exposure to Previous research has suggested that TV is a potent
socialization agents and actively construct their own shop- force shaping consumer values. This is, however, not the
ping and consumption scripts. case for the ATCKs in our study. Our participants opt for
the Internet and teen magazines to enrich their market-
Methodology place knowledge of the first culture, even though the first-
culture marketplace is thousands of miles away. Broadly
To conduct a discovery-oriented study addressing speaking, these media are selected because they enable
complex and dynamic issues of consumer socialization, the ATCKs to stay connected to their cultural roots.
we adopted phenomenological interviewing as the method
for data collection. Textual data were generated by means Contrary to past research which suggests that adoles-
of in-depth interviews with 13 English-speaking female cents from higher-income families are more materialistic,
ATCKs (ages 15 to 17) in a cosmopolitan city. The data the well-off ATCKs in our study generally do not consider
were analyzed using a modified meaning condensation themselves to be brand-conscious. Situated in a heteroge-
technique. After establishing an internal structure through neous brand environment in the diverse expatriate com-
an iterative process, we generated meaning categories via munity, ATCKs appear to be less vulnerable to the hege-
collective coding. Finally, inter-relationships among the mony of leading global brands.
categories were established in keeping with the overarching
theme of consumer socialization. Implications
Findings Unlike much of the empirical research on consumer

socialization, our ATCK study has brought the cultural
The ATCKs in our study have very strong ties to the context to the forefront. In light of the encouraging
immediate family. In contrast to previous findings of findings of this study, the door is wide open for follow-up
waning parental influence during adolescence, parents studies on consumer socialization of ATCKs, and for

investigations into the impacts of global marketing on the unyielding local differentiation, given that young con-
socialization of consumers in different cultures. A consumers are active participants in the socialization process.
certed effort in cross-cultural research on consumer so-
cialization would contribute to our understanding of inter- Many ATCKs in our study long for a connection to
national consumer behavior and the broader impacts of their cultural roots while adapting to the second culture.
globalization. Notably, previous research has found that expatriate ex-
ecutives, ATCKs’ adult counterparts, also share this sen-
Some have envisaged a homogeneous youth market. timent. This has important implications for global market-
The socializing influence of global marketing, however, ers targeting the growing expatriate communities around
is mediated by socialization agents. There are hints of the world. Arguably, the marketing and public policy
implications apply to immigrant communities as well.

Haksin Chan
Shatin, Hong Kong
Phone: 852.2609.7637
FAX: 852.2603.5473
E-Mail: hchan@cuhk.edu.hk

CONSUMER ETHNOCENTRISM IN THE GERMAN MARKET
Heiner Evanschitzky, University of Muenster (MCM), Germany
Florian V. Wangenheim, Universität Dortmund, Germany
SUMMARY termed consumer cosmocentrism (CC), which we intro-

duce due to the specific circumstances surrounding etno-
International marketing researchers have for long censtrism in Germany is able to explain DCB better than
been concerned with determining whether consumers are the CE construct.
predisposed towards a preference for domestic as op-
posed to foreign products. Empirical studies have consis- Results
tently confirmed the existence of such as “domestic-
country bias” (DCB), which is manifested in stronger In a German context, we examine the relationship
product preferences and buying intentions for home- between CE and CC on consumer preferences and inves-
made products (for an overview, see Verlegh and tigate whether competitiveness and cultural similarity can
Steenkamp 1999). DCB has typically been explained by help explain the varying strength of CE effects. The
an individually varying, trait like property named “con- inclusion of a new construct, CC, adds little explanatory
sumer ethnocentrism” (CE; Shimp and Sharma 1987). In power to our models. It seems important to verify that CE
brief, the more ethnocentric consumers are, the stronger does in fact explain consumer preferences better than its
the DCB, and consequently, their predisposition to prefer counterpart, even in a culture which is likely to be biased
domestic over foreign products. towards the formulation of the CE scale.
In a recent study, Balabanis and Diamantopoulos In general, it seems that preference rankings can be
(2004) identify a number of weaknesses of prior research better explained by the combination of demographic
linking CE to DCB. First, earlier research had typically variables and CE for Germany than for Britain. The
been restricted to one product category, thereby not allow- highest R 2 -value in the study by Balabanis and
ing researchers to investigate potential variation of DCB Diamantopoulos (2004) is .14 (Cars/Britain), while the
across product categories. Second, previous studies had highest explained variation in the present study is .18
focused on a very limited number of countries of origin (Shoes/Germany). Taken together, the fact that the high-
(COO) of the researched product categories, which in turn est R2-value in both studies is obtained for the home
prevented findings regarding varying degrees of DCB for country suggests that CE is better suited for explaining
different COOs included in the study. Since there are domestic rather than foreign-country bias. Still, the level
some indications that the effect of CE on DCB depends on of explained variance is far from satisfactory. Since the
the specific configuration of COO and product category, CC construct has not yielded satisfactory results, one
it is important to consider both aspects jointly. In sum, implication from this study is that further research is
there are reasons to believe that the effect of consumer needed to understand more clearly how preference judg-
ethnocentrism on domestic country bias is likely to differ ments for domestic versus foreign products are formed.
both across countries and product categories, but prior
research has not been designed to test this assumption. The effect of CE on domestic product preference is a
consistent finding in both Balabanis and Diamantopoulos
Based on those weaknesses, Balabanis and (2004) and the present study. Domestic firms in Germany
Diamantopoulos (2004) investigate the effect of CE on can well rely on a “safeguarding” effect when marketing
DCB for one domestic (Britain) and five foreign countries their products to consumers high in CE. At the same time,
of origin in eight product categories, and find initial managers from foreign countries cannot rely on CE as a
support for the assumption that this effect varies across reliable indicator of the inclination of consumers to down-
product categories and COOs. grade their products.
Replicating and extending Balabanis and Diamanto- The findings of the study confirm that CE effects are
poulos (2004), we contribute to the literature by (1) product- and country-specific, which confirms Balabanis
conducting a similar study using a larger sample of Ger- and Diamantopoulos’ (2004) findings. However, the re-
man consumers, (2) investigating a total number of 14 sults of the study contradict Balabanis and Diamantopoulos
product categories (including the eight categories re- (2004) somewhat, in that at least economic competitive-
searched by Balabanis and Diamantopoulos 2004) and (3) ness of the country-of-origin plays a role in determining
testing whether the addition of an alternative construct, respondents’ judgments. One explanation for this finding

may be that Germans are higher in uncertainty avoidance atory insights that could help shape further research
than Britons. Therefore, Germans tend to choose products questions. First, it seems that CE affects preference rat-
that they believe to be superior rather than to “experi- ings for the home country negatively in product categories
ment” with home country products when competitively (a) that are perceived as being strong drivers of the
better offers are available. This finding is also important economy (in Germany: cars, TV sets, electronics) and/or
from a managerial perspective, as Balabanis and buying from foreign firms may endanger employment in
Diamantopoulos (2004) concluded that in Britain, man- the home economy (fashion wear, toys). Consequently,
agers from economically strong countries cannot count on CE exhibits the expected effects in product-country con-
a country-of-origin effect in their favor, due to economic figurations that are likely to be perceived as threats for the
competitiveness. For the German market, however, that home economy (e.g., French food products, Italian fash-
seems to be the case: American and British firms are not ion wear and shoes, Japanese electronic products and TV
negatively affected by CE effects, and in a few cases it sets). From that perspective, it may be concluded that
even seems that CE is positively related to preference for German ethnocentrist view British and American prod-
them. ucts as good and competitive, but not as harmful for the
German economy, and therefore do not discount them in
When further examining the structure of effects of their preference ratings. References and tables with re-
CE on preferences, we are able to generate some explor- sults available upon request.

Heiner Evanschitzky
Marketing Center Muenster
Am Stadtgraben 13–15
D-48143 Muenster
Germany
Phone: +49.251.83.22036
FAX: +49.251.83.22032
E-Mail: evanschitzky@ifhm.de

IDENTIFYING INFORMATION SEARCH PATTERNS IN A WEB-BASED
ENVIRONMENT: DEVELOPMENT OF A SEARCH PATTERN INDEX
Morris K. George, University of Connecticut, Storrs
Girish N. Punj, University of Connecticut, Storrs
ABSTRACT However, based on previous research on information

source usage in traditional settings (Bruner 1986;
We propose an index (labeled as Search Pattern Westbrook and Fornell 1979) it is possible to hypothesize
Index) that can be used to capture distinctive patterns of likely influences on search in a web environment. These
consumers in an online store setting. A sensitivity analysis may be broadly classified as Consumer Characteristics
reveals that the proposed index can capture a variety of and Product and Situational Factors and are briefly de-
search sequences, thereby enabling a better understand- scribed next.
ing of underlying online behavior.
Consumer Characteristics
INTRODUCTION
Domain expertise is defined as the ability of consum-
The internet has become a dominant source of infor- ers to navigate through a website to access relevant
mation about products and services. The information information. Consumers construct navigational maps,
search sequences used during consideration set formation consisting of places or “landmarks” (Hodkinson et al.
stage may vary across consumers depending on several 2000). Landmarks are “features of the online environment
consumer, product and situational influences. The ob- which are relatively stable and conspicuous” (Dillion
served search sequences are an external (and measurable) et al. 1993, p. 173), such as known websites. Domain
manifestation of what the consumer is perhaps thinking expertise can be expected to be an important influence on
while evaluating a particular brand for inclusion in a search patterns.
consideration set. Buried in the search sequences are
insights into the behavior of consumers in an online store System Expertise: Is the ability of an individual to
environment. use the web and it includes skills such as being able to
navigate through menus, plan and execute an online
The purpose of this study is to propose a methodol- search, and the ability to manipulate and interact with a
ogy by which distinct search sequences can be measured search engine or decision aid. System expertise can also
and classified. Specifically we seek to develop a Search be expected to be an influence on search patterns
Pattern Index (SPI) that can be used to capture the distinc- (Hodkinson et al. 2000).
tiveness of common search sequences used by consumers
in accessing web-based product information. Demographic Characteristics: Such as education,
occupation and income can also be expected to influence
Consumers are likely to use the web as an information consumer search patterns. Education is assumed to in-
source during the early phases of the consumer decision crease the buyer’s need for information relating to the
making process. Later, this information is likely to be purchase decision. Occupation and income are also im-
supplemented by (and integrated with) information gath- portant influences because of their effect on search costs
ered through a mix of traditional sources such as personal (Westbrook and Fornell 1979).
(friends, family, etc.), neutral (books, magazine articles,
etc), and marketer-dominated (advertisement, etc.) Search Attitude Towards the Web: Can be expected to
sequences in these environments, while also important, influence search patterns because of consumers varying
are much more difficult to capture. But, by then the beliefs of the credibility and reliability of information
purchase decision has already been framed by the infor- posted on the web. Consumers may be skeptical of mar-
mation initially gathered on the web. Hence, it is impor- keter-dominated information and may perform unneeded
tant to understand the search sequences used by consum- search solely to establish the credibility of such informa-
ers during this important initial phase. tion.
FACTORS INFLUENCING INFORMATION Brand Preference: Consumers with a strong brand

SEARCH PATTERNS preference normally limit their search to information
about their preferred brand. A consumer who has decided
There is relatively little research on the factors that to buy a particular brand will primarily be interested in
may determine search patterns in an online store setting. comparing prices for the selected model and choosing the

best store/outlet. Hence, search patterns are likely to focus Knowledge Building Visit: The search patterns are
few attributes. likely to be similar to those in a hedonic browsing visit but
with more structure to them. They are less likely to be
Price and Non-Sensory Attributes: The search at- affected by “flow.” Fewer websites are likely to be visited
tributes can be categorized into (1) brand name, (2) price, with more pages being accessed at a visited site.
(3) searchable sensory attributes, and (4) non-sensory
attributes (Degeratu et al. 2000). Non-sensory attributes Web-Design Features: The amount of information
form an important evaluative criterion in a web-based stored on a website and the navigational tools or decisions
store environment. The importance given to price and aids that facilitate search between and within web pages
other non-sensory attributes as an evaluative criterion by that are linked to that site have an important influence on
a consumer depends on the price sensitivity and utilitarian search patterns. These design features determine to a great
needs of the consumer. extent the total number of pages visited, the frequency
with which pages are visited more than once and the
Product and Situational Factors sequence of page visits. There are at least two different
approaches that can be used to understand how consumers
Nature of the Product: The complexity and value of navigate across the webpages linked to a website.
the product and the consumer’s prior experience are likely
to influence search patterns. For a complex, high value According to the economics of information perspec-
product, consumers are likely to access more information tive, there is a likely negative relationship between search
about product attributes and product reviews or ratings. cost and extent of search (Stigler 1961). Search cost in a
For low value products, information search is likely to be web-based environment is closely related to the time or
limited to brand and/or price comparisons. effort required to execute the search (Johnson et al. 1999).
Web-design can play an important role to reduce the time
Shopping Strategies: Are likely to be an important or effort required for search by providing readily identifi-
influence on consumer search patterns because of their able links to related attribute information on other pages.
relationship to two important goals, namely, the urgency
of the purchase and the motivation for the website visit According to the information foraging framework,
(Moe and Fader 2001). A taxonomy based on these two information resides in “patches” and consumers allocate
dimensions suggests four shopping strategies: directed time to searching between-patches versus within-patch
purchase visits, search/deliberation visits, hedonic brows- search, such that they optimize the rate of information
ing visits, and knowledge building visits. gain per unit cost (Pirolli and Card 1999). In an online
store setting, consumers will allocate time to within-page
Directed Purchase Visits: Since an immediate pur- versus between-page search in order to maximize infor-
chase is planned, consumer search will be product specific mation gain. The easier it is to access information across
and of the “drill-down” type (Lynch and Ariely 2000). In web pages, the more within-page search there will be.
other words, consumers will visit few sites, but examine
more pages at a visited website. Another phenomenon To summarize, information search patterns in a web-
consumers are likely to display is referred to as “harking based environment are likely to be influenced by the
back” (Green and Jackson 1976), which refers to fre- Consumer Characteristics and Product and Situational
quently going back to pages visited earlier. Factors as discussed above. The manner in which these
factors interact to influence search patterns is depicted in
Search/Deliberation Visits: The search patterns will Figures 1 and 2.
be similar to those in a directed purchase visit, except that
there will be less “harking back” due to the diminished Figure 1 shows the factors influencing the search
urgency for the purchase. Also, consumers are likely to activity of a consumer. Regardless of the product, con-
visit more websites, and examine fewer pages at a visited sumers have certain inherent consumer characteristics
site. such as demographic, domain and system expertise and
attitude toward web, brand preference, etc. that they bring
Hedonic Browsing Visit: The search patterns are to the search task. Once web-based search is initiated,
likely to be more random, since an immediate purchase is purchase related product and situational factors, such as
not planned and neither is there a sense of purpose. shopping strategy (e.g., directed search vs. hedonic brows-
Consequently, consumers are more likely to be suscep- ing), web-design features, especially amount of informa-
tible to the effects of “flow” (Hoffman and Novak 1996). tion and the organization of information in different
They are likely to visit many websites, but only access a webpages, will either facilitate or distract the consumers
few pages at a visited site. in their search for information. Taken together, the above-
mentioned factors determine the information search pat-
terns of consumers in an online store setting.

FIGURE 1
Factors Influencing the Information Search Pattern in a Web-Based Environment
Consumer
Characteristics
• Domain expertise
• System expertise
• Demographic
characteristics
• Attitude towards web
• Brand preference
• Price & non-sensory
attributes
Direct
Purchase
visit
Consumer Web Search Purchase

Planned
Search/
Deliberation
visit
Product &
Situational Factors
• Nature of the product Web-design features
• Shopping strategy • Amount of information Hedonic- Knowledge-
Browsing visit Building visit
• Navigational tools’
decision aids
Figure 2 shows the different paths consumers may DEVELOPMENT OF A SEARCH

take in a web-based information search depending on the PATTERN INDEX
influence of the above factors at different stages of the
search process. Each of these paths reflects different Search patterns are important for designing websites
patterns of information search. For example, a consumer for different kinds of products. A manager is interested in
in a directed search visit (immediate planned purchase) understanding specific search patterns consumers display
who has decided the brand to buy will directly access any for their product. The formation of a consideration set is
known website to search for evaluative information. On often an outcome of a web-based search. It is therefore
the other hand, a consumer without a pre-search decision important for the manager to understand the search pat-
about brand and with less product knowledge will first try terns used in the formation of a consideration set. Many
to access product information through a search engine or websites of electronic stores or retail chains are designed
decision aid (depending upon domain and system exper- in such a way that a consumer can search for alternatives
tise). At this stage too, the amount of information accessed in a product category using certain attributes or attribute
varies from consumer to consumer depending on the values as search criteria. Depending upon customers’
motivation to acquire information. In both the above preferences, these alternatives can also be sorted based on
cases, the number of sites visited (inter-site search) and price, consumer ratings, or best-selling brands.
number of pages visited per site (intra-site search), links
taken from a site and the type of sites visited depend to a In developing a Search Pattern Index, it is important
great extent on the design of the web-sites. to understand the underlying search behavior of consum-

FIGURE 2
Flow Chart of Web-Based Information Search for a Directed Purchase Visit
Brand loyalty Domain and

Prior search system expertise
Brand site
Yes Yes
Directed Brand price Specific site
Purchase visit decision address known?
No No Shop bots
Search engine- Search engine-
keyword search keyword search
Experience/
Prior Search Shop bots
Sufficient No Low
knowledge about Involvement of
product category? Consumer
Visit e-
Yes stores
High
Product &
Company/ Reviews/ Product Siturational Factors
Shop bots Consumer reports
Brand site sites
ers. Consumers start their search with the “ultimate, all Search Index = Alternative transitions – Attribute transitions
singing, all dancing version of the product rather than with Alternative transitions + Attribute transitions
a straightforward replacement” (Smith 2000). In other
words consumers search for a product which scores high where alternative (attribute) transitions represent the num-
on all their important evaluative criteria (attributes). This ber of instances in which the ith +1 piece of information
gives them an opportunity to know what is available in the accessed was of the same alternative (attribute) as the ith.
market and at what price. But very often this ideal product This will give an indication whether it is an alternative-
will be a high priced premium product and not what the based search or attribute-based search depending on
consumer wants. Consumers then “trade down” (Smith whether search index is positive or negative. This index
2000) to evaluate more realistic brands for the purpose of captures the consumers’ search pattern in cases where the
forming a consideration set. Since information search is information accessed can be segregated as attribute based
associated with search cost in terms of time (and cognitive or alternative based.
effort), customers attempt to arrive at the consideration set
with minimum number of searches. They select criteria However, in most searches, a combination of at-
which, according to them, will reduce the number of tributes and attribute values are used as search criteria and
searches required for the formation of consideration set. this makes it difficult to use this index. The general search
In order to achieve this goal, consumers may engage in pattern indicates that instead of accessing information
attribute (alternative) based search. Some of the earlier about an attribute, consumers use attribute-values as search
studies have focused on developing a search index, which criteria to arrive at a set of alternatives. This brings up the
reflects this kind of search pattern in a web based decision- need for a new search index, which captures the search
making environment. Payne (1976) proposed Search In- pattern described above.
dex as:

It is evident from the general pattern of search that (1) attributes selected remains the same, (b) the attribute
consumers use attribute-values as search criteria (2) the value remains the same, but the number of attributes
number of attributes used in a search and attribute-values change, and (c) both the attribute value and the number of
differ across searches (3) consumers try to minimize the attributes change.
number of searches in arriving at a desired set of alterna-
tives. They use attributes and attribute-values in such a Here, we are ignoring the scenario where both at-
way that it helps to reduce the number of searches. An tribute values and numbers of attributes remain unchanged
ideal situation would be one in which the consideration set because, in such cases, the alternatives selected will be the
is arrived at in one search using the correct number of same and it would not be an effective search. In most
attributes and the right attribute-values. But in reality, cases, the attribute value, which has the maximum impact
consumers tend to over calibrate or under calibrate, or, in on selection, is price range. Hence, we are considering
other words, they start with more stringent or less strin- only price range as the attribute value.
gent criteria for selection. In subsequent searches, this
over/under calibration is corrected by relaxing/tightening In the three cases mentioned earlier, identifying tran-
the criteria. For an index to reflect the search pattern, it sition is easy in the first two cases. In the first case, when
should capture this over/under calibration and transitions the number of attributes remains unchanged, narrower/
in consecutive searches, i.e., from under calibration to wider price range will indicate more/less stringent criteria
over calibration or vice versa. for selection, and hence becomes a case of transition to
over/under calibration. In the second case, when attribute
Measuring Over/Under Calibration values remain unchanged, more/less number of attributes
in the second search indicates more/less stringent criteria
As discussed above, attributes and attribute-values and thus is transition to over/under calibration. However,
are used as criteria in searches. Therefore, the number of in the third case, it is very difficult to identify transitions.
attributes and attribute-values used in a search gives us an To identify transition, one can follow a general rule of
indication of how stringent the criterion is. Since this thumb like the one mentioned below:
varies with consumers and searches, the index of search
pattern should include the following variables to repre- 1. When the number of attributes is less and the price
sent over/under calibration: range is more, in the second search, it is a case of
under calibration.
1. Number of attributes used in the first search, A1
2. When the number of attributes is less and the price
2. Number of attributes used in the last search, AL range smaller, it is over calibration.
3. Total number of attributes, NA 3. When the price range is larger and the number of
attributes more, it is under calibration.
4. The attribute value: in most cases, the price range, in
the first search, R1 4. When the price range is smaller and the number of
attributes less, it is considered over calibration.
5. Price range in the last search, RL
In (3) and (4) it is assumed that when the price range
6. Maximum price range, RMAX. is smaller, the selection criteria becomes stricter even
though the number of attributes is less and is treated as
In addition to the above variables, we need to mea- transition to over calibration. Similarly, larger price range
sure the transitions- under calibration transitions, NU, and indicates less strict selection criteria and the transition is
over calibration transitions, NO. In two consecutive treated as under calibration transition.
searches, if the search criteria are more stringent in the
second search than in the first, it is a case of transition to Using the above parameters, an index of search
over calibration and NO is taken as 1. Similarly in the pattern can be expressed as:
second search, if the search criteria are less stringent
compared to the first, it is considered as transition to under SPI = 100 {(A1 – AL)/2NA + (RL – R1)/2RMAX} NU/NT
calibration and NU is taken as 1.
where A1 = number of attributes in the first search
Identifying Over/Under Calibration Transitions AL = number of attributes in the last search
NA = total number of attributes in the search
In consecutive searches, the following scenarios are R1 = price range in the first search
probable: (a) the attribute value changes, while number of RL = price range in the last search

RMAX = maximum price range The term, NU/NT is a measure of consumers’ willing-
NU = number of transitions to under calibra- ness to correct his/her initial over calibration in the subse-
tion quent searches. Together with the first term, this indicates
NT = total number of transitions the magnitude of over/under calibration.
The term, {(A1 – AL)/2NA + (RL – R1) / 2RMAX} is a The index, we proposed allows for range specifica-
measure of over all under/over calibration. Theoretically, tion only in price. However, in reality, the websites allow
the most extreme values for this term are -1 and +1. Here, for range specification in more number of attributes. In
we are interested in the sign of the expression. The order to capture the search patterns in such cases, one can
following scenarios are possible with respect to the sign of modify SPI by adding terms, similar to (A1 – AL)/2NA,
SPI. representing the over/under calibration with respect to the
range specification of different attributes. However, by
Positive Sign: This can happen in the following doing so the interpretation of SPI becomes more complex.
conditions. Since we thought price is a surrogate measure in many
cases for product features and performance, only price
• Customer specifies relatively stringent criteria on the range is included in our search pattern index.
attributes as well as the price range (i.e both (A1– AL)
and (RL – R1) are positive) in the initial search. This is AN ILLUSTRATION OF THE INDEX
a simple case of over calibration. The customer
searches for the ideal or dream product in the first The Search Pattern Index is tested using data from an
search itself. experiment involving apartment search in a web-based
environment. A total of 19 attributes (NA = 19) of the
• Other conditions, in which SPI is positive, are not so apartment like location of the apartment, availability of
straightforward. Customer may over calibrate in one heat/hot water, facilities such as swimming pool, fitness
criterion but under calibrate in the other. In this case, room in the complex are used as search criteria. Along
SPI will be positive only when the under calibration with these attributes, participants specified maximum and
in attributes/price range is canceled out by the over minimum monthly rent for the apartment as other search
calibration in price range/attributes. For example, if criteria. The maximum monthly rent that can be entered
a customer specified 4 attributes initially and 8 at- for search was $1300 and minimum was $400. Hence the
tributes in the final search, and if NA=10, then under maximum price range, Rmax is 900 (i.e., 1300 – 400).
calibration by the customer in attributes is -0.2 (i.e., Participants used a combination of these search criteria in
(A1 – AL)/2NA). However, if the same customer had a number of searches to arrive at a list of apartments,
specified a price range of 100 initially, 700 in the last which they consider for renting. The participants’ choice
search, and if RMAX is 1000, then the over calibration of attributes and the price range specified in each of the
in price range, (RL – R1) / 2RMAX) is 0.3. Hence, the searches was captured in log files. The experiment was
overall over calibration in this case is 0.1. conducted in severe to slight pressure and few alternatives
vs. many alternatives (2 x 2 experiments).
SPI will be negative in the following conditions:
The search histories of eight participants who display
• Customer specifies relatively less stringent criteria different search patterns are studied. The search param-
on the attributes as well as the price range (i.e both eters A1, AL, R1, RL, NO, and NU, are calculated from the log
(A1 – AL) and (RL – R1) are negative) in the initial files and used to arrive at the Search Pattern Index. The
search. This is a simple case of under calibration. The SPI, calculated for different combination of search pa-
customer searches for what is available in the market rameters are given in Table 1.
and then goes toward the product which best suits his/
her needs. Another explanation may be that a cus- The search histories selected show great amount of
tomer who under calibrates may not have sufficient variability. The number of attributes used in the first
knowledge about the product. search varies from 13 to 4 and that in the last search ranges
from 1 to 8. Similarly the price range selected varies from
• Other necessary condition for a negative sign for SPI 200 to 900. In all these cases, the Search Pattern Index
is the overall under calibration by the customer, (SPI) captures the underlying aspect of search pattern.
similar to the second condition explained in the case The sign of SPI indicates whether it is overall over/under
of positive sign. Here, for SPI to be negative, the over calibration and magnitude represents the extent of over/
calibration in attributes/price range has to be can- under calibration. The index is positive with higher mag-
celled out by under calibration in price range/at- nitudes when stricter criteria (over calibration) are used in
tributes.

TABLE 1
Search Pattern Index in a Web-Based Information Search
Maximum price range, RMAX = 900

Total No. of Attributes, NA = 19
Search
Pattern
Sl. No Search Parameters Index
A1 AL R1 RL NU NT
1 12 5 300 700 14 18 31.61

2 7 4 250 900 5 8 27.50
3 7 8 900 600 3 8 -7.24
4 11 4 600 200 10 12 -3.17
5 6 1 700 200 1 2 -7.31
6 8 4 900 450 4 7 -8.27
7 13 3 600 300 3 6 4.82
8 4 1 400 900 4 4 35.67
the first search compared to the last search, in terms of more efficiently. One way of customization, which many
both the number of attributes and price range (cases 1, 2, companies do, is to make the search more interactive and
and 8 in Table 1). When less stringent criteria are used in prompt customers to visit more appropriate sites. This is
the first search, in terms of both the number of attributes done by tracking the sites and pages visited by the con-
and price range, the index is always negative (case 3). sumers in their search. Identification of search pattern and
However in cases where participants used more attributes more importantly making inferences about the consumers
but wider price range, the sign and magnitude of the index from the search pattern will help to customize the search
depends on the extent of overall over/under calibration for individual consumer. In this paper we identified cer-
specified by both price range and number of attributes tain factors that influence search patterns. These factors
used in the search. In cases 4, 5, and 6, the under calibra- have varying influence on consumers’ search activity at
tion in terms of price range was large enough in order to different stages of their search. Also, in order to reflect the
nullify the over calibration specified in terms of A1 and AL individual search pattern, especially in the search for
and therefore the overall index is negative. But in case 7, formation of consideration set, a Search Pattern Index is
the over calibration specified in terms of A1 and AL is developed. This can be used as a dependent measure in
sufficient to nullify the under calibration expressed in future researches to understand and predict consumers’
terms of price range and hence the search index is positive. search pattern in a web-based environment. The index can
This illustrates the SPI’s ability to capture the search be easily modified to suit websites of different e-stores
pattern in terms of overall over/under calibration in differ- because the basic nature of search remains the same- i.e.,
ent searches. searching for alternatives using attributes and attribute
values as criteria.
CONCLUSIONS AND FUTURE RESEARCH
We developed a uni-dimensional index to represent
Emergence of web as an important source of informa- search pattern, which is a multi-dimensional activity.
tion has necessitated the identification of search patterns. Future research can focus on developing a multi-dimen-
More and more e-businesses are trying to customize their sional index without sacrificing the ease of interpretabil-
web sites to help the consumers search for information ity.

REFERENCES Web,” International Journal of Human-Computer
Studies, 52, 805–30.
Bruner, C. Gordon (1986), “Problem Recognition Styles Johnson, J. Eric, Gerald L. Lohse, and Naomi Mandel
and Search Patterns: An Empirical Investigation,” (1999), “Designing Marketplaces of the Artificial:
Journal of Retailing, 62 (3), (Fall) 281–97. Four Approaches to Understanding Consumer Be-
Degeratu, Alexandru M., Arvind Rangaswamy, and Jianan havior in Electronic Environments,” INFORMS Con-
Wu (2000), “Consumer Choice Behavior In Online ference “Marketing and the Internet.”
and Traditional Supermarkets: The Effects of Brand Moe, W. Wendy and Peter S. Fader (2001), “Uncovering
Name, Price, and Other Search Attributes,” Interna- Patterns in Cybershopping,” California Management
tional Journal of Research in Marketing, 17, 55–78. Review, 43 (4), (Summer), 106–17.
Smith, Gerald E. (2000), “Search at Different Price Lev- Payne, J. Stephen, Andrew Howes, and William R. Read-
els: The Impact of Knowledge and Search Cost,” The er (2001), “Adaptively Distributing Cognition: A
Journal of Product and Brand Management, 9 (3), Decision-Making Perspective on Human – Comput-
164–78. er Interaction,” Behaviour & Information Technolo-
Green, T.R. and P.R. Jackson (1976), “Hark-Back: A gy, 20 (5), 339–46.
Simple Measure of Search Patterns,” British Journal Pirolli, Peter and Stuart Card (1999), “Information Forag-
of Mathematical and Statistical Psychology, 29 (1), ing,” Psychological Review, 106 (4), 643–75.
(May), 103–13. Westbrook, Robert A. and Claes Fornell (1979), “Patterns
Hodkinson, Chris, Geoffrey Kiel, and Janet R. McColl- of Information Source Usage among Durable Goods
Kennedy (2000), “Consumer Web Search Behav- Buyers,” Journal of Marketing Research, 16 (Au-
iour: Diagrammatic Illustration of Wayfinding on the gust), 303–12.

Morris K. George
University of Connecticut
2100 Hillside Road
Storrs, CT 06269
Phone: 860.486.1102
FAX: 860.486.5246
E-Mail: mgeorge@business.uconn.edu

PERCEIVED RISK AND CONSUMER INNOVATIVENESS HIERARCHY:
AN EMPIRICAL STUDY OF RESISTANCE TO HIGH TECHNOLOGY
PRODUCT ADOPTION
Tanawat Hirunyawipada, University of North Texas, Denton
Mohammadali Zolfagharian, University of North Texas, Denton
SUMMARY construct containing common dimensions across product

categories as well as category-specific dimensions. Since
Consumer innovativeness is a central element in the this study focuses on high-tech products, network exter-
studies of diffusion of innovation (Midgley and Dowling nalities is introduced as a category-specific risk while
1978), and it identifies early adopters from general con- financial, performance, physical, time, social, and psy-
sumers (Roehrich 2002). Finding early adopters can ac- chological risks are included as perceived risk dimensions
celerate the diffusion of innovation and minimize the generally found across categories. The seven dimensions
chances of new product failure (Robertson 1971, pp. 112– of perceived risk are collectively tested to explore the
113). Consumer innovativeness studies generally seem to effect of innovation-resistance on the consumer innova-
focus on arousal and novelty seeking as the underlying tiveness hierarchy.
reasons for consumers to seek novel products (Hirschman
1980). However, new products also inherently encom- Data collected from 746 graduate and undergraduate
pass an element of risks associated with resistance to students at a major public university in Southwestern U.S.
adoption (Ram and Sheth 1989). This study examines: (1) show strong supports for all 12 hypotheses. The positive
consumer innovativeness is the trait that engenders the relation between CGI and ADI suggests that CGI is a
adoption of innovation; (2) new products encompass propensity to enjoy cognitive process that needs practical
resistance to adoption. We hypothesize that perceived and pragmatic activities and that can be obtained through
risk, which is a highly salient, influential attribute of using products. This trait can drive consumers toward the
innovation resistance (Bettman 1973; Dholakia 2000), actual innovation adoption. The positive relation between
moderates the relationship between consumer innova- SNI and VCI shows that SNI trait relates to consumers’
tiveness and new product adoption. tendency to seek fantasy and arousal for optimizing their
level of stimulation. This arousal is achieved by the mere
Reviewing consumer innovativeness literature (e.g., adoption of stimulus information. SNI is therefore a trait
Midgley and Dowling 1978, 1993; Hirschman 1980; in consumers with positive propensity toward VCI. The
Roger 1983; Venkatraman and Price 1990), we delineate positive relation between DSI and AI (both ADI and VCI
the consumer innovativeness hierarchy, which includes dimensions) may suggest that consumer innovativeness
the three different aspects: global innovativeness (GI), becomes stronger within specific product categories. DSI
domain-specific innovativeness (DSI), and actualized encompasses consumers’ interests in specific categories
innovativeness (AI). GI is defined as the degree to which and constitutes higher propensity to adopt innovations.
an individual is receptive to new ideas and independently
adopts the ideas through various forms of new products. The results confirm the moderating effect of per-
GI is also disaggregated into cognitive innovativeness ceived risk, that is, consumers with high overall perceived
(CGI) and sensory innovativeness (SNI). DSI is consid- risk typically score high on VCI (the higher perceived
ered an individual’s innovativeness predisposition to- risk, the higher VCI). This is a key answer to the ongoing
ward generic product class. AI is defined as the degree to dispute concerning the proneness of globally innovative
which consumers are relatively earlier in adopting new consumers (GI) toward new product adoption. The resis-
products than other members of their societies. We sug- tance to innovation adoption, represented by perceived
gest that AI be disintegrated into the adoption of actual risk associated with new products, clearly counteracts the
products (Adoptive innovativeness, ADI) and the acqui- influence of the CGI trait. It is only the SNI trait that
sition of novel information about products (Vicarious significantly influences the adoption behavior of innova-
innovativeness, VCI). tive consumers with high concern about risk. This effect
causes innovative consumers to reject new products but
This study also proposes that perceived risk, i.e., still acquire and enjoy products’ novel information. Per-
instigating resistance to the adoption of innovation, mod- ceived risk changes the innovators from “doers” to “dream-
erates the relation between consumer innovativeness and ers.” However, perceived risk has insignificant impact on
innovation adoption. Perceived risk is a multidimensional the relation between DSI and AI (both ADI and VCI). DSI

consumers’ interests and knowledge in certain product warranty programs are among considerable activities to
categories may mitigate perceived risk from his/her adop- mitigate consumers’ perceived risk.
tion.
Future study should consider more diversified prod-
The result implies that marketing managers should uct categories. By doing so, the other dimensions of
first target DSI consumers as early adopters for high-tech perceived risk might be identified, and the external valid-
products. GI consumers make up the second best potential ity of the model might be increased. Additionally, the
adopters. To achieve this objective, managers should hierarchy of consumer innovativeness needs further theo-
manipulate the market mix so that both groups of consum- rization, especially on other moderating variables that can
ers are targeted. Promotional messages, while stressing significantly influence AI. The dimensions of perceived
CGI and SNI, should ensure consumers that they can risk, which may be salient in different aspects of consumer
overcome the risk associated with innovation adoption. innovativeness, are another interesting avenue for future
Product demonstration, product trials, and appropriate research.

Tanawat Hirunyawipada
University of North Texas
P.O. Box 311396
Denton, TX 76203
Phone: 940.565.3120
FAX: 940.565.3837
E-Mail: Hirunyat@unt.edu

PERCEIVED ENTITATIVITY AS A MODERATOR OF FAMILY
BRAND EVALUATIONS
Joseph W. Chang, University of Regina, Regina
Yung-Chien Lou, National Cheng-Chi University, Taiwan
SUMMARY Jews, women, may be perceived as low or moderate

entitative groups with insignificant coherent interactions.
Research in family brand evaluations uncovers that Social groups with coherent interaction among group
the reciprocal effects of brand extensions on family brands members, such as task or intimacy groups, are normally
are moderated by the categorical similarity of brand perceived as high entitative and strongly valued (Lickel
extensions (e.g., Chang 2001; Loken and John 1993; et al. 2000). Social groups with different perceived
John, Loken, and Joiner 1998) and the accessibility and entitativity activate different psychological processes and
diagnosticity of brand extension information (e.g., lead to different subsequent impression formation (or
Ahluwalia and Gurhan-Canli 2000; Chang 2002). When judgments) of groups. Moreover, Crawford et al. (2002)
the accessibility of extension information is high, con- find that the abstracted traits of the individual behaviours
sumers engage in piecemeal or systematic processing on of high entitative groups are more likely to be applied to
brand evaluations, where detailed brand extension infor- revise group stereotypes than those of low entitative
mation is processed for the impression formation of fam- groups, which suggests that new individual members of
ily brands. Under the circumstances, the valence of highly high entitative groups are more influential than those of
accessible extension information is more diagnostic than low entitative groups on group impression formation. The
the contextual factor of categorical similarity on family findings suggest that, except for similarity and informa-
brand evaluations. The images of family brands are weak- tion accessibility-diagnosticity, group impression forma-
ened by negative extension information and are enhanced tion is moderated by the perceived entitativity of groups.
by positive extension information, regardless of the cat- For brand evaluation research, the results may also imply
egorical similarity of brand extensions. that the new extension information of high entitative
family brands is more influential than that of low entitative
Most recently, a considerable amount of attentions in family brands on family brand evaluations.
the research of social cognition has been given to the
influences of new group members on the impression Based on group impression formation and accessibil-
formation of social groups with high and low perceived ity-diagnosticity theories, experimental hypotheses are
entitativity (e.g., Crawford, Sherman, and Hamilton 2002; developed to examine the intermediating roles of per-
Lickel, Hamilton, Wieczorkowska, Lewis, Sherman, and ceived entitativity, information valence, and categorical
Uhles 2000; McConnell, Sherman, and Hamilton 1997). similarity on family brand evaluations with laboratory
A social aggregate is perceived as having “the nature of an experiments under high accessibility situations. Both high
entity, of having real existence” (Campbell 1958, p. 17; and low entitative family brands are enhanced and diluted
McConnell et al. 1997, p. 750), whereas perceived by positive and negative extension information respec-
entitativity is defined as the degree to which a collection tively, regardless of the categorical similarity of brand
of persons is perceived as being bonded together in a extensions and the perceived entitativity of family brands.
coherent unit (Campbell 1958; Lickel et al. 2000, p. 224). The polarization effect of perceived entitativity on the
The concept of perceived entitativity emphasizes the quality judgements of family brands is also observed.
interactive coherence among group members (Gaertner High entitative family brands are more favorably evalu-
and Schopler 1998). Expected variability discusses the ated than low entitative family brands. The polarization
similarity among existing group members in a more static effect then serves as a mediator on reciprocal extension
fashion, such as race, gender, and quality, whereas per- effects and leads to the results that high entitative family
ceived entitativity emphasizes the relationship among brands are more significantly diluted by negative exten-
group members on the dynamic perspective of interactive sion information and low entitative family brands are
coherence (Gaertner and Schopler 1998; Hamilton, more significantly enhanced by positive extension infor-
Sherman, and Rodgers 2003; Lickel et al. 2000). Low mation. References available upon request.
variable groups with same ethnicity or gender, such as

Joseph W. Chang
University of Regina
1320C Grace Street
Regina, SK S4T 5M8
Canada
Phone: 306.949.9523
FAX: 310.356.4934
E-Mail: mcyihwc2@yahoo.com

THE PERFORMANCE IMPLICATIONS OF SYNERGISTIC
KNOWLEDGE RESOURCE EFFECTS IN DIFFERING
ENVIRONMENTAL CONDITIONS
David A. Griffith, Michigan State University, East Lansing
Stephanie M. Noble, The University of Mississippi, University
Qimei Chen, University of Hawaii, Honolulu
SUMMARY all standardized factor loadings were well above .40, a

minimum threshold of acceptability (Hair et al. 1998).
Introduction The measurement model produced a chi-square value of
399 with 216 degrees of freedom, a chi-squared/df ratio of
Despite the growing interest among scholars and 1.85, GFI of .90, IFI of .943, and CFI of .942, indicating
practitioners, there is a lack of research on and a need to that the model provides a good overall fit.
improve our understanding of the process of implement-
ing a knowledge management strategy. This article ad- Results and Discussion
dresses this area and contributes to our understanding by
providing insights into both synergistic knowledge re- This study makes several important contributions to
source effects and the environmental influences on knowl- the literature. First, researchers have yet to fully explore
edge management strategy implementation effectiveness. the synergistic effects of knowledge resources in their
Specifically, we investigate the synergistic effect of com- conversion into capabilities. Day (1994) identifies capa-
bining knowledge of customers, industry, and firm prac- bilities and skills related to “market-sensing” and “cus-
tices to develop the capability of ability to meet customer tomer-linking,” as critical to a firm’s competitive advan-
needs and the influence of the capability of ability to meet tage. Our results build on this and indicate that when the
customer needs on performance. Further, as our focus is firm’s knowledge resources (i.e., knowledge of a firm’s
resource conversion into the capability of ability to meet customers, its industry and the firm’s practice) are devel-
customer needs, we examine the moderation effect of two oped and implemented in conjunction with one another
environmental conditions i.e., competitive intensity and the firm is able to establish more well developed capabili-
market dynamism, on the conversion effectiveness. ties in relation to meeting customer needs.
Method and Analysis The findings of this study also demonstrate the influ-
ence of environmental conditions on resource conversion.
We tested the hypotheses using data collected in the The results indicate that competitive intensity does not
retail industry. Trained market researchers were then sent diminish the conversion of knowledge resources into the
into the field to conduct 320 pre-arranged in-office inter- capability of ability to meet customer needs. In fact, the
views with retailers systematically selected from a direc- conversion of knowledge resources had a positive influ-
tory. A total of 293 usable responses were received, ence on the ability to meet customer needs even in highly
yielding a 60 percent response rate (when taking the competitive markets. These finding are encouraging as
replacement sample into account). Retail sectors in the they indicate that even in markets stronger in competitive
sample consisted of airlines, major household appliance, intensity, synergistically combining knowledge resources
automobile dealers, banks, beauty salons and services, can result in the development of the capability of ability to
clothing retail, computer dealers, convenience stores, meet customer needs.
copy and duplication service, cosmetics retail, depart-
ment stores, electronic appliances, florists retail, furniture Further, the results pertaining to the influence of
retail, gift shops, golf, grocer, hotels, insurance, jewelers, market dynamism on resource conversion indicated that
pet shops, pharmacies, photo finishing, pizza, real estate market dynamism did in fact influence a firm’s knowl-
agent, restaurants, sporting goods, toys, and travel agen- edge resource conversion. Specifically, the results indi-
cies. cated that firms who were able to synergistically combine
knowledge resources operating in highly dynamic mar-
The fit of the measurement model was assessed by kets were less effective at developing the capability of
examining factor loadings from latent variables to indica- ability to meet customer needs when compared to firms
tor variables, the chi-squared test, and fit indices. All operating in less dynamic markets. As theorized, dynamic
factor loadings were significant at the p < .001 level and markets are characterized by ever changing consumer

demands and as such even firms who are able to synergis- knowledge resources to a firm is often viewed in terms of
tically combine resources are faced with unique market its performance outcomes, without careful consideration
challenges. This is not to indicate that firms that were able of its intervening firm capabilities, such as the ability to
to synergistically combine resources were not able to meet customer needs. As such, the findings of this study
develop the capability of ability to meet customer needs, underscores, and provides empirical support for, the theo-
for the results clearly indicate that there was a strong retical precepts put forth in the capabilities literature, i.e.,
relationship between these two elements in markets char- that resources are necessary yet insufficient for the estab-
acterized by high and low market dynamism, but rather lishment of enhanced performance. Here, the findings
that the market environment itself hampered the firm from demonstrate that it is through the conversion of knowl-
fully capitalizing on its ability to synergistically combine edge resources into capabilities that a firm is able to
knowledge resources. develop a unique market position allowing for enhanced
performance. As such, this study extends the literature on
In addition, most studies of knowledge resources knowledge resources and the RBV of the firm by explor-
have focused almost exclusively on direct performance ing the employment of knowledge resources under a
implications, thus neglecting the importance of resource capabilities perspective. References available upon re-
conversion into viable capabilities. The importance of quest

David A. Griffith
East Lansing, MI 48824–1122
Phone: 517.432.5535, Ext. 260
FAX: 517.432.1112
E-Mail: griffith@bus.msu.edu

THE IMPACT OF MARKET CHARACTERISTICS ON ORDER-OF-
BRAND ENTRY STRATEGY: AN EMPIRICAL STUDY
Danielle A. Chmielewski, The University of Melbourne, Australia
Bryan A. Lukas, The University of Melbourne, Australia
Robert E. Widing II, The University of Melbourne, Australia
ABSTRACT highly competitive nature of the consumer goods indus-

tries, it is important for firms to better understand the
This study empirically investigates the impact of impact of market factors on its order-of-entry strategy.
competitive intensity, market turbulence and market This will help ensure that a firm’s order-of-entry decision
growth potential on a firm’s order-of-brand entry deci- for a brand reflects the conditions it faces in the market it
sion. The results indicate that the greater the competitive is choosing to enter.
intensity and market growth potential, the earlier the
brand entry. Directions for future research are discussed. Against this background, the purpose of this study is
to empirically examine the impact of key characteristics
INTRODUCTION of the market, namely competitive intensity, market tur-
bulence and market growth potential, on a firm’s order-of-
Order-of-brand entry refers to the sequence of entry brand entry strategy in a range of consumer goods indus-
of brands into a market (Schoenecker and Cooper 1998). tries. These three characteristics affect the attractiveness
The issue of when to introduce a new brand into a market of the market into which a firm wishes to enter. This is a
is a complex one, because as Schnaars (1986) asserts, very important and managerially relevant issue, because
different entry strategies are optimal for different firms as it suggests that different market characteristics require
well as for different market conditions. Whilst earlier different order of entry decisions. It has been argued that
entry into a market can lead to the development of a first- the order-of-entry of a product into a market is a key
mover advantage (see Lambkin 1988; Robinson 1988; determinant of the success or failure of the product
Robinson and Fornell 1985), there are also risks and costs (Fuentelsaz, Gomez, and Polo 2002). Understanding the
associated with introducing a brand earlier into the mar- interplay between the market and order of entry strategy
ket. Development and particularly promotion costs of the will allow firms to make a more informed decision about
brand are high, and the risk of failure is considerable, as the most appropriate entry strategy to employ, given the
demand uncertainty exists (Lambkin 1988; Urban et al. market conditions.
1986). In a new market, it can also be difficult to accu-
rately forecast the size of the market and the optimal The paper is structured as follows. First, we present
positioning of the new brand (Sullivan 1991). Entering a the conceptual model and put forth our hypotheses by
brand later into the market can allow a firm to benefit from providing a brief review of the relevant literature. Then,
a late-mover advantage (see Sullivan 1991; Shankar, we identify the methodology employed in this empirical
Carpenter, and Krishnamurthi 1998), although later entry study, before presenting our results. In the final section,
also has its risks, because the brand is entering a more we provide a discussion of the results and some future
competitive marketplace. Therefore, by understanding research avenues.
the factors affecting the order-of-entry decision for a
brand, a firm is better placed to enter a brand into a market CONCEPTUAL BACKGROUND AND
in such a way as to balance the risks of entering too early HYPOTHESES
with the problems of missing opportunities by entering
too late (Lilien and Yoon 1990). Figure 1 is a conceptual framework for the discussion
that follows. We now discuss each of the proposed rela-
The strategic marketing and management literature tionships and develop hypotheses based on existing litera-
has long argued that a firm’s environment affects a firm’s ture.
strategy (e.g., Miller 1987; Tan and Litschert 1994). A
firm’s external environment can be a source of informa- An attractive market is defined as one “where the
tion uncertainty (Rajagopalan and Spreitzer 1996). Con- average competitor consistently earns a return above its
ducting an analysis of the competitive environment en- cost of capital, i.e., it is creating value for shareholders”
ables a firm to better understand and measure “the attrac- (Doyle 2000, p. 157). Market attractiveness is a function
tiveness of industries for long-term profitability and the of the structure of the market in which a firm competes,
factors that determine it” (Porter 1985, p. 1). Given the and the extent to which the market is turbulent influences

FIGURE 1
Order-of-Brand Entry Strategy: A Conceptual Model
Market Attractiveness
(a) Competitive Intensity Order-of-Brand Entry
(b) Market Turbulence
(c) Market Growth Potential
the attractiveness of that market (Glazer 1991). A turbu- entrants with a “head start,” because they increase the
lent, dynamic market requires a firm to be responsive and lead-time between early-movers and late-movers.
adapt to changes in its environment (Rajagopalan and
Spreitzer 1996). A number of factors influence the attrac- Barney (1991) argues that early entrants can develop
tiveness of the market, including customer turbulence, an advantage over late entrants, because they have the
competitive intensity, technological intensity, buyer power, opportunity to develop solid relationships with suppliers,
and supplier power (Doyle 2000; Jaworski and Kohli establish strong brand awareness, and obtain a large share
1993; Miller 1987; Porter 1980). This study will focus on of the market. Consequently, they are able to lock in
three key factors; competitive intensity, market turbu- suppliers and customers and create barriers to entry, thus
lence, and market growth potential (Doyle 2001). We now making entry more difficult for subsequent entrants. If
turn to a brief discussion of each of these factors. firms know that their industry will be characterised by
intense competition, they have an incentive to make a pre-
Competitive Intensity emptive move by entering earlier and erecting barriers to
entry, rather than entering later and facing barriers to
Competitive intensity refers to the extent to which the entry. On this basis, we put forth the following hypothesis:
composition of the market and competitive actions change
over time (Gatignon and Xuereb 1997; Kohli and Jaworski Hypothesis 1: The greater the competitive intensity in
1990; Slater and Narver 1994). According to Fuentelsaz, a market, the earlier the brand entry.
Gomez, and Polo (2002, p. 250), “competition stems from
firms interacting and striving in the same environments Market Turbulence
and for the same resources.” Competitive intensity ad-
dresses the breadth and aggressiveness of a competitor’s Market turbulence is defined as the extent to which
actions (Slater and Narver 1994). Day (1994) asserts that the composition of customers and their preferences change
the intensity of competition within a market determines over time (Han, Kim, and Srivastava 1998; Kohli and
the profit potential of that market. Jaworski 1990). Market turbulence is similar to “environ-
mental heterogeneity,” which refers to “the change in
According to Porter (1980, p. 17), competitive inten- diversity of production methods and marketing tactics
sity in a market occurs “because one or more competitors required to cater to customers’ needs” (Miller 1987,
either feels the pressure or sees the opportunity to improve p. 62).
position.” Rivalry can take the form of price competition,
aggressive advertising and marketing tactics, product Customers are an important component of an indus-
introductions, and increased levels of customer service or try. They can influence the nature of competition within
warranties. an industry (and hence industry profitability) by forcing
prices down, demanding higher quality or more services,
A firm achieves superior performance by choosing to and playing competitors against each other (Porter 1980).
enter an attractive market, where it can defend its position
against competitors. An increased level of competitive A market that is characterised by high levels of
intensity decreases the attractiveness of the market, in that turbulence implies that a firm must constantly modify its
it can result in the erection of barriers to entry such as portfolio of brands or develop new brands in order to
economies of scale and strong brand name recognition continuously meet customers’ changing preferences or
(Urban and Hauser 1993). Entry barriers provide early latent needs (Jaworski and Kohli 1993). This makes a

market quite unattractive, because brand preferences are wear, food, household products, motor vehicles, personal
frequently changing. In addition, in turbulent markets, care products, and computers. Respondents were asked to
increased customer pressure puts profit margins under complete the survey with respect to a new brand that their
threat (Doyle 2000). Thus market turbulence makes it SBU had introduced into the market at least one year ago.
difficult for firms to accurately forecast the demand po-
tential for their brand, as well as to develop an appropriate Data Collection
positioning strategy for their brand (Sullivan 1991). Where
market turbulence is present, later brand entry may be the The sampling frame for this research comprised 2894
best option for a firm. The reason for this is as follows. consumer goods manufacturers in Australia and New
Entering earlier into a market requires high set up costs, Zealand. The mailing list was obtained from a public
such as R&D, advertising, and market development costs. mailing list company. T-tests found that there were no
A market that is turbulent is not particularly attractive, significant differences between the Australian and New
therefore in order to maximise its return, a firm should Zealand responses. The study employed direct mail data
engage in a strategy that is cost-effective and less risky. collection in the form of an on-line survey. An on-line
Later entry allows a firm to free-ride on the investments survey was chosen because past research has found that an
made by earlier entrants, hence reducing the costs associ- on-line survey generally “produces an acceptable re-
ated with entry and increasing the profitability of the sponse rate at a lower cost per returned questionnaire than
brand entry (Sullivan 1991). Lowest-cost entrants are mail” (Tse 1998, p. 353).
most likely to generate profitability for their new brand
(Doyle 2000). In light of this, we hypothesize the follow- E-mail invitations containing a hyperlink leading to
ing: a web-site with the on-line survey were sent to marketing
executives, brand managers, or general managers of stra-
Hypothesis 2: The greater the market turbulence, the tegic business units in our sample firms. The initial e-mail
later the brand entry. was followed by two subsequent follow-up e-mails. The
response rate, after taking into account a number of
Market Growth Potential ineligible and/or unreachable respondents, was six per-
cent, with 149 useable surveys completed. Whilst the
Market growth potential refers to the level of poten- response rate is low, it is not unusual. Alreck and Settle
tial and demand growth of the market at the time of entry (1995) note that it is not uncommon for direct mail data
of the brand strategy (Fuentelsaz, Gomez, and Polo 2002). collection response rates to fall within the range of five to
According to Lilien and Yoon (1990), a firm’s order of ten percent. Indeed, our response rate (6%) was consistent
entry decision depends on the market potential and de- with the response rate obtained by Tse (1998) (7%) in his
mand growth of a market. A higher growth and demand study comparing response rates when using e-mail versus
potential increase the attractiveness of the market, be- mail data collection methods. Low e-mail response rates
cause it is easier for a brand to be successful if the market may be attributed to the sharp increase in junk e-mails in
into which a firm is entering is growing (Doyle 2000). recent years, and increasingly effective anti-spam soft-
According to Doyle (2000), “growth markets are non- ware blocking unsolicited e-mails (Tse 1998).
zero-sum: all the competitors can grow, which acts to
reduce destructive price competition and margin erosion” For the purposes of this study, a high internal validity
(p. 158). However, in order to capitalise on the growth was more important than external validity, which is con-
potential of the market and capture early-mover advan- sistent with the view put forth by Wittink (2004) in his
tages, high market growth potential is most likely to lead recent Editorial Statement of the Journal of Marketing
to early brand entry. Therefore: Research. We nonetheless controlled for a possible non-
response bias in three ways. First, in the spirit of Armstrong
Hypothesis 3: The greater the market growth poten- and Overton (1977), we tested for differences between
tial, the earlier the brand entry. early and late respondents by dividing the data into thirds
using the three response waves as the grouping variable.
METHODOLOGY The t-tests between mean responses of first-wave and
second-wave, first-wave and third-wave, and second-
Industry Selection wave and third-wave responses indicated no statistically
significant differences (p < .05) across market turbulence,
To be consistent with previous research (e.g., Golder competitive intensity, market growth potential, order of
and Tellis 1993; Kalyanaram and Urban 1992; Schoenecker entry, and business unit size. Second, we conducted a one-
and Cooper 1998; Sullivan 1991; Urban et al. 1986), our way between-groups analysis of variance to explore the
unit of analysis is the strategic business unit (SBU) of impact of response wave on market turbulence, competi-
firms operating in a range of consumers goods industries. tive intensity, market growth potential, order-of-brand
These industries include beverages, clothing and foot- entry, and size of business unit. Respondents were divided

into three groups according to the response wave. There using exploratory factor analysis (EFA) run with SPSS
were no statistically significant differences (p < .05) in the 11.5 and confirmatory factor analysis (CFA) run with
mean scores of first-wave, second-wave, and third-wave LISREL 8.5. EFA identifies the structure of the factors to
responses. Third, we used the Mann-Whitney U Test to be tested (Gerbing and Anderson 1988). Market turbu-
compare the median responses between first-wave and lence, competitive intensity and market growth potential
second-wave, first-wave and third-wave, and second- were subjected to EFA using principal components ex-
wave and third-wave responses across market turbulence, traction method with oblique rotation in order to allow the
competitive intensity, market growth potential, order-of- factors to correlate with each other (Tabachnick and
brand entry, and size of business unit. No statistically Fidell 2001). Factors with eigenvalues greater than 1.0
significant differences (p < .05) were found. These find- were retained, and items with low loadings (less than
ings indicate that non-response bias was not a problem in 0.40) were deleted (Hair, Anderson, Tatham, and Black
this study. 1998). The results supported both discriminant and con-
vergent validity for all three constructs.
The Measures
After conducting EFA, we subjected market turbu-
This study used a combination of existing scales and lence, competitive intensity and market growth potential
new scales. to CFA to test the underlying structure identified in the
EFA (Gerbing and Anderson 1988). The chi-square of the
Order of Brand Entry: The scale was adopted from measurement model was statistically significant (χ2 =
Green, Barclay, and Ryans (1995) and Schoenecker and 96.84, p = 0.04, df = 74). The goodness-of-fit index (GFI),
Cooper (1998), and comprised a single item, “how many adjusted goodness-of-fit index (AGFI), root mean square
brands were in the market prior to entry of your brand?” error of approximation (RMSEA), parsimonious normed
The item was scored using a 7-point Likert-type scale, fit index (PNFI), comparative fit index (CFI) and normed
were 1 = 7, 2 = 6, 3 = 5, 4 = 4, 5 = 3, 6 = 2, and 7 = 1 brands. incremental fit index (NFI) indicate an acceptable fit with
the hypothesised measurement model (GFI = 0.86, AGFI =
Competitive Intensity: We adopted the 7-point scale 0.81, RMSEA = 0.06, PNFI = 0.59, CFI = 0.88, and NFI =
used by Jaworski and Kohli (1993) to assess competitive 0.72), and meet the benchmarks suggested by the litera-
intensity. The scale consisted of six items, where 1 = ture (Baumgartner and Homburg 1996).
strongly disagree, and 7 = strongly agree.
Reliability of the three constructs was measured by
Market Turbulence. We adopted the 7-point scale looking at the composite reliability and cronbach alpha.
used by Jaworski and Kohli (1993) measure market turbu- Convergent validity was assessed by examining the pa-
lence. The scale consisted of six items, where 1 = strongly rameter estimates, t-values and average variance extracted
disagree, and 7 = strongly agree. (AVE). Discriminant validity was assessed by comparing
the squared correlations for all pairs of constructs in the
Market Growth Potential: A new scale was devel- measurement model with the AVE for each construct
oped for this construct in this study. The scale for market (Fornell and Larcker 1981).
growth potential comprised 4 items: (i) the market for our
brand did not grow as fast as expected, (ii) the untapped The three constructs exhibited acceptable reliability
dollar value of our market was not as large as expected, levels, with the cronbach alphas ranging from 0.70 to 0.82
(iii) our market did not provide the expected sales poten- (exceeding the recommended level of 0.70) (Churchill
tial for the brand, and (iv) our market was not as profitable 1979), and the composite reliability ranging from 0.69 to
as expected. These were assessed using a 7-point scale, 0.83 (exceeding the recommended threshold of 0.60)
where 1 = strongly agree, and 7 = strongly disagree. (Churchill 1979; Fornell and Larcker 1981). All of the
parameter estimates were significant, with the t-values
Control Variables: Consistent with prior research ranging from 3.70 to 7.79. All the t-values were signifi-
studies on new brands and order of entry (see Fuentelsaz, cant (p < 0.01) (Hair et al. 1998). Moreover, nearly all
Gomez, and Polo 2002; Lilien and Yoon 1990; Robinson parameter estimates met or exceeded the threshold of 0.60
and Fornell 1985; Smith and Park 1992), this study recommended by Bagozzi (1981), providing evidence of
employed brand development time, quality of the brand, convergence validity. The AVE ranged from 0.33 (for
degree of similarity of new brand with existing brands, customer turbulence), 0.50 (for competitor intensity), and
and business unit size as control variables. 0.40 (for market growth potential). Whilst two of the
constructs had AVE values that did not meet the recom-
Assessment of Measures mended threshold of 0.50 (Fornell and Larcker 1981),
they exhibited acceptable levels of reliability and face
A pretest was first conducted to determine face valid- validity. There was also evidence of discriminant validity
ity. We then assessed measure reliability and validity between the constructs. Looking at the correlations in

Table 1, it can be seen that the squared correlation be- Hypothesis 2 predicted that the greater the market
tween any pair of the three constructs does not exceed turbulence, the later the brand entry. Order of brand entry
either of the construct’s AVE. is negatively and marginally significantly related to mar-
ket turbulence (ß = -0.206, p < 0.10). While Hypothesis
RESULTS 2 could not be supported, it was marginally significant and
thus warrants future research. This will be discussed later.
Regression analysis was conducted to test the hy-
pothesized relationships. Table 2 provides a summary of Hypothesis 3 predicted that the greater the market
the results. growth potential, the earlier the brand entry. Order of
brand entry is significantly and positively related to mar-
Hypothesis 1 predicted that the greater the competi- ket growth potential (ß = 0.691, p < 0.01). Hypothesis 3
tive intensity, the earlier the brand entry. Order of brand was supported.
entry is significantly and positively related to competitive
intensity (ß = 0.715, p < 0.01). Hypothesis 1 was sup- Two of the four control variables were significant.
ported. The results show a significant and negative relationship
TABLE 1
Pearson Correlations and Descriptive Statistics
Mean S.D 1 2 3 4
1. Order of Brand Entry 3.57 2.26 1.0

2. Competitive Intensity 4.83 1.25 .01 1.0
3. Market Turbulence 4.54 .93 -.06 .24* 1.0
4. Market Growth Potential 4.52 1.07 .30 -.02 -.02 1.0
* p ≤ .01 (two-tailed test)
TABLE 2
Regression Analysis: Standardized Regression Coefficients
Dependent Variable
Independent Variables Order of Brand Entry
ß t
Control Variables
Brand development time -.421 -2.338**
Quality of brand -.435 -2.711**
Degree of similarity of brand .196 1.290
Business unit size -.170 -.845
Direct Effects
Competitive Intensity .715 4.195*
Market Turbulence -.206 -1.570***
Market Growth Potential .691 3.950*
R2 .922
Adjusted R2 .832
F 10.188
* p ≤ .01; ** p ≤ .05; *** p ≤ .10 (one-tailed test)

between brand development time and order of brand entry decision. A strong market growth potential denotes a
(ß = -0.421, p < 0.05), and a significant and negative more favourable, attractive market, because it reduces the
relationship between quality of the brand and order of risks of entry and increases the likelihood of a firm
brand entry (ß = -0.435, p < 0.05). These two significant capturing an early-mover advantage (Doyle 2000).
results will be discussed briefly later.
In sum, our results, taken together, suggest that mar-
DISCUSSION AND FUTURE RESEARCH ket characteristics are indeed important drivers of order-
of-brand entry. Therefore, a thorough knowledge of the
The purpose of the study was to empirically test market is necessary to make an informed entry decision.
several hypotheses grounded in the literature regarding Importantly, these market characteristics will enable deci-
some possible antecedents of order-of-brand entry strat- sion-makers to have a more complete understanding of
egy. The findings support our hypothesis that the attrac- competitor reaction by enhancing the likelihood of cor-
tiveness of the market influences a firm’s order-of-brand rectly predicting their competitors’ entry decisions (Por-
entry strategy. ter 1980). This knowledge may subsequently influence a
firm’s decision to enter earlier or later.
The study suggests several factors as important deter-
minants of a firm’s order-of-brand entry strategy. First, Two control variables, brand development time and
competitive intensity in a market appears to be an impor- brand quality, were significantly and negatively related to
tant determinant of early brand entry. In competitively order-of-brand entry, thus indicating that both these vari-
intense markets, a firm must be strategically aware that ables are also important determinants of later brand entry.
their competitors are likely to enter earlier rather than Future research needs to be conducted in order to further
later, so this should be factored into a firm’s entry deci- our understanding of the impact of these two variables on
sion. A firm should recognize that entering such a market order-of-brand entry.
earlier helps a firm generate switching costs and establish
barriers to entry through the development of strong brand There are some other possible avenues for future
awareness and strong relationships with buyers and sup- research that would help to develop a more comprehen-
pliers (Barney 1991). In addition, as Mueller (1997) sive understanding of the market attractiveness – order-
states, early-movers face lower costs than late entrants, of-brand entry relationship. Market turbulence was nega-
because they can disregard these sunk costs when making tively and marginally significantly related to order-of-
strategic decisions, whereas later entrants must incur both brand entry. Further investigation of this relationship is
the sunk costs and any other costs in order to compete on necessary to more fully understand the impact (if at all)
the same level as the early-movers. that market turbulence has on a firm’s order-of-entry
decision for a brand.
Second, as market turbulence had only a marginally
significant impact on later brand entry, this warrants In addition, it seems desirable to include performance
further investigation. To the degree that future research indicators (such as return on investment, return on sales,
supports market turbulence as being an important deter- and market growth) in the model in order to empirically
minant of order-of-brand entry, this suggests that a firm determine the performance of the order-of-brand entry
should factor market turbulence into its entry decision. strategies given a market’s competitive intensity, market
Market turbulence reduces the attractiveness of a market turbulence and market growth potential.
and increases the risks of entry for a new brand, in that it
becomes more difficult to accurately forecast the demand Also, by looking only at market attractiveness in this
potential for a brand (Sullivan 1991). Entering later in the study, this research has focused on the opportunities and
market allows a firm to free-ride on the investments made threats (i.e., external environment) that influence a firm’s
by established competitors in the market, hence reducing order-of-brand entry strategy. Including a firm’s resources
the financial risks of entry into such an unstable market and capabilities in the conceptual model would help to
(Doyle 2000). Again, in this study, the marginally signifi- empirically determine how a firm’s strengths and weak-
cant result should be interpreted with caution. nesses (i.e., its internal environment) affects the order-of-
entry decision for a brand.
Finally, the results indicate that the growth potential
of the market is also an important determinant of early Finally, it would also be interesting to replicate this
brand entry. In markets with strong growth potential, a study in the industrial goods industries to see whether the
firm must be aware of the likelihood of its competitors attractiveness of a market plays an important role in
entering earlier rather than later, thus indicating that a firm determining a firm’s order-of-entry decision for a brand in
must factor the market growth potential into its entry these industries.

REFERENCES Ryans (1995), “Entry Strategy and Long-Term Per-
formance: Conceptualization and Empirical Exami-
Alreck, Pamela L. and Robert B. Settle (1995), The Survey nation,” Journal of Marketing, 59 (October), 1–16.
Research Handbook. New York: Irwin. Hair, Jr., Joseph F., Rolph E. Anderson, Ronald L. Tatham,
Armstrong, J. Scott and Terry S. Overton (1977), “Esti- and William C. Black (1998), Multivariate Data
mating Nonresponse Bias in Mail Surveys,” Journal Analysis, 5th ed. New Jersey: Prentice-Hall Inc.
of Marketing Research, 14 (August), 396–402. Han, Jin K., Namwoon Kim, and Rajendra K. Srivastava
Bagozzi, Richard P. (1981), “Evaluating Structural Equa- (1998), “Market Orientation and Organizational Per-
tion Models with Unobservable Variables and Mea- formance: Is Innovation a Missing Link?” Journal of
surement Error: A Comment,” Journal of Marketing Marketing, 62 (October), 30–45.
Research, 18 (August), 375–81. Jaworski, Bernard J. and Ajay K. Kohli (1993), “Market
Barney, Jay (1991), “Firm Resources and Sustained Com- Orientation: Antecedents and Consequences,” Jour-
petitive Advantage,” Journal of Management, 17 nal of Marketing, 57 (July), 53–70.
(March), 99–119. Kalyanaram, Gurumurthy and Glen L. Urban (1992),
Baumgartner, Hans and Christian Homburg (1996), “Ap- “Dynamic Effects of the Order of Entry on Market
plications of Structural Equation Modelling in Mar- Share, Trial Penetration, and Repeat Purchases for
keting and Consumer Research: A Review,” Interna- Frequently Purchased Consumer Goods,” Marketing
tional Journal of Research in Marketing, 13 (April), Science, 11 (Summer), 235–50.
139–61. Kohli, Ajay K. and Bernard J. Jaworski (1990), “Market
Churchill, Jr., Gilbert A. (1979), “A Paradigm for Devel- Orientation: The Construct, Research Propositions,
oping Better Measures of Marketing Constructs,” and Managerial Implications,” Journal of Marketing,
Journal of Marketing Research, 16 (February), 64– 54 (April), 1–18.
73. Lambkin, Mary (1988), “Order of Entry and Performance
Day, George S. (1994), “The Capabilities of Market- in New Markets,” Strategic Management Journal, 9
Driven Organizations,” Journal of Marketing, 58 (Summer), 127–40.
(October), 37–52. Lilien, Gary L. and Eunsang Yoon (1990), “The Timing
Doyle, Peter (2000), Value-Based Marketing: Marketing of Competitive Market Entry: An Exploratory Study
Strategies for Corporate Growth and Shareholder of New Industrial Products,” Management Science,
Value. England: John Wiley & Sons, Ltd. 36 (May), 568–85.
____________ (2001), “Building Value-Based Branding Miller, Danny (1987), “The Structural and Environmental
Strategies,” Journal of Strategic Marketing, 9, 255– Correlates of Business Strategy,” Strategic Manage-
68. ment Journal, 8 (January/February), 55–76.
Fornell, Claes and David F. Larcker (1981), “Evaluating Mueller, Dennis C. (1997), “First-Mover Advantages and
Structural Equation Models with Unobservable Vari- Path Dependence,” International Journal of Indus-
ables and Measurement Error,” Journal of Marketing trial Organization, 15 (October), 827–50.
Research, 18 (February), 39–50. Porter, Michael E. (1980), Competitive Strategy. New
Fuentelsaz, Lucio, Jaime Gomez, and Yolanda Polo (2002), York: The Free Press.
“Followers’ Entry Timing: Evidence from the Span- ____________ (1985), Competitive Advantage. New
ish Banking Sector After Deregulation,” Strategic York: The Free Press.
Management Journal, 23 (March), 245–64. Rajagopalan, Nandini and Gretchen M. Spreitzer (1996),
Gatignon, Hubert and Jean-Marc Xuereb (1997), “Strate- “Toward a Theory of Strategic Change: A Multi-
gic Orientation of the Firm and New Product Perfor- Lens Perspective and Integrative Framework,” Acad-
mance,” Journal of Marketing Research, 34 (Febru- emy of Management Review, 22 (January), 48–79.
ary), 77–90. Robinson, William T. and Claes Fornell (1985), “Sources
Gerbing, David W. and James C. Anderson (1988), “An of Market Pioneer Advantages in Consumer Goods
Updated Paradigm for Scale Development Incorpo- Industries,” Journal of Marketing Research, 22 (Au-
rating Unidimensionality and its Assessment,” Jour- gust), 305–17.
nal of Marketing Research, 25 (May), 186–92. ____________ (1988), “Sources of Market Pioneer Ad-
Glazer, Rashi (1991), “Marketing in an Information- vantages: The Case of Industrial Goods Industries,”
Intensive Environment: Strategic Implications of Journal of Marketing Research, 25 (February), 87–
Knowledge as an Asset,” Journal of Marketing, 55 94.
(October), 1–19. Schnaars, Steven (1986), “When Entering Growth Mar-
Golder, Peter N. and Gerard J. Tellis (1993), “Pioneer kets, Are Pioneers Better than Poachers?” Business
Advantage: Marketing Logic or Marketing Legend?” Horizons, 29 (March/April), 27–36.
Journal of Marketing Research, 30 (May), 158–70. Schoenecker, Timothy S. and Arnold C. Cooper (1998),
Green, Donna H., Donald W. Barclay, and Adrian B. “The Role of Firm Resources and Organizational

Attributes in Determining Entry Timing: A Cross- Sullivan, Mary (1991), “Brand Extension and Order of
Industry Study,” Strategic Management Journal, 19 Entry,” Marketing Science Institute Working Paper
(December), 1127–43. No. 91–105, Cambridge, MA: Marketing Science
Shankar, Venkatesh, Gregory S. Carpenter, and Lakshman Institute.
Krishnamurthi (1998), “Late Mover Advantage: How Tabachnick, Barbara G. and Linda S. Fidell (2001), Using
Innovative Late Entrants Outsell Pioneers,” Journal Multivariate Statistics, 4th ed. New York: Harper &
of Marketing Research, 35 (February), 54–70. Row, Publishers, Inc.
Slater, Stanley F. and John C. Narver (1994), “Does Tan, J. Justin and John R. Hauser (1993), Design and
Competitive Environment Moderate the Market Ori- Marketing of New Products, 2d ed. New Jersey:
entation-Performance Relationship?” Journal of Prentice Hall, Inc.
Marketing, 58 (January), 46–55. ____________ and Robert J. Litschert (1994), “Environ-
Smith, Daniel C. and C. Whan Park (1992), “The Effects ment-Strategy Relationship and its Performance Im-
of Brand Extensions on Market Share and Advertis- plications: An Empirical Study of the Chinese Elec-
ing Efficiency,” Journal of Marketing Research, 29 tronics Industry,” Strategic Management Journal, 15
(August), 296–313. (January), 1–20.
Danielle A. Chmielewski
Department of Management (Marketing Program)
Alan Gilbert Building
The University of Melbourne
Parkville Vic 3010
Australia
Phone: +61.3.8344.1886
FAX: + 61.3.9348.1921
E-Mail: dchmi@unimelb.edu.au

ORGANIZATIONAL CULTURE ANTECEDENTS OF MARKET-DRIVEN
POSITIONAL ADVANTAGE AND ORGANIZATIONAL
PERFORMANCE CONSEQUENCES
Artur Baldauf, University of Bern, Switzerland
David W. Cravens, Texas Christian University, Fort Worth
Christian Bischof, University of Bern, Switzerland
SUMMARY Hypothesis 1: There is a positive relationship be-

tween (a) market focus, (b) participative decision
Research attention has been given in several studies making, (c) support and collaboration, (d) power
to organizational culture, market orientation, innovation, sharing and market intelligence.
and organizational performance. Dimensions of culture
are potential antecedents to market-driven positional ad- Hypothesis 2: There is a positive relationship be-
vantage which is expected to impact organizational per- tween (a) market focus, (b) participative decision
formance consequences. Nonetheless, only limited re- making, (c) support and collaboration, (d) power
search has considered the simultaneous antecedent and sharing and innovativeness.
consequence relationships. In the focal construct we view
positional advantage as a multi-dimensional construct Hypothesis 3: There is a positive relationship be-
consisting of market intelligence, innovativeness, and tween (a) market focus, (b) participative decision
learning and development. The proposed organizational making, (c) support and collaboration, (d) power
culture antecedent dimensions are market focus, partici- sharing and learning and development.
pative decision making, support and collaboration, and
power sharing. We also relate the focal construct to Market intelligence, innovativeness, and develop-
desired consequences and argue that a market-driven ment and learning positional advantages are expected to
positional advantage positively impacts market perfor- have a positive impact on the market and financial perfor-
mance, profitability, and sales growth. mance of the business unit. We offer the following hy-
pothesis:
The organizational culture, positional advantages,
and performance conceptualization adds to prior cultural, Hypothesis 4: There is a positive relationship be-
market orientation, and performance research in the fol- tween (a) market intelligence, (b) innovativeness,
lowing ways: (1) cultural and process perspectives of and (c) learning and development and market and
market orientation are considered in an antecedent conse- financial performance of the business unit.
quence context; (2) positional advantage which could be
termed as a receptivity to innovate is conceptualized as a The data for examining the hypotheses were col-
three dimensional construct consisting of market intelli- lected from senior managers employed by companies in a
gence, innovativeness, and learning and development; German-speaking business environment. The sampling
and (3) a multi-dimensional view of organizational per- objective was to include a wide range of larger firms in a
formance is examined in terms of culture and innovation wide range of different businesses. A judgment sampling
antecedents. procedure was applied to identify candidate companies.
We utilized a standardized questionnaire which was pre-
We view organizational culture in terms of widely tested for wording and understanding before final mail
shared and strongly held values and belief systems and distribution. After several follow-up activities we re-
consider the higher level construct as a source of competi- ceived 204 usable questionnaires were returned reflecting
tive advantage. Our market-driven positional advantage a response rate of 21 percent. Established multiple item
dimensions are proposed to capture relevant dimensions measures were used for the ten construct measures which
of the capacity to innovate, and represent a more compre- we purified applying state-of-the art methodologies. Ac-
hensive reflection of innovative behavior than is consid- ceptable reliability and validity of the scales was indi-
ered in prior research. Based on our conceptual logic we cated. Besides investigating the above stated direct rela-
specifically interested in examining the following cultural tionships we also controlled in our path models for poten-
and positional advantage hypotheses: tial effects of customer type and company size on the

organizational culture and positional advantage relation- laboration has no significant impact on the positional
ships. In addition, market turbulence, competitive inten- advantage dimension.
sity, and technological turbulence were included as mod-
erators for the positional advantage and performance Higher market intelligence activities result in higher
relationships. Regression analysis was used to test the market, profitability, and growth performance. Innova-
hypotheses. tiveness positively impacts profitability but not market
and growth performance. Learning and development pos-
The results for H1, H2, and H3 are encouraging but itively impacts all three performance consequences. Hence
mixed. Market focus and participative decision making we find partial support for H4a and H4c and mixed
are strong predictors of market intelligence, innovativeness, support for H4b. The moderating effects of the environ-
and learning and development. Power sharing is only a ment constructs are not supported.
predictor of learning and development; support and col-

Artur Baldauf
Management Department
University of Bern
Engehaldenstrasse 4, 3012 Bern
Switzerland
Phone: +41.31.631.5331
FAX: +41.31.631.5332
E-Mail: baldauf@imu.unibe.ch

SEE NO EVIL, HEAR NO EVIL, SPEAK NO EVIL: A STUDY OF
DEFENSIVE ORGANIZATIONAL BEHAVIOR TOWARDS
CUSTOMER COMPLAINTS
Christian Homburg, University of Mannheim, Germany
Andreas Fürst, University of Mannheim, Germany
SUMMARY complaint intention), representing the consequences of

DOB.
Despite substantial benefits of an effective complaint
management, there is ample evidence that many organiza- Results
tions do not handle customer complaints appropriately.
Instead, organizational members often exhibit an appar- With respect to the (direct and indirect) antecedents
ently irrational and dysfunctional defensive behavior to- of the prevalence of DOB, the results of our study support
wards complaints. This paper aims at providing a theoreti- the prediction that the prevalence of negative attitudes
cal explanation for this phenomenon. Furthermore, based towards complaints is negatively affected by customer
on a dyadic data set, it analyzes antecedents and conse- orientation of HRM and customer orientation of corporate
quences of the prevalence of defensive organizational culture, respectively (p < .01). Moreover, as suggested,
behavior towards complaints (DOB). the prevalence of negative attitudes towards complaints,
in turn, has a positive effect on the prevalence of DOB (p <
Conceptualization of DOB .01). In addition, our results provide evidence for the
prediction that the prevalence of DOB is also (directly)
Based on individual psychology and organizational negatively affected by customer orientation of HRM (p <
theory as well as in line with literature on organizational .01). However, we fail to find statistical support for our
behavior and complaint management, we argue that indi- assumption that the prevalence of DOB is too (directly)
viduals in organizations perceive customer complaints as negatively influenced by customer orientation of corpo-
a source of actual or potential threat to self-esteem, rate culture (p > .10).
reputation, autonomy, resources, or job security. Thus, in
order to protect themselves against this threat, they exhibit Furthermore, all hypotheses related to the conse-
different types of DOB. quences of the prevalence of DOB are confirmed by the
data. More specifically, the prevalence of DOB is found
Overall, we identify seven types of DOB that can be to have a negative effect on complaint satisfaction as well
assigned to one of the three following categories: com- as on perceived complaint-based improvements (p < .01).
plaint acquisition (i.e., isolation from complaints, hostile Moreover, we observe that complaint satisfaction posi-
behavior towards complainants), complaint transmission tively influences overall customer satisfaction and future
(i.e., no/biased transmission of complaints to complaint complaint intention, respectively (p < .01). In addition,
managers, no/biased transmission of complaints to senior our findings confirm a positive impact of perceived com-
managers), and complaint utilization (i.e., no/inadequate plaint-based improvements on overall customer satisfac-
handling of complaints, no/inadequate analysis of com- tion as well as on future complaint intention (p < .01).
plaints, and no/inadequate use of complaint information
in decision making). Research Issues
Methodology First, drawing upon individual psychology and orga-

nizational theory, we provide a theoretical explanation for
Our data analysis is based on 110 dyads. Each dyad the phenomenon of DOB.
consists of a managerial assessment of the antecedents
(i.e., customer orientation of human resource manage- Second, our study provides evidence for the high
ment (HRM), customer orientation of corporate culture, relevance of this phenomenon. More specifically, we
prevalence of negative attitudes towards complaints) and show that the presence of DOB has a significant negative
types of DOB in the focal company and five customer impact on customers’ complaint satisfaction and on the
assessments related to their post-complaint responses ability of the firm to learn from complaints (i.e., the
(i.e., complaint satisfaction, overall customer satisfac- implementation of complaint-based improvements).
tion, perceived complaint-based improvements, future

Third, our research provides an understanding of the First, managers can work directly on this phenom-
forces within an organization that can influence DOB. A enon. Our conceptualization of DOB (i.e., the identifica-
particularly strong influence on the prevalence of DOB tion of seven different types) provides managers with a
comes from the customer orientation of HRM. checklist type of structure. Based on this structure, they
can analyze the prevalence of DOB in their company and,
Managerial Implications in turn, initiate activities to reduce this behavior.
Our research also provides guidance for managers on Second, managers may also work on the antecedents
how to improve a firm’s complaint management. More of the prevalence of DOB. In this context, the customer
specifically, the findings of our study suggest that manag- orientation of a firm’s HRM is particularly important.
ers should strive to reduce the prevalence of DOB in their
company. This can be done in two ways:

Christian Homburg
Marketing Department I
University of Mannheim
L 5-1
68131 Mannheim
Germany
Phone: +49.621.181.1555
FAX: +49.621.181.1556
E-Mail: prof.homburg@bwl.uni-mannheim.de

THE REST OF THE ICEBERG: AN EXAMINATION OF
NONCOMPLAINING SERVICE CUSTOMERS
Clay M. Voorhees, Florida State University, Tallahassee
Michael K. Brady, Florida State University, Tallahassee
David M. Horowitz., Florida State University, Tallahassee
SUMMARY provider. Moreover, consumers that simply exit a failed

encounter may demonstrate more favorable intentions
While most service recovery strategies have focused toward the firm than those that opt to complain and
on customers that actively complain to the firm, it has been experience poor recovery efforts.
suggested that these customers represent only the “tip of
the iceberg” (Diener and Greyser 1978, p. 22). This means The analysis reveals significant differences between
that the majority customers that experience poor service the different consumer groups with respect to their repeat
simply exit the encounter and managers receive no oppor- purchase intentions and several negative outcomes vari-
tunity to recover. Davidow (2003) suggests that in order ables. Specifically, the results confirm the hypotheses and
to fully understand customer behavior in failed encoun- demonstrate that consumers that receive a satisfactory
ters a study must be forwarded that investigates the recovery effort have the most favorable intentions toward
responses of the consumers that do not register com- a firm. Following this, noncomplainers demonstrate the
plaints. next most favorable intentions. Finally, consumers that
complain but receive a dissatisfactory recovery effort are
In an effort to address this critical gap in the literature, least likely to return to do business with a service provider.
the current study conducts a comprehensive comparison
of groups of customers that opted not to complain and Ultimately, the findings offer significant implica-
those that registered a complaint to the firm. The compari- tions for both service researchers and managers. In par-
sons are made across repeat purchase intentions and a ticular, the results provide further justification for re-
number of negative outcome variables. It is hypothesized search on service recovery strategies as it appears that
that consumers that end a failed encounter with a satisfac- poor recovery efforts can be costly to a firm. References
tory recovery are most likely to return to the service are available upon request.

Clay M. Voorhees
Florida State University
Tallahassee, FL 32306–1110
Phone: 850.645.1519
FAX: 850.644.4098
E-Mail: cmv6779@cob.fsu.edu

FROM EMPATHY TO FORGIVENESS: A PROSOCIAL PERSPECTIVE IN
SERVICE FAILURE AND RECOVERY RESEARCH
Felix T. Tang, The Chinese University of Hong Kong, Hong Kong
SUMMARY reflects a concern for others and incorporates the concepts

of sympathy, compassion, and tenderness (Batson 1990).
Consider a dinning scenario in which your server has When people feel empathy towards the transgressor, they
completely forgotten your order. How do you feel? This are more likely to forgive him/her (McCullough et al.
is an outcome failure and the extant literature predicts that 1997, 1998). Such evidences have been found across
you will become dissatisfied, and you may also engage in demographics variables, across cultures, and across con-
negative word-of-mouth and exit behaviors. What if your texts (Eisenberg and Miller 1987). Thus, consumers are
server collapses into tears in front of you after finding out capable of displaying prosocial behavior, even if such
that he/she has completely forgotten your order? Would behaviors may not be directly beneficial. The concept of
you feel sympathetic towards the server despite of the empathy has also been discussed by Parasuraman et al.
service failure? Would you abandon your resentment (1988), Tax et al. (1998), but the conceptualization is
towards the service provider? In such situation, consum- quite different in this paper. While Parasuraman et al.
ers might feel empathetic and behave in a merciful way (1988) and Tax et al. (1998) saw consumers as recipients
towards the server even if there has not been a word of of empathy, this paper advocates the possibility for con-
apology or compensation. sumers to be givers of empathy.
Although this scenario may be pushing the limits, it Observational set and perceived similarity are the
exemplifies a boundary condition where the concepts of two most relevant antecedents in the service recovery
justice, equity, fairness, disconfirmation, and attribution context. Observational set is the focal of one’s attention
are not adequate to explain all consumer behaviors. The and it can be instructed or cued. When people focus on the
traditional view downplayed consumers’ ability to self- needy other’s emotional state or perspective, they are
restore satisfaction, ignored consumers’ consideration for more likely to feel empathetic towards those who are in
other’s welfare, and overlooked the effect of positive need than when people focus on the facts (Eisenberg and
emotions on post-recovery evaluation. Miller 1987). Perceived similarity concerns one’s percep-
tion of the resemblance between oneself and a comparison
Prosocial Behavior and Forgiveness object. In their meta-analysis of sixteen studies, Miller
et al. (2001) concluded that perceived similarity with
Prosocial behavior research in psychology suggests another was associated with or led to feelings of empathy
consumers may look beyond their own well being (Batson or sympathy.
et al. 1995). Prosocial behavior is broadly defined as
“social behaviors oriented to benefit another, regardless Forgiveness as Consumer Behavior
of potential outcomes for oneself” (Miller, Kozu, and
Davis 2001). Such behavior includes but not limited to What marketers are doing during service recovery
donating, forgiving, helping, sharing and volunteering. are essentially wooing for consumer forgiveness with
Philosophers, such as Hume (1949), considered prosocial compensation and apology, asking consumer to replace
behavior as moral actions of human instinct. Of the resentment (e.g., dissatisfaction) with beneficence (e.g.,
various prosocial behaviors, forgiveness is one of the satisfaction). The two consequents of forgiveness, lower
most relevant concepts in the service marketing context. avoidance and lower revenge, are also strikingly similar to
Modern philosophers, such as North (1987), agreed that the marketing consequents of successful service recovery
the central theme in forgiveness involves replacing re- (e.g., increase repatronage and lower negative word-of-
sentment with beneficence. Using McCullough, Parga- mouth activity). Thus, forgiveness is a valid, alternative
ment, and Thoresen’s (2000) definition, forgiveness is an perspective of service recovery.
intraindividual and prosocial change toward a perceived
transgressor. Hypotheses
Empathy Based on insights from prosocial behavior research,

six hypotheses are postulated and are summarized in
The most proximal determinant of forgiving of for- Figure 1. Due to space constraint, the hypotheses and their
giveness is empathy (McCullough et al. 1998). Empathy development are not discussed here. The Figure should be

FIGURE 1
An Empathy-Forgiveness Model in the Context of Service Failure
Satisfaction
Observational
Set + +
Empathy
+ Forgiveness + Repatronage
Perceived Intention
Similarity + Negative WOM
Intention
self-explanatory (the positive and negative signs indicate our ability to explaining consumer behavior; (3) postulate
the predicted directions of relationship) and the logics a new perspective by seeing service recovery efforts as
flow from the discussion above. attempts to seek consumers forgiveness.
Discussion No doubt, the concept of justice, equity, fairness,

disconfirmation, and attribution remain the cornerstones
This paper aims to (1) call for attention to the limita- of our understanding on consumer behavior in the recov-
tions of the current views and theories on service recovery process. However, let us not blindly rely only on these
ery; (2) introduce the concepts of prosocial behavior, concepts for they do not give us a complete picture of
empathy, and forgiveness into service recovery to broaden consumer behavior. References available upon request.

Felix Tang
Hong Kong
Phone: 852.9678.5958
FAX: 852.2603.5473
E-Mail: felixtang@cuhk.edu.hk

INTERNATIONAL MARKETING ALLIANCE DYNAMICS: EMPIRICAL
FINDINGS FROM THE PHARMACEUTICAL INDUSTRY
Sengun Yeniyurt, Michigan State University, East Lansing
Janell D. Townsend, Michigan State University, East Lansing
Erin Cavusgil, Michigan State University, East Lansing
SUMMARY uncovering the effects of competition for collaboration,

experiential learning and culture distance on the propen-
Collaborative ventures are an indispensable tool for sity to engage in a specific mode of entry: international
executives in the quest for achieving a sustained competi- marketing alliances.
tive advantage in the marketplace. The significance of this
instrument is evidenced by a marked increase in the use of A co-evolutionary dynamic framework is introduced
cooperative arrangements as a business form, and the as a means to understand the complex phenomena in-
variety of studies related to this phenomenon. The bound- volved in forming international marketing alliances. The
aries of what define the marketplace, however, have advantages derived from international marketing alli-
evolved in ways previously unimaginable. The world has ances are first explored based on the concepts of resource
become more integrated in terms of infrastructure and complementarities and extension. In order to reveal the
dependence; the drivers of these global trends leave few consequences of the intensity of collaboration and com-
industries untouched by the increased velocity and inten- pany experience on international collaborative venture
sity of competition for resources and customers (Wolf formation in the global marketplace, we draw on the
2000). In this domain, collaboration has become an indis- organizational ecology perspective. Further, we employ
pensable means through which firms are able to respond the concepts of organizational learning and the assump-
to environmental turbulence (Achrol 1991), effectively tions of the internationalization process to delineate the
extending the reach and resources they would not have effects of experience and culture on international collabo-
otherwise. The use of alliances in marketing and interna- rative ventures.
tional contexts is inextricably linked.
The hypotheses are tested through the analysis of the
This complex phenomenon encompasses dynamic global marketing alliance activity announced by U.S.
aspects of the competitive environment and various per- pharmaceutical firms from 1984 to 2003; this includes
spectives of firm behavior. As firms collaborate to com- 792 international marketing alliance formations engaged
pete (Ohmae 1989a), the relative population of potential by 317 firms. An event history was constructed for each
partner firms remains somewhat static over time, eventu- U.S. pharmaceutical company starting with the date of its
ally leading to competition for collaboration in an indus- first international marketing alliance. The data were ar-
try. What remains to be discerned is the nature of the ranged into yearly spells updated when an event occurred
competition for alliance partners, particularly in light of and at the end of each year to account for the changes in
the increased environmental turbulence and diversity variables. A continuous time event history analysis with
established by the globalization of industries (Achrol time varying covariates was employed in order to estimate
1991). Additionally, experiential learning is posited to be the effects of the independent variables on the probability
fundamental to a firm’s ability to accumulate knowledge of a company engaging in a new international alliance.
(Huber 1991; Sinkula 1994) and conduct international
operations (Cavusgil 1980; Johanson and Vahlne 1977); The findings of a Cox hazard rate model support the
yet, empirical findings supporting this position remains hypothesized effects of legitimation and competition,
sparse. While the literature generally proposes cultural sustaining the assumptions of organizational ecology,
distance as a significant factor in the internationalization extending the theory to the global context of inter-firm
process (Johanson and Vahlne 1977; Johanson and relationships. Particularly important to the literature is
Weidersheim-Paul 1975; Kogut and Singh 1988), recent that companies are eager to follow their competitors in the
findings do not support this proposition with respect to early stages of industry level internationalization utilizing
international market entry timing (Mitra and Golder 2002). marketing alliances as a mode of entry; yet once a critical
The question remains as to whether the diversity of mass of international marketing alliances is attained in the
cumulative culture distance experiences affects an industry, the propensity to engage in alliances declines.
organization’s inclination toward future international col- Our results provide strong support for the proposition that
laborative ventures. This study contributes to the litera- companies learn from their international alliance forma-
ture through the reconciliation of these perspectives by tions. Cultural distance is a significant factor in interna-

tional marketing alliance engagement, as the cumulative cumulative cultural distance experiences on an organiza-
cultural distance of previous alliances increases the like- tion’s inclination toward future actions. Finally, organiza-
lihood that a company will engage in a future international tional size has a positive effect on the propensity to engage
marketing alliance. Previous studies have only consid- in a new international marketing alliance. References are
ered the effect of culture distance experience in specific available upon request.
markets, as opposed to the impact of the diversity of

Sengun Yeniyurt
The Eli Broad Graduate School of Management
Phone: 517.353.6381
FAX: 517.432.4322
E-Mail: yeniyurt@msu.edu

A CROSS-NATIONAL STUDY OF CONSUMER-FIRM EXCHANGE
RELATIONSHIPS WITHIN THE CONTEXT OF MARKET MILIEUS
Patrick Lentz, University of Dortmund, Germany
Deepak Sirdeshmukh, North Carolina State University, Raleigh
Ed Nijssen, Radboud University, The Netherlands
Hartmut H. Holzmüller, University of Dortmund, Germany
Jagdip Singh, Case Western Reserve University, Cleveland
SUMMARY other control mechanisms) and relationship-based (char-

acterized by reciprocity, fairness, and personal exchange
Broadly, two fundamental mechanisms are relevant processes) market milieus. We propose hypotheses re-
for understanding the dynamics that promote or impede garding the effects of market milieus on micro relation-
consumer-firm market exchanges in a global society of ships, mainly drawing from Granovetter (1985), Kollock
networked economies. One fundamental mechanism fo- (1994), and Zucker (1986):
cuses on the ongoing exchanges between individual con-
sumers and sellers, and involves understanding the moti- H1: Consumers’ perceptions of exchange value and satis-
vations for each party to enter into, consummate and faction will be significantly higher in rule based
continue market exchanges, from here on referred to as market milieus, relative to relationship based market
micro mechanism (e.g., Nicholson, Compeau, and Sethi milieus.
2001). A second fundamental mechanism focuses on the
embeddedness of market exchanges and involves under- H2: Consumers’ judgments of trust will be significantly
standing the social and market context that provides an higher in relationship based market milieus, relative
arena for individual consumer-firm exchanges to occur, to rule based market milieus.
referred to as macro mechanism (e.g., Shapiro 1987;
Zucker 1986). Few if any studies have focused on the H3: Consumers’ loyalty intention judgments will not
macro mechanisms of individual consumer-firm ex- vary significantly for different configurations of
changes, which is particularly troubling since different market milieus.
markets around the globe are likely to vary both in terms
of their micro and macro mechanisms. H4: Relative to relationship based market milieus, the
influence of consumers’ satisfaction and value judg-
Conceptual Background and Research Hypotheses ments on loyalty will be significantly higher in rule
based market milieus.
We draw from extant literature to develop a micro
model of Satisfaction-Trustworthiness-Trust-Value-Loy- H5: Relative to rule based market milieus, the influence
alty (ST 2VL) relationships (e.g., Liu, Furrer, and of consumers’ trust judgments on loyalty will be
Sudharshan 2001; Sirdeshmukh, Singh, and Sabol 2002). significantly higher in relationship based market mi-
Regarding the larger context in which exchanges occur, lieus.
we draw from sociology (Shapiro 1987; Zucker 1986) and
economics (Hirschman 1970; Nooteboom, Berger, and Research Design
Noorderhaven 1997) to define market milieus. These
involve consumers’ and sellers’ expectations regarding Two different samples were used to test the hypoth-
the norms governing customer-firm interactions in a given eses. First, we collected a consumer sample regarding the
market. Market milieus thus refer to the dominant and individual-level dynamics between satisfaction, trustwor-
shared normative structure that forms the context for thiness, trust, value, and loyalty with their insurance
consummating exchanges between providers and their company (n = 365, 504, and 316 for the U.S., Germany,
customers. As Kramer (1999) notes, in the absence of and the Netherlands, respectively). Second, a separate
specific knowledge about each other, consumers and consumer sample was used to evaluate the macro environ-
firms can revert to formal rules at one end of a continuum ment, i.e., the market milieus at the industry level in the
to mutual relationships at the other end, to solve their same three countries (n = 18, 30, and 31 for the U.S.,
problem or uncertainty. We thus define two dominant Germany, and the Netherlands, respectively), which helped
configurations of market milieus, that is, rule-based (char- circumvent method bias. To test our hypotheses, the two
acterized by contracts, detailed legal environments, and types of data were joint in an hierarchical type of analysis

using both mean analyses and multiple-group SEM, with and partial support for H1, since only value shows higher
the three countries as separate groups. Items for analyzing levels in rule based relative to relationship based market
the micro environment have been adapted from existing milieus; however, H2 cannot be supported. Third, turning
literature, whereas items capturing the level of rules and to the moderating effects of market milieus, we find
relationships have been developed for the purpose of this partial support for H4 and H5. While both satisfaction and
study, and pretested using “think aloud” exercises with trust in frontline employees differ in their influence on
small convenience samples. The measurement model loyalty according to our hypotheses, trust in management
evidenced good levels of reliability and validity. policies and practices as well as value do not.
Results Discussion
Our analysis revealed substantial and statistically Our study results demonstrate that examining con-
significant differences across the three market milieus at sumer-firm relationships in situ is a promising approach
the macro level. We found that the U.S. insurance market for better understanding exchange relationships, particu-
milieu is more relationship based, the German insurance larly when considering markets in different countries.
market is rule based, and the milieu of the Dutch insurance This extends contemporary understanding of the exchange
market is located between the two extreme positions. dynamics and thus provides both theoretical and manage-
Although additional differences were found in the micro rial relevance. The results clearly reveal that an isolated
model relationships between countries, the more interest- investigation of micro mechanisms is inappropriate, espe-
ing result is related to the influence of market milieus. cially in cross-cultural settings. We believe that our work
First, the macro level variables add significantly to model enhances the knowledge of macro effects on micro mecha-
fit. Second, regarding direct effects of market milieus on nisms by conceptualizing and empirically demonstrating
exchange-specific constructs, we find full support for H3 the different effects. References available upon request.

Patrick Lentz
University of Dortmund
Otto-Hahn-Str. 6
D-44227 Dortmund
Germany
Phone: +49.231.755.3277
FAX: +49.231.755.3271
E-Mail: Patrick.Lentz@udo.edu

AN EXAMINATION OF IMC AT THE TACTICAL LEVEL: DIFFERENCES
ACROSS TIME AND PRODUCT TYPE
Stephen J. Grove, Clemson University, Clemson
Les Carlson, Clemson University, Clemson
Michael J. Dorsch, Clemson University, Clemson
Christopher D. Hopkins, Clemson University, Clemson
SUMMARY A research design was employed to compare services

print advertisements and goods print advertisements on
Since its inception, Integrated Marketing Communi- their degree of integration and the trend towards integra-
cation (IMC) has been viewed as a valuable concept by tion over time. First, a sample of services and goods print
marketing practitioners (McArther and Griffith 1997). advertisements was generated to obtain a wide range of
IMC has been characterized as the coordination of com- advertising messages over three different time periods.
munication tools (e.g., advertising, publicity, sales pro- Next, a data collection process was established for ex-
motion, etc.), for a brand. A review of this phenomenon in tracting data from each print advertisement. The data
practice uncovers three broad categories including IMC extraction process consisted of two steps. During the first
as “integrated communication,” IMC as “one voice” com- step, content analysis of each print advertisement was
munication and IMC as a “coordinated marketing com- conducted in order to profile it in terms of the communi-
munication campaign.” This study focuses on the inte- cation tools that it employed. During the second step, the
grated communication approach because it appears to be profiled print ads were re-examined to determine the
the one most often used in practice. Relative to services, extent to which they were integrated. This two-step pro-
it is argued that advertising in this area should be charac- cess produced a data set consisting of count data. Once the
terized by a greater presence of IMC than the advertising data set was created, the data were examined using both
of physical goods. This is largely do the fact that IMC may inferential and descriptive analyses.
provide a mechanism that addresses some of the problems
associated with marketing a product the customer has Results indicate that while IMC is a reality among
difficulty comprehending clearly and thus, may alleviated both services and physical goods advertisements, IMC
a degree of perceived risk. was found to exist in services advertising to a greater
extent relative to goods advertising. Moreover, the inci-
This study empirically investigates the utilization of dence of IMC at the advertisement level, regardless of the
IMC across product type and over time. It is expected that product type that is being advertised, has not increased
there is greater usage of IMC among sellers of services significantly over recent years. As a result, there appears
than among sellers of goods. It is also expected that the to be much room for further and more enlightened adop-
usage of IMC has grown over the years, reflecting its tion of IMC principles. References available upon re-
increased emphasis in the marketing discipline. To mea- quest.
sure the nature and incidence of IMC, the study focuses on
its manifestation at the tactical level.

Christopher D. Hopkins
College of Business and Behavioral Science
Clemson University
233 Sirrine Hall, Box 341325
Clemson SC, 29634–1325
Phone: 864.656.3952
FAX: 864.656.0138
E-Mail: chopkin@clemson.edu

ASSESSING THE EFFECTS OF IN-SCHOOL POINT OF PURCHASE AND
SAMPLING ON THE CHOICE OF A HEALTHY FOOD OPTION
Dafina Rexha, The University of Western Australia, Australia
Katherine Mizerski, Edith Cowan University, Australia
Richard Mizerski, The University of Western Australia, Australia
SUMMARY influence and socialisation can have a significant impact

on children’s first purchasing decisions.
Childhood obesity has tripled in the U.S., U.K., and
Australia in the past decade with one in four children now School canteens are one of the first places where
being classified as overweight or obese (Azhar 2004; de children are faced with the decision of what food to
Brito and Segal 2002). These children are more likely to purchase by themselves. Usually parents give their chil-
develop health problems such as diabetes, asthma, cardio- dren money to purchase from the school canteen. The
vascular disease, joint problems, back pain, and psycho- parents, however, are not present at the time the purchase
logical problems (Anthony, Patterson, and Kemp 2002; decision is made.
Cameron et al. 2003).
Many suggest that an immediate change in the school
There are a number of factors thought to contribute to food and nutrition environment needs to occur (Stitzel
the “obesity epidemic” (Time 2004) with increased con- 2003; Wallis 2004) so that schools are healthy environ-
sumption of energy-dense foods and decreased levels of ments where the food related policy of the canteen allows
physical activity, receiving most of the blame (Waters and children the opportunity to make healthy choices (Baxter
Baur 2003). Advertisers of these energy-dense foods and 1998; Conento et al. 1995). O’Dea (2003), in her research
fast foods have come under scrutiny and are accused of into children’s eating practices, identified convenience,
targeting their products directly to children. This targeting taste and social factors as barriers to healthy eating. The
has led for a call to ban food advertising during children’s children in the study noted, however, that they would eat
programmes in a number of countries (Canning 2004) and what was available and “allowed.” This would suggest
has forced manufacturers to re-evaluate their offerings. that if healthy food were available for children to purchase
and eat at school, they would do so. There is a lack of
Recently, a new culprit has been identified as a research, however, on how to create this change and how
contributor to childhood obesity. It has been suggested to encourage children to eat healthier. The current study
that schools, that routinely educate children about good attempts to fill this gap and provides the results of a
nutrition in the classroom, have failed to send the right multiple observation period experiment undertaken at a
messages from their school canteens. A study by Which? primary school canteen in Australia. The experiment was
Magazine (2003) found that school canteen offerings read designed to determine the effects of sampling, subtle
more like fast food menus and contributed less than one promotion, and availability of healthy food in children’s
portion of fruit and vegetables to a child’s daily intake. preferences and purchases of canteen food.
This has led to criticism being levelled at schools for
failing to fulfil their duty of care towards students by The sample consisted of 166 children in grades one to
providing “unhealthy” food choices (Stanton 2002; Wallis four. Each child who purchased food from the canteen
2004). during recess was observed and their food choices re-
corded. After obtaining a baseline of existing product
With one third of all children’s meals coming from sales, Smoothies, a healthy food choice consisting of
outside the home and more than half (52%) of these being strawberries, skim milk, and orange juice, were made
consumed at school, many food preferences and food available as a new option. This new option was supported
acceptance patterns are developed in schools (Birch 1999; with point of purchase promotion in the form of a large
Story 2002). Schools are also a critical component of the poster at each window. In addition, two of the classes were
child’s social environment and play a significant role in provided with samples prior to recess.
shaping children’s food preferences and eating behaviours
(Baxter 1998). Not only are children reinforced through The results support previous research into the effect
their own choices at school, they are also affected by the of advertising on children’s preferences but go beyond
choices of their peers (Kubik et al. 2003). This peer this to look at the effect of alternative promotional tools on

actual purchase. While the point-of-purchase ad did have mation in class on nutrition, these promotional tools could
the effect of increasing sales, sampling, an often used tool serve as reinforcement for the educational message.
in adult food purchases (e.g., Fazio 1986; Smith and
Swinyard 1980), was not as effective. Although more Although the findings of the current study are based
children who had sampled ultimately purchased, the dif- on children’s choice behaviour in a school canteen set-
ferences were not significant. ting, the findings may be generalizable to a broader venue.
As canteens are one of the first places where children can
Evidence from the study suggests that the availability make independent food choices, the results of this study
and promotion of a healthy food product can reduce the can be used by other outlets where children make choices
purchase and consumption of other “less healthy” op- (i.e., fast food outlets). The results can provide valuable
tions. This effect, however, was short-lived. The rebound information for managers of food outlets that target chil-
of these less healthy options in the absence of promotional dren and for producers of food targeted at children. By
tools may provide a strategy for long-term behavioural providing information on healthier options, food advertis-
change. If schools are serious about encouraging good ers may be able to circumvent restrictions on advertising
eating habits, and joining the fight to reduce childhood during children’s programs. References available upon
obesity, perhaps they need to consider the use of “promo- request.
tional” materials. Given that most schools provide infor-

Katherine Mizerski
MTL/BPM
Edith Cowan University
Pearson St.
Churchlands 6018
Western Australia
E-Mail: k.mizerski@ecu.edu.au

TESTING WHY ADULTS PURCHASE FAST FOOD CARTOON
CHARACTER TOY PREMIUMS
Claire Lambert, University of Western Australia, Australia
Richard Mizerski, University of Western Australia, Australia
SUMMARY among children. However, few of the three to seven year

old consumers that are targeted actually purchase the
The Fast Food Industry has come under extensive premiums. Because independent visits to Fast Food res-
criticism for its “junk food” content and its aggressive taurants appear rare before the age of eight, the vast
marketing to children (Sieders and Petty 2003). The use of majority of toy premiums are bought by adults. It is
extensive advertising to children and toys as premiums are assumed that adults purchase Fast Food premiums after
alleged to cause young consumers to develop poor eating being “pestered” by the child to obtain the items
habits, become overweight or obese, and labor under (McKimmie 2004).
long-term health problems due to a diet of too much Fast
Food (Dore, Harris, and Whittaker 2002; WHO 2003; The most effective time for a child to request a Fast
McKimmie 2004). Food toy premium would appear to be at the Fast Food
restaurant, where the decision is made and the retail
New restrictions, including warning labels on fast environment showcases the product. One would expect
food meal packaging, the banning of advertising to chil- that meal buyers accompanied by a child to a Fast Food
dren, and eliminating promotions that involve toys as facility, would have a greater probability of purchasing a
premiums, are being studied by some Australian States cartoon toy premium aimed at that child than a buyer
and their Federal Government in order to tackle rising without a child with them (AFA 2001). Fast Food and
obesity in young Australian children (Cumming 2002). brand past purchase frequency and perceived value of the
Similar sentiments have been expressed in the United premium were also tested for an effect in the premium’s
States and the United Kingdom, where there are initiatives purchase.
to restrict or ban fast food advertising and promotion
aimed at young consumers (Banzhaf 2003; Barboza 2003; The research was conducted using a field experiment
House of Commons 2004). There are present bans on at a major Fast Food brand’s retail facility located in the
“junk food” advertising during children’s television pro- Perth, Australia metropolitan area. The promotion was a
grams and restrictions concerning the type of premiums four-week long continuity premium promotion that of-
used in five other countries (Lambert and Mizerski 2003). fered 16 versions of a Snoopy plush toy. The campaign
included local media advertising that almost exclusively
The success of premium based promotions appears aired on children’s television, as well as point-of-pur-
rare for many product categories, with consumer partici- chase material meant to generate awareness and inten-
pation rates often reported as one percent or less of a tions to visit the fast food retail operations.
brand’s buyers (Schultz, Robinson, and Peterson 1993).
However, the experience of fast food retailers is quite The study used a cross-section sample of meal buyers
different (Taylor 2001). McDonalds, Burger King, and that were personally interviewed over three time periods
Wendy’s lead all other companies in the use of premiums to investigate the effect of the Snoopy premium on con-
in their promotions. Most of their premiums appear to be sumer decision-making and purchase in this category. A
aimed at children, with some Fast Food retailers begin- total of 50 surveys were completed per day (n = 100 per
ning their targeting to three years olds (AFA 2001). weekend) over three periods, and sampled meal buyers
that visited the Fast Food facility in-store, during the
Critics of the Fast Food Industry (Schlosser 2001; weekend lunch period.
WHO 2003; McKimmie 2004) charge that the special
allure of toy type premiums, and the sheer volume of their It was found that the accompaniment of children with
use, may train young and naïve children into repeatedly the meal buyer was not a significant effect in the purchase
requesting or purchasing Fast Food meal items. The of a Snoopy cartoon toy premium, nor for the number of
promotion of these meals frequently feature cartoon char- Snoopy premiums reported purchased in the past. Buyers
acter premiums tied into television programs or movies in of the premium provided a significantly higher mean
their advertising and sales promotion (Spethmann 2002). value of the Snoopy premium than non-buyers of the
The repeated use of heavily promoted premiums is as- premium, but these values were significantly lower if the
sumed to build Fast-Food brand and category loyalties buyer was accompanied by a child or children. However,

there were no differences in the groups’ perceived value rant should be re-evaluated as it appears they may tend to
of the Snoopy premium if pre-campaign respondents’ primarily lure adults. These results suggest that potential
mean estimation of the premium’s value was included in Public Policy interventions need to be aimed at the parents
comparisons. and caretakers that largely make the purchase decisions
for what children eat. The need for bans on the use of
The effect of previous purchase of the Snoopy pre- cartoon–based premiums targeted to the children’s mar-
mium, not purchase of the brand of Fast Food, was a ket (cf., Sieders and Petty 2003; McKimmie 2004), needs
significant effect in buying the Snoopy premium on the more evidence before causality can be assumed. Of course,
visit that was surveyed. In addition, the reported purchase it may be that simple possession of a premium is enough
of the premium reflected a distribution not significantly as reported by Pierce et al. (1998), but rejected by Lee
different from the Negative Binomial Distribution often et al. (2004), in regard to finding of an effect of Tobacco
observed with fast moving consumer package goods and premiums in smoking initiation.
gambling products (Mizerski et al. 2004). This suggests
that purchase of the continuity premium may have be- Researchers that have looked into the stochastic
come habitual. This apparently habitual component was a pattern for fast moving consumer package goods have
significant effect in repurchase of that premium, and argued (cf., Ehrenberg 1969) for there generally being
potentially more influential than the respondent’s per- little or no effect of information processing or attitudinal
ceived value of the premium. precursors in frequent choice behavior. Information-based
programs and remedies may be expected to be less effec-
The study’s results provide an insight into the effective than if the buyer of the premium was more cognitively
tiveness of children’s requests for Fast Food toy premi- involved in the decision. Clearly, more research needs to
ums targeted at them. The claims that cartoon-based be conducted to see if these early findings can be repli-
premiums necessarily lure children to a Fast Food restau- cated and explained. References available upon request.

Richard (Dick) Mizerski
School of Economics and Commerce M261
The University of Western Australia
35 Stirling Highway, WA 6009
Australia
Phone: +618.6488.7210
FAX: +619.6488.1055
E-Mail: dickm@biz.uwa.edu.au

MULTIMARKET CONTACT AND THE MODERATING ROLE OF
DOMINANT LOCAL PLAYERS: A CONCEPTUAL OVERVIEW
Sweta Chaturvedi Thota, James Madison University, Harrisonburg
SUMMARY the relationship between multimarket contact and mutual

forbearance have not received adequate attention in ex-
This paper identifies the factors that moderate the tant literature. In this paper, the factors that moderate the
relationship between multimarket contact and mutual relationship between multimarket contact and mutual
forbearance observed by firms that operate in multimarkets. forbearance are identified. The paper draws from game-
A framework is developed that identifies the market theoretic propositions and hypercompetition literatures
factors that moderate the relationship between multimarket and develops a framework that identifies the market
contact and mutual forbearance. Specifically, the paper factors that moderate the relationship between multimarket
discusses the moderating effects of factors, which arise contact and mutual forbearance. It is proposed that the
mainly due to the presence of dominant local players, on various moderating variables would reduce the level of
the levels of mutual forbearance observed by firms. It is mutual forbearance observed by firms when firms operate
proposed in this paper that the suggested moderating in multimarkets. In this paper, primary market factors that
factors would reduce the level of mutual forbearance moderate the relationship between multimarket contact
observed by firms when firms operate in multimarkets. and mutual forbearance are identified, which would help
in understanding and determining the relationship be-
Multimarket competition is defined as a situation tween multimarket contact and the mutual forbearance
where firms compete with each other simultaneously in observed by firms.
several markets (Karnani and Wernerfeldt 1985). This
situation is present in multiproduct industries or industries The paper discusses the moderating effects of factors
with different geographic markets (Fernandez and Marin that arise mainly due to the presence of single-market
1998). A better understanding of multimarket competi- firms or dominant local players on the levels of mutual
tion is of great importance to marketing strategy research forbearance observed between firms. The factors identi-
and practice because of the central nature of the issue of fied are the presence of single-market firms or dominant
interfirm rivalry to marketing strategy issues. Thus, it is of local players, the level of competition triggered by dom-
extreme importance to firms that compete in multiple inant local players, and resource similarity between the
markets to develop an understanding of the behavior of single-market firms and the firms operating in multimar-
other firms that operate in these markets. Further, in kets.
formulating a strategy for multimarket competition, it is
important to consider the impact of various factors that The discussion in this paper, that describes the pro-
affect the level of competition between firms. Despite cess through which multimarket competition influences
attempts by past research to explore issues revolving the intensity of competition and leads to reduced mutual
around multimarket competition, there is a dearth of forbearance, has been represented in a framework. The
research that explores the roles of the various moderating framework, demonstrates the role of moderating variables
variables on the relationship between firms that engage in that reduce mutual forbearance and heighten the intensity
multimarket competition. of competition when firms operate in multimarkets.
Strategy researchers and industrial economists have The framework and propositions in this paper have
examined whether firms that engage in multimarket com- several research implications for managers and research-
petition would compete aggressively against each other or ers. The impact of the several moderating variables on the
observe mutual forbearance (i.e., tacitly collude). Mutual level mutual forbearance and, consequently, on the level
forbearance is a form of tacit collusion in which firms of competition between firms that operate in multimarkets
avoid competitive attacks against the rivals they meet in has been examined in this paper. The reduced levels of
multiple markets (Jayachandran, Gimeno, and Vardarajan mutual forbearance and the increased intensity of compe-
1999). Interestingly, past research has argued both for and tition may have serious implications on the success of
against whether mutual forbearance is a deterministic multimarket firms and on the level of competition that
outcome of multimarket contact. multimarket firms may face from a single-market firm or
a dominant local player operating in that market. This
It is stressed here that although the relationship be- study will help shed light on how managers of multimarket
tween multimarket contact and mutual forbearance has firms should mentally map their market and determine
been examined in prior research, the factors that moderate their rivals.

Sweta Chaturvedi Thota
College of Business MSC 0205
James Madison University
Harrisonburg, VA 22807
Phone: 540.568.6817
FAX: 540.568.3587
E-Mail: thotasc@jmu.edu

THE NATURE OF CO-OPETITION: LITERATURE REVIEW
AND PROPOSITIONS
Pilsik Choi, University of Illinois, Champaign
SUMMARY ances mainly because it focuses on both cooperation and

competition with competitors simultaneously while the
Interfirm interactions have been extensively studied other three focus only on cooperation with suppliers,
in economics, marketing, strategic management, and or- customers, competitors, and/or noncompeting firms. Com-
ganization theory. Over the last two decades, research on petitors in this paper are defined as direct competitors who
interfirm interactions has evolved into two research are competing in the same market(s).
streams: one based on firms’ competitive behavior and the
other on firms’ cooperative behavior. Before the mid Co-opetition is categorized into channel co-opetition,
1980’s, interfirm interactions were analyzed with great marketing co-opetition, and R&D co-opetition. Channel
emphasis on firms’ competitive behavior. This stream of co-opetition denotes a situation where one competitor
research was influenced by economic theory, which gen- becomes the other competitor’s buyer or supplier or a
erally embraces intense competition. In the mid and late situation where one competitor uses the other’s distribu-
1980’s, while other researchers dwell on the tradition of tion channels or production facilities while they compete
firms’ competitive behavior, a number of researchers with each other in a final goods market. Marketing co-
turned their attention to the cooperative aspects of inter- opetition is a situation where direct competitors cooperate
firm interactions. Since then, they have identified various for joint marketing efforts (e.g., brand alliances, joint
types of interfirm cooperation, e.g., symbiotic marketing, promotion, bundling, etc.) while they compete with each
co-marketing, strategic alliances, etc. All of these con- other in the same market. R&D co-opetition occurs when
cepts emphasize cooperation between firms. direct competitors join forces in research and develop-
ment or other similar activities (e.g., joint new technology
In the early 1990’s, sharing the similar background to development, joint industry standard development, etc.)
that of the concepts emphasizing cooperation, a new while they compete in other activities.
concept, “co-opetition,” was created. Unlike the other
concepts, co-opetition (blend of cooperation and compe- Drawing upon the literature on co-opetition, strategic
tition) focuses on both cooperation and competition at the alliances, co-marketing, and symbiotic marketing, I iden-
same time. The book, Co-opetition, by Brandenburger tify complementarity, compatibility, managerial ability,
and Nalebuff (1996) ignited the popularity of the concept. and competition level as antecedents of co-opetition.
The book has generated tremendous interest among prac- Complementarity occurs when the pooled skills and re-
titioners and researchers. In light of the amount of inter- sources can create excess value relative to their value
ests in co-opetition, however, it is surprising that aca- before the pooling. Since the cooperation part of co-
demic publication on co-opetition is very limited thus far. opetition requires mutual benefits, one partner should
have certain skills or resources that the other partner needs
Addressing this issue in the literature, this paper in order for such cooperation to be formed. Thus, a high
attempts to investigate the nature of co-opetition. The level of complementarity between competitors will likely
purpose of this paper is to review the literature related to lead them to enter into co-opetition. In this process,
co-opetition and examine its antecedents and conse- technological change moderates the influence of
quences. Specifically, this study attempts to address the complementarity on co-opetition. Compatibility, which is
following questions: (1) What is co-opetition? (2) How characterized as similarities in management style and
can it be classified? (3) Which firm characteristics cause company culture, is another important factor for the
co-opetition? (antecedents), and (4) What benefits does cooperation part of co-opetition. Although competitors
co-opetition generate for firms that adopt the concept? are complementary with each other, if they are not com-
(consequences). patible in the areas where they cooperate, it is difficult for
them to cooperate with each other in those areas. Hence,
Co-opetition is defined as the situation where a group a high degree of compatibility between competitors will
of competitors cooperate in activities associated with likely lead them to engage in co-opetition. Management
creating mutual benefits while at the same time they ability is also an important factor for co-opetition forma-
compete with each other in activities associated with tion. Management that has the ability to recognize skills
dividing up the benefits. Co-opetition is different from and resources it needs and seek those skills and resources
symbiotic marketing, co-marketing, and strategic alli- from other companies will more likely identify co-opetition

opportunities with other competitors that can provide available to the firms in the co-opetitive relationship, their
such skills and resources. Therefore, competitors with products and services are not likely to match those from
such management ability will likely forge a co-opetitive the firms in the co-opetitive relationship. Thus, the firms
relationship. Finally, competition level can influence co- in the co-opetitive relationship gain a competitive edge
opetition formation as well. When two companies com- over competitors outside the relationship, which increase
pete fiercely in a market, they likely perceive each other competitiveness. In addition to increased competitive-
as an enemy to defeat and have less willingness to collabo- ness, co-opetition also provide the participating firms
rate, even if they have complementary skills and re- with efficiency because they do not need to develop skills
sources. Thus, it is proposed that a high level of competi- and resources they need internally.
tion between competitors will decrease the possibility of
formation of a co-opetitive relationship. The most important contribution of this paper is that,
to my knowledge, this study is the first study that formally
In terms of consequences, co-opetition brings better defines co-opetition and examines both its antecedents
competitiveness and efficiency to participating competi- and consequences. Given the fact that research on co-
tors. Since competitors outside the co-opetitive relation- opetition is at its early stage, this study provides ground-
ship do not possess all the skills and resources that are work on which future research on co-opetition can build.

Pilsik Choi
Department of Business Administration
339 Wohlers Hall
University of Illinois at Urbana–Champaign
1206 South Sixth Street
Champaign, IL 61820
Phone: 217.333.4240
FAX: 217.244.7969
E-Mail: pchoi@uiuc.edu

TECHNOLOGY VERSUS PEOPLE: TWO SCHOOLS OF THOUGHT ON
PRICING CAPABILITY DEVELOPMENT
Lewis K.S. Lim, Indiana University, Bloomington
Rebecca J. Slotegraaf, Indiana University, Bloomington
Rockney G. Walters, Indiana University, Bloomington
SUMMARY investment priorities for firms. Specifically, the techno-

logical school would place a greater priority on the exter-
A growing number of studies in marketing and stra- nal acquisition of material, tangible resources such as
tegic management are concerned with the notion of pric- software and consultancy services, whereas the people
ing capabilities. Within this emerging literature, two school would prescribe the internal development of ab-
schools of thought dominate the intellectual discourse. stract, intangible resources such as tacit market knowl-
The “technological” school, reflected mainly in industry- edge, and interfunctional conflict resolution mechanisms.
oriented and business practice writings, emphasizes com- Each school inherently emphasizes resource acquisition
putational precision, and objectivity as prerequisites to from a different kind of factor market and may be suited
effective pricing. Accordingly, it views pricing capabili- for adoption in a different type of organizational environ-
ties as driven by the installation of information systems ment. These issues call for linking the study of pricing
and the procurement of pricing engineering expertise. In capabilities to the broader theories of management, such
contrast, the “people” school, espoused largely by schol- as the resource-based view, transaction cost theory, and
ars subscribing to an organizational capabilities or orga- contingency theory.
nizational learning perspective, sees market sensing and
concerted organizational action as the fundamental me- In exploring these two schools of thought, this paper
diators of effective pricing. Focusing on such higher- contributes to marketing knowledge in a number of ways.
order people-dependent processes, this school advocates First, the current state of understanding concerning pric-
the development of coordination mechanisms, tacit knowl- ing capabilities is put into proper perspective, with the
edge, and routines over time to improve pricing effective- delineation of two possible alternative positions scholars,
ness. consultants, and practitioners alike might hold toward the
topic. Certain implicit theories or otherwise unstated
Noting the divergent viewpoints, we construct two premises on either side are thereby illuminated. Second,
“implied” models of pricing capability development in within each school of thought, the notion of a pricing
order to bring to light the mediating processes and driving capability itself is further clarified with the identification
mechanisms assumed by proponents of the respective of the key enabling skills and resources. Third, the alter-
schools of thought. In the technological model of pricing native processes of improving pricing effectiveness are
capability development, we depict the availability of explicated, thus allowing us to understand how different
accurate and timely information on cost and revenue compositions of resources influence the effectiveness of
flows, the acquisition of pricing engineering knowledge, price decision making.
and the customization of data display in information
systems as antecedents, and pricing precision and objec- As an agenda for future research, we outline two
tivity as mediators, of pricing astuteness. In the people primary directions in which the framework proposed in
model on the other hand, we suggested collective task this paper could be extended. For a start, investigations
experience in developing the pricing process, a shared/ could be carried out into the managerial-perceptual and
dominant logic about the marketplace, and interfunctional organization-specific factors that lead a firm to adopt
goal alignment within the organization as antecedents, either the technological or the people model of pricing
and market sensing and concerted organizational action capability development. Thereafter, research could focus
as mediators, of pricing astuteness. on determining the contingent deployment of different
pricing capabilities to suit the organizational decision
By juxtaposing the two alternative models, we are environment. More definite conclusions could then be
able to draw differential implications of each school. drawn about the role of pricing capabilities in influencing
Evidently, each school suggests a different set of resource firm performance.

Lewis K.S. Lim
Kelley School of Business
Indiana University
1309 East Tenth Street
Bloomington, IN 47405
Phone: 812.855.1116
FAX: 812.855.6440
E-Mail: lewislim@indiana.edu

DIMENSIONS AND OUTCOMES OF RELATIONAL EXCHANGE IN A
BUSINESS-TO-BUSINESS CONTEXT: A META-ANALYSIS
Mohammadali Zolfagharian, University of North Texas, Denton
Rajasree K. Rajamma, University of North Texas, Denton
SUMMARY methodological moderators below were identified and

used in our analysis.
Research in the last two decades identifies relational
exchange as the key underlying factor driving business 1. Contextual Moderators:
performance. However, the theoretical foundation of re- 1.1 The country from which data was collected (U.S.
lational exchange is far from clear. Relational exchange vs. other countries).
has been operationalized using over 37 different dimen- 1.2 Whether the data was collected from a single
sions, many of which overlapping and interrelated industry; and two.
(Gundlach et al. 1995). This might be due to the contextual
and methodological differences of the studies. The contri- 2. Methodological moderators:
bution of our meta-analysis is twofold: (1) to establish 2.1 Whether upstream or downstream firm in the
whether an overall positive association exists between relationship responded to the survey.
relational exchange and its purported outcome variables 2.2. Respondent’s status in the firm (whether the
and (2) to identify the contextual, methodological and respondent was a top executive or not).
measurement peculiarities that instigate variation in find-
ings reported by different researchers. Hence, our first We hypothesized that relational exchange dimen-
proposition is that there is an overall positive relationship sions have a stronger impact on its outcome variables (a)
between relational exchange and its outcomes. when the sample is drawn from within the U.S. (b) when
it is drawn from a single industry (c) when the upstream
The database for meta-analysis was constituted of 34 partner is the respondent in the study and (d) when the
studies, altogether contributing 37 dimensions of rela- respondent has a high rank in the firm.
tional exchange and 13 outcome variables. After collaps-
ing dimensions with similar definitions and eliminating Analysis was done using the multiple regression
those that have less than 10 correlations, our database technique with dummy coded variables (0/1). The z-
consisted of 19 studies (55%) providing 89 usable corre- transformed correlations formed the criterion variable
lations. These correlations were associated with five rela- and the five dimensions of relational exchange, the three
tional exchange dimensions: solidarity, durability, flex- outcome variables and the four moderator variables formed
ibility, information exchange and mutuality, and three the predictor variables in the regression equation. All the
outcome variables: satisfaction, performance, and com- data, except the effect size, were dummy-coded.
mitment. These dimensions and outcomes were catego-
rized into (1) instrumental/concrete including flexibility, The findings offer insights to researchers as well as
information exchange (two dimensions) and performance practitioners. We could not find any significant overall
(one outcome) and (2) abstract including solidarity, dura- relationship between relational exchange and its outcome
bility, mutuality (three dimensions), satisfaction and com- variables. The results also partially support our argument
mitment (two outcomes). We hypothesized that the instru- that instrumental dimensions (flexibility) as well as in-
mental dimensions and outcomes of relational exchange strumental outcomes (performance) of relational exchange
would have significant, positive relationship with the are more salient than abstract dimensions. Further, sig-
effect size. nificant moderators of the relationship were found to be
sampled industry (studies that collected data from a single
Differences across correlations in meta-analyses are industry had stronger positive effect sizes) and the
often attributed to four broad categories of characteristics: respondent’s rank in the firm. Responding firm’s position
measurement method, research context, estimation proce- in the relationship was found to marginally moderate the
dure, and model estimation (Sultan et al. 1990). As our relationship. By far, it was seen that, sampled industry is
analysis focused on correlations, model estimation proce- the strongest moderating factor of the relationship.
dure and model specification (i.e., omitted variable bias)
were not considerable issues for this study (Henard and In sum, the findings indicate that the association
Szymanski 2001). Therefore, two contextual and two between relational exchange and outcome variables is

highly dependent on three factors: (1) how relational and its outcomes. This calls for more studies in the area.
exchange is operationally defined, (2) who evaluates the Moreover, we found that other meaningful dimensions of
relationship (upstream firm versus downstream firm and relational exchange such as trust, harmonization of con-
high-status employee versus low-status employee within flict have received barely any attention from the research-
the organizational hierarchy), and (3) the data source ers. Hence, this is an avenue for further research on the
(single- vs. multiple-industry). subject. Future research should also include the views of
all the parties involved in the relationship in order to get
One major limitation of our findings stems from the a more comprehensive view of what is happening within
small number of empirical studies that are available and a given relationship.
that explore the relationship between relational exchange
For further information contact

Rajasree K. Rajamma
College of Business
P.O. Box 311396
Denton, TX 76203–7231
Phone: 940.565.4787
FAX: 940.381.2374
E-Mail: RajammaR@unt.edu

ENTRY MODE AND LEVEL OF EQUITY: A SIMULTANEOUS
EXAMINATION OF FOREIGN DIRECT INVESTMENT
GOVERNANCE
Sudha Mani, University of Western Ontario, London
Kersi D. Antia, University of Western Ontario, London
Aric Rindfleisch, University of Wisconsin – Madison
SUMMARY inclusion of these dual viewpoints allows us to assess their

relative theoretical merit upon both aspects of FDI deci-
Entering a foreign market is an expensive, risky, and sions.
daunting task comprised of multiple decisions. Among
the most important of these decisions is the governance Using a rich dataset of 4,459 subsidiaries of 858
arrangement of a new subsidiary. Firms entering foreign Japanese firms across 38 countries over a nine-year pe-
markets must first select a mode of entry (e.g., a wholly- riod, we specify a cross-classified, multilevel, bivariate
owned subsidiary versus a joint venture with a local model of FDI governance. This model enables simulta-
partner). For joint ventures, the entering firm must also neous estimation of entry mode and level of equity deci-
decide upon its level of equity investment. Although these sions and integrates theoretical viewpoints from both the
two governance decisions are strategically inter-related, TCA and experience perspective.
the extant FDI literature has examined them in isolation.
The bivariate model in this paper addresses the problem The results from the bivariate multilevel analysis
associated with the correlated nature of these two deci- suggest a need to integrate the TCA and experience
sions. perspective. Our findings reveal that TCA factors have a
greater significant impact on choice of entry mode than it
The key theoretical perspectives explaining entry does on the level of equity participation. In contrast,
mode and level equity have been the transactional cost and experience considerations significantly determine both
experience perspective. Prior studies have typically as- aspects of governance. The much larger effect sizes ob-
sumed common antecedents of transaction cost analysis tained for entry mode and host country experience, rela-
(TCA) and experience for both aspects of governance, tive to that of advertising intensity lend support to the
and little is known about their specific effects on each assertion that models of FDI activity that focus solely on
aspect of governance. TCA emphasizes capability pro- TCA-related concerns are underspecified (Blodgett 1991;
tection, while the experience perspective considers gov- Brouthers 2002). Further, our analysis reveals that al-
ernance with a view to capability enhancement. Despite though experience exerts a powerful influence on FDI
recent research examining these dual objectives (Delios governance, the direction of its effects is dependent upon
and Henisz 2000; Lu 2002; Madhok 1997), the relative the type of experience in question.
predictive ability of these two theoretical perspectives
across both aspects of FDI governance is unknown, as This research was supported by a grant from the
there has been no simultaneous estimation of their influ- Social Sciences and Humanities Research Council of
ence. Our paper addresses this lacuna by simultaneously Canada (#411-98-0393) at the University of Western
estimating both mode and level of FDI governance via the Ontario. References available upon request.
lens of both the TCA and experience perspectives. Our

Kersi D. Antia
Richard Ivey School of Business
University of Western Ontario
1151 N. Richmond Street
London, Ontario N6A 3K7
Canada
Phone: 519.661.4231
FAX: 519.661.3959
E-Mail: kantia@ivey.uwo.ca

TOWARDS AN UNDERSTANDING OF THE GOVERNANCE OF
COMPLEX NETWORKS OF RELATIONSHIPS: UNCERTAINTY,
GOVERNANCE SYSTEM CHOICES, AND
PERFORMANCE OUTCOMES
Andrew T. Stephen, University of Queensland, Australia
SUMMARY works (cf., Bradach and Eccles 1989; Heide 1994, 2003).
Plural governance is intended to afford firms in networks
The governance of interfirm relationships in market- with greater flexibility to adapt to changing conditions in
ing channels is complex. Much extant literature has con- their environments or channels. This notion is applied to
sidered relationship governance from the perspective of more general considerations of how uncertainty can be
discrete, isolated dyadic relationships (e.g., a single buyer dealt with in relationship networks through governance
and a single supplier). In reality, relationships in market- mechanisms. The discussion then turns to introducing the
ing channels, supply chains, and distribution networks notion of governance system compatibility; that is, opti-
combine to form relationship networks that are inherently mal configurations or combinations of governance modes
complex and can have high levels of uncertainty. The under plural governance must be compatible across levels
governance of relationships in these networks is likely to or branches of relationship networks. The most efficient
vary significantly from the governance of the single governance systems or plural forms of governance are
dyadic relationships that have been the focus of many those that provide sufficient flexibility and greater cer-
previous studies. An emerging stream of literature consid- tainty, and that are compatible across levels. The concep-
ers the governance of multiple, linked relationships in tual framework builds on these foundations and advances
network-type structures (e.g., Heide 2003; Mishra, Heide, propositions that predict (1) how downstream channel
and Cort 1998; Wathne and Heide 2004). While it is conditions can influence upstream channel governance
understood that complex systems of interfirm relation- choices, and (2) when plural governance is required based
ships exist and are relatively commonplace in practice on prevailing conditions elsewhere in a network.
(cf., Anderson et al. 1994), research into the governance
of these networks of relationships is needed. The concepts and propositions discussed in this paper
have a number of implications for theory and practice.
Continuing in the tradition of this emerging stream of Studying relationship networks and how they might be
literature, and seeking to extend previous, more descrip- efficiently governed in order to increase flexibility and
tive conceptualizations of relationship networks in mar- certainty for firms in these networks is of importance to
keting, this paper advances a more general conceptualiza- academic and practitioner audiences. At a theoretical
tion of governance in relationship networks. The paper level, the framework proposed in this paper is novel in the
focuses specifically on how governance system choices sense that it considers contingent effects across levels of
are made and what factors influence these decisions, and networks (e.g., upstream vs. downstream) in determining
the performance-related outcomes of systems of gover- the most appropriate systems of governance that are
nance. An important element of the conceptual discussion compatible with each other. This builds on recent work by
in this paper is the notion that governance choices for one Mishra et al. (1998) and Wathne and Heide (2004).
level in a network are strongly influenced by a firm’s Including plural governance in these conceptualizations
assessments and evaluations of conditions of uncertainty adds to the depth and richness of the framework. For
and relationship characteristics at other levels in the same practitioners, this paper offers suggestions for how to
network. The authors present an extended conceptual make optimal and efficient governance system choices
discussion and formal propositions in an attempt to ad- under conditions of uncertainty and complexity in a
vance thought towards a more general framework of relationship network context. The conceptualizations pro-
network governance. vide guidelines for what practitioners should consider and
evaluate as part of the process of designing the most
A central issue considered in the conceptual frame- appropriate governance systems for their needs. Refer-
work relates to the use of plural forms, or plural gover- ences available upon request.
nance, in managing multiple, linked relationships in net-

Andrew T. Stephen or Leonard V. Coote
UQ Business School
Brisbane, Qld 4072
Australia
Phone: +61.7.3365.9721
FAX: +61.7.3365.6988
E-Mail: a.stephen@business.uq.edu.au

MARKET DRIVING RELATIONSHIP MARKETING FOR
RADICAL INNOVATIONS
Helder J. Sebastiao, University of Oregon, Eugene
SUMMARY merce, and Dell driving the standard for modular com-
puter manufacturing processes (Kumar, Scheer, and Kotler
Firms that launch radical product and process inno- 2000).
vations are market driving: creating, shaping, and accel-
erating markets for their radical innovations and redefin- The market driving entrepreneur typically creates
ing customer expectations, value propositions, and busi- and enters new markets through alliances and similar
ness processes. Successful market driving is dependent on cooperative strategies rather than competitive positioning
developing, growing, and strategically leveraging net- (Sarasvathy 2001). Success is dependent on leveraging
works of alliances and key customers, and relationship and building upon who they know: industry contacts,
marketing plays a critical role in fostering collaboration. funding sources, supporters, suppliers, and customers
This suggests a fundamental shift in marketing focus from (Sarasvathy 2001). This emphasis on alliances and col-
targeting and communicating with prospective customers laboration is the most significant influence on the selec-
towards the cultivation, management, and leveraging of tion of relationship marketing objectives and strategies.
multiple collaborative relationship ties. Thus initial rela- Alliances facilitate widespread industry adoption and
tionship marketing efforts should be concentrated on provide access to markets and customers that otherwise
establishing strong alliance network ties and conducting may not be available to the entrepreneurial firm. Initial
frequent and intensive mutual learning-oriented commu- customers provide feedback and the input that drives
nications with initial customers. successive product iterations and service enhancements
(Gatignon and Xuereb 1997; Hill 1997).
Contrasting Being Market Driven with Market Driv-
ing Market driving firms both drive the formation of
expectations and use an iterative approach to refining new
Market Driven Firms Are Responsive: Market driven products and services to meet customer and alliance
relationship marketing strategies emphasize responsive- partner expectations that evolve through successive inter-
ness to both channel member and existing customer needs. action with their networks. Due to limited resources
When a firm is market driven its primary objectives are to entrepreneurs must achieve this incremental learning within
gain full channel member support of a product launch and a relatively short timeframe. Interaction with the firm’s
to leverage existing customers in adopting an incremental networks must also reinforce strategically important rela-
innovation, often through the use of product/service mi- tionships.
gration strategies. The result of a successful market driven
strategy should be expanded market share or share of Market Driving Creates an Innovation Network
wallet/budget, reinforced brand loyalty, and enhanced
existing customer satisfaction. While firms can maintain A collaboration-oriented relationship marketing strat-
competitive advantage by being market driven in existing egy should result in lower product launch costs, reduced
markets, they become vulnerable to firms pursuing a time for market acceptance, and reduced market uncer-
market driving strategy. tainty. Embedded relationships provide an additional bar-
rier against competitive threats and help reduce overall
Market Driving Firms Are Opportunistic and Col- marketing costs. The strategic leveraging of these rela-
laborative: Market driving firms are generally new en- tionships leads to ongoing product, service, and market
trants who revolutionize an industry by delivering a opportunities as the network expands and becomes in-
substantial leap in customer value through either a break- creasingly interconnected to other alliance and customer
through technology or marketing system made possible networks. The desired end state is status as an industry
by a unique business process (Kumar, Scheer, and Kotler standard and possession of a highly embedded network of
2000). The most successful radical innovations produce relationships contributing to future innovations based on
what Arthur (1990) refers to as a market in lock-out, joint value creation. This is a natural evolution for the
where a technology or technology-based business process initial customer relationships in which the seeds of col-
becomes an industry standard that is extremely difficult to laboration were planted. This environment of shared risk
dislodge. Examples include Amazon.com and EBay be- and reward fosters greater learning between the firm and
ing perceived as the standards for conducting e-com- the customer, which in turn can lead to improved efficien-

cies, higher quality, and faster turnaround times (Prahalad Table 1 summarizes the differences between market
and Ramaswamy 2000). driving and being market driven.
TABLE 1
Market Driving Versus Market Driven Relationship Marketing
Market Driving Market Driven
Type of Innovation Radical product and/or process innovation Incremental product / service
innovation
Firm Situation Typically entrepreneurial; new product to new markets New product/service to existing
markets
Basic marketing Essentially proactive, iterative, and committed to Essentially reactive/responsive to

philosophy creating technology standards and markets market demand
Basic strategic Strategy evolves from an iterative process, testing as Select an optimal strategy among
approach many alternatives as possible by leveraging alternatives based upon expected
♦ Firm technical, functional, market expertise returns, using
♦ Emerging networks of prospective customers/alliances ♦ Pre-specified criteria
♦ Collaboration ♦ Pre-determined goals
♦ Research and projections
♦ Knowledge of existing channels
and customers
Involvement of Initial target customer input is integral to the iterative Possible contact via context of NPD
target customer process of product/service and segment refinement; research
in the innovation Dialog is continuous, and network of potential Leverage existing relationships for
process customer/collaborators continually expands potential migration
Key Relationship Identifying, recruiting and forming networks of Customer adoption/migration to new
Mktg Strategies alliances and customers product
Building commitment to the technology as a standard Full channel cooperation
Customer as input to future product
ideas
Joint value creation via embedded network relationships Customer as referral source
Key Relationship (1) Well defined market (1) Expand market share
Mktg Goals (2) Standards as an effective barrier to entry (2) Expand share of wallet/budget
(3) Access to additional markets and opportunities (3) Reinforce brand loyalty
Key Mktg-Based (1) Industry standard technology or process (1) Expanded cash flow from
Assets from (2) Embedded product innovation network incremental sales
RM Activities acceptance time (2) Enhanced cash flow from
reduced market
Conclusion cesses. Successful market driving is dependent on devel-

oping, growing, and strategically leveraging networks of
Traditional relationship marketing models for estab- alliances and key customers. Relationship marketing plays
lished, market driven firms do not adequately account for a critical role in market driving and ultimate radical
the characteristics and challenges of entrepreneurial mar- innovation launch success, particularly in fostering col-
ket driving firms. These firms redefine markets with laboration with alliances and customers. References are
radical innovations that trigger dramatic changes in cus- available upon request.
tomer expectations, value propositions, and business pro-

Helder J. Sebastiao
University of Oregon
Lillis Business Complex
1208 University of Oregon
Eugene, OR 97403–1208
Phone: 541.346.4179
FAX: 541.346.3341
E-Mail: hsebasti@uoregon.edu

AN INVESTIGATION OF PERCEPTUAL FACTORS INFLUENCING
CONSUMER’S INTENTION TO ADOPT RADICAL VERSUS
INCREMENTAL NEW PRODUCTS
Audhesh K. Paswan, University of North Texas, Denton
Lisa C. Troy, University of North Texas, Denton
SUMMARY and inter-item correlations. The results indicate accept-

able levels of reliability and validity. We next tested the
The objective of this paper is to enhance our under- hypotheses using pair wise t-tests and multiple regres-
standing of key perceptual dimensions influencing con- sions to test the hypotheses of interest. Separate models
sumers’ intentions to adopt radical and incremental new were run using purchase intention towards the RNP and
products. We do so by investigating five key perceptual the INP, and for both models the multi-collinearity esti-
factors as antecedents to intention to adopt (or purchase) mates were well within acceptable limits.
a radical new product (RNP) or an incremental new
product (INP): perceived knowledge about the RNP and Findings from the study indeed indicate that the very
INP, perceived complexity of the RNP and INP, perceived same consumer can have similar adoption intentions for
financial risk and perceived performance assurance (op- different types of products (i.e., radical versus incremen-
posite of risk) associated with the RNP and INP, and tal), possibly desiring to adopt both a radical and an
perceived comfort associated with purchase of the INP or incremental new product within the same broad category
RNP. Our investigation of whether these perceptual fac- of consumer electronics. Additional findings suggest that
tors vary across radical versus incremental new products, consumers seem to differ in the way they perceive the two
and whether they influence the adoption of radical versus types of innovative products (radical versus incremental)
incremental new products similarly differs from existing along perceptual dimensions such as complexity, finan-
literature which typically focuses on consumer innova- cial risk and performance assurance dimensions but do
tiveness as a trait or a behavior and which often lumps not differ (in our sample) across product types in their
consumers together into radical versus incremental groups perceptions of perceived knowledge or perceived comfort
despite the fact that consumers may actually adopt incre- with the purchase process. Finally, we find that the influ-
mental products some of the time and radical products in ence of key perceptual factors on adoption intentions may
other situations. be similar for radical versus incremental new products.
The respondents for the study were composed partly Our findings are important for marketers who have
of senior and graduate students at a major university in a historically sought to lump consumers into innovative
suburban campus and partly of non-students contacted categories with the assumption that these would hold
personally through student interviewers from two market across products. Furthermore, consumers seem to differ in
research classes. The scale items for measuring the five the way they perceive the two types of innovative prod-
perceptual factors were adapted from existing studies. ucts along perceptual dimensions such as complexity,
The same scale items were repeated in two separate financial risk, and performance assurance dimensions but
sections of the survey – one for Incrementally New do not differ (in our sample) across product types in their
Products (INPs) and another for Radical New Products perceptions of perceived knowledge or perceived comfort
(RNPs) in the context of consumer electronics. A total of with the purchase process. These findings are also impor-
609 surveys were first analyzed for reliability and validity tant to marketers as they seek to design marketing plans to
(convergent and discriminant) using Cronbach’s Alpha address consumer’s perceptions.

Audhesh K. Paswan
Department of Marketing and Logistics
College of Business Administration
P.O. Box 311396
Denton, TX 76203–1396
Phone: 940.565.3121
FAX: 940.565.3837
E-Mail: paswana@unt.edu

IT’S ALL ABOUT LEARNING: WHAT FIRMS CAN LEARN FROM
CONSUMER PIONEERS
Yun Ye, The University of Arizona, Tucson
ABSTRACT literature and developing a mediation model. The model

indicates that the resource of consumer pioneers can
This paper identifies an implicit potential resource generate sustained competitive advantage through the
that can generate sustained competitive advantage for mediator of higher-level organizational learning, espe-
firms that produce innovative products: consumer pio- cially under high-level environmental uncertainty for
neers. Examining the mechanism of consumer learning firms that produce high-level innovative products.
and organizational learning, this paper emphasizes the
mediating role of higher-level organizational learning and In the next section, I will discuss the concept of
the dynamics how consumer pioneers generate sustained consumer pioneers from a consumer learning perspective,
competitive advantage. followed by a discussion of why the resource of consumer
pioneers could be a potential sustained competitive ad-
INTRODUCTION vantage by addressing its four attributes. Then I will
discuss the mediating role of higher-level organizational
Identifying resources that can generate sustained learning. Finally, a conceptual framework and some addi-
competitive advantage has been an important area of tional propositions are presented to build a comprehen-
study in strategic management. In the existing literature, sive mediation model incorporating consumer pioneers,
some unique resources, such as strategic planning, infor- organizational learning and sustained competitive advan-
mation processing systems and positive reputations, are tage.
considered to be able to generate sustained competitive
advantage for firms (Barney 1991), but resources which CONSUMER LEARNING AND
can generate sustained competitive advantage have be- CONSUMER PIONEERS
come more implicit and limited today. As environmental
uncertainty is growing faster in this technology era, it has Consumer Learning
become more crucial to exploit new resources to serve as
sustained competitive advantages for firms, especially for Consumers learn, but their learning varies. The basic
those firms whose products are highly innovative and definition of learning, which refers to the relatively endur-
need to be upgraded constantly, such as digital product ing changes in the response to stimuli (Rogers 1962),
industry (e.g., computer, digital camera, PDA, etc.) already points out the key issue: change. But how do
changes happen? How are they different from consumer
Thus, the purpose of this study is to address another to consumer? And what does the difference mean to
implicit and important potential resource that can gener- researchers and marketers? These are the major issues I
ate sustained competitive advantage for firms that pro- will address here. Consumer learning occurs not only
duce innovative products: consumer pioneers. Defined as through exposure to external information sources such as
lead consumers with high learning capability and low advertising and word-of-mouth, but also through a pro-
learning costs, consumer pioneers as a resource have four cess of internal knowledge transfer from familiar to novel
empirical indicators of the potential to generate sustained domains (Gregan-Paxton and John 1997). Employing the
competitive advantage: value, rareness, imperfect imita- internal knowledge transfer view, this study focuses on
bility, and non-substitutability (Barney 1991). The under- the endogenous mechanism of consumer learning. Two
lying idea is learning transfer. That is, firms can generate popular views that apply to the endogenous mechanism of
consumer pioneers to sustained competitive advantage by consumer learning are the hierarchical connections be-
learning from what consumer pioneers have learned. Prior tween stored pieces of information (Miller 1956) and the
studies have considered working with lead customers to model of Consumer Learning by Analogy (Gregan-Paxton
recognize strong needs before the rest of the market and to and John 1997). According to the hierarchical connection
find solutions to those needs as a salient strategy for firms theory, the information previously stored in consumers’
to gain competitive advantage (Slater and Narver 1995). memory will associate their learning when they encounter
But what is the underlying mechanism that such a re- similar information. Similarly, the model of Consumer
source generates sustained competitive advantage? Un- Learning by Analogy predicts that previously acquired
der what conditions does this resource become crucial to knowledge is transferred by analogy in the process of
firms, and to what kind of firms? This article provides consumer learning. These two views reveal that learning
answers to the above questions by examining the relevant is a relatively harder task to consumers when they encoun-

ter an innovative or a new product and use the product for located on different positions of the diffusion curve.
the first time, as they may not be able to refer to similar Among them there exists such a group that enjoys learn-
previous experience. Thus, firms that have innovative ing, trying and sometimes adopting new products, espe-
products face harder tasks of consumer learning than cially high technological products such as digital cam-
those firms that don’t. eras, MP3 players, new models of computer, and hybrid
electric vehicles. These consumer pioneers are active and
Innovation Adoption and Diffusion effective learners compared to average consumers. Their
learning usually does not stop at adoption, or even at
As Rogers (1962) stated, the process by which inno- rejection of a product. Unlike average consumers, these
vations are adopted by consumers is a special example of pioneers will extend their learning beyond adoption or
consumer learning. There are five stages in the adoption rejection with high interest and enthusiasm. For example,
process: awareness, interest, evaluation, trial and adop- after adoption, when they find what they like or dislike
tion. Depending on the degree of innovativeness, differ- about the product, they will express their satisfaction,
ent consumers belong to different categories of the diffu- dissatisfaction, or further demands to the marketers. On
sion curve: innovators, early adopters, early majority, late the other hand, if they reject the product, they clearly
majority, and laggards. Consumer innovation adoption know why they rejected the product and what needs to be
and diffusion are especially important to firms entering improved in the future. In both cases, they are willing to
the market with innovative products because consumers’ deliver their perceptions (e.g., satisfaction, expectations,
learning costs are usually high for such products. Just as and demands) to others (other consumers or the firm). In
Mistri (2002) indicated, “the consumer’s learning behav- the beginning, when they are interested in the product,
ior presupposes that he is in a situation of bounded they can be very passionate and highly motivated. Later,
rationality and must proceed with a progressively more when they evaluate the product, they can be very rational
and more refined classification of a product or group of and professional.
products, and that he is seeking a strategy that enable him
to classify the goods with a minimum research effort” In this article I use the term “consumer pioneers”
(p. 310). Thus, the main barrier for consumers to adopt an rather than “consumer experts” or “innovators/early adopt-
innovative product is the relatively high learning cost. ers” because these three groups are from different con-
sumer behavior perspectives. Precisely, consumer ex-
Learning Costs perts emphasize expertise and knowledge (consumer be-
havior status) and early adopters emphasize adoption
Since most of the time consumers do not have com- (consumer behavior consequence), while consumer pio-
plete information of the product, they are more likely to neers emphasize learning (consumer behavior process).
learn inaccurately and misperceive product quality. Per- Consumer experts are “people whose prior knowledge is
fect learning only occurs when consumers can accurately well developed, in part because they have had a lot of
evaluate product quality (Goering 1985). Not all consum- experience, knowledge, and familiarity with an object or
ers can perform perfect learning for all products. a task” (Hoyer and MacInnis 1997, p. 101). Thus, these
Gabszewicz, Pepall, and Thisse (1992) explained “as with consumers may have high expertise in necessary choice
most new products, consumers need to learn how to use and purchase, but may not be willing to learn and try new
the new good, and consumers differ in their ability to products. Some firms use consumer experts in test mar-
learn, or in their willingness to adapt to something new” keting because of their expertise. However, since they are
(p. 399). Thus, the combination of consumers’ willing- not necessarily interested in new products in the first
ness and ability to learn can be seen as their learning place, their opinions may not be that reliable to represent
capability. As Gabszewicz, Pepall, and Thisse (1992) average consumers in the real market, and may cause
stated, a consumer with low learning capability has a high firms to overestimate the market. In contrast, consumer
learning cost, and is predicted to dislike trying new pioneers may not have as much expertise as experts, but
products. Meanwhile, a consumer with high learning they have more enthusiasm to learn and try new products
capability has a low learning cost and will enjoy trying as well as communicate with marketers and other consum-
innovative products. Thus, I refer to consumers with high ers. Additionally, since their actions are more spontane-
learning capability and low learning costs as consumer ous, firms could lower consumers’ learning costs by
pioneers, and I refer to those with low learning capability studying consumer pioneers’ learning processes. This
and high learning costs as average consumers. may not only help the firm to explore flaws of their
products, but also to accelerate the diffusion process.
Consumer Pioneers Also, a consumer pioneer can be a member of either the
first two categories of adopters: innovators and early
As we can see, for an innovative product, there are adopters. The reason it is not necessary to distinguish
always various groups of consumers who have different consumer pioneers from innovators or early adopters is
learning capabilities and adoption actions, and they are that this article focuses on the process of consumers’

choice (learning) rather than the consequence of choice consumers and market really expect from the product and
(adoption). It’s not important whether a consumer pioneer exploit new opportunities. Second, as mentioned earlier,
eventually adopts the product, but he/she engages in the since high learning cost is a main barrier for average
learning process with high learning capability during the consumers to adopt new products, firms may exploit the
choice. strategy that can lower consumers’ learning costs through
consumer pioneers’ experience. For example, firms could
Nowadays, it is common to see numerous online find how to improve consumers’ willingness to try a new
communities in which people communicate with each product by stimulating their interest, which can be discov-
other about their experiences with certain products and ered from consumer pioneers’ experience. Learning from
brands. Consumer pioneers often are active members, consumer pioneers’ experience, firms can also improve
though not necessarily opinion leaders (Rogers 1962). consumers’ learning ability by designing education pro-
One salient characteristic of them is their high willingness grams and upgrading products. Thus, the above argument
to communicate with other consumers and also with indicates that consumer pioneers can be considered a
marketers. Unlike consumer experts, who are usually valuable source that can help firms to exploit new oppor-
located by firms to ask for opinions, consumer pioneers tunities in the market, and finally generate sustained
voluntarily locate firms to provide suggestions. Such competitive advantage.
activities include writing online reviews, providing feed-
back cards, e-mailing to technology support departments, Rareness
and so on. The underlying motivation could be their needs
for self-esteem and self-fulfillment, which needs further Just as not all firms can have sustained competitive
research. But at least one thing is clear: such communica- advantage, not all consumers can be consumer pioneers.
tion between the marketers and consumers as well as that Only those active and effective learners with high learn-
among consumers is crucial to innovation adoption and ing willingness and ability can be consumer pioneers. A
diffusion processes (Schiffman and Kanuk 2000). few factors put constraints on consumers’ learning will-
ingness and ability, such as interest, financial support,
CONSUMER PIONEERS AS A POTENTIAL education level, time, and so on. For example, a consumer
RESOURCE OF SUSTAINED COMPETITIVE with a tight schedule may not have time to try new
ADVANTAGE products even though he may be interested in them. A
consumer with limited financial support also will not try
The definition of sustained competitive advantage new products because he will not risk his money at all
that Barney (1991) provided is “a firm is said to have a (Rogers 1962). A consumer with limited education may
sustained competitive advantage when it is implementing have difficulty understanding a high technology product
a value creating strategy not simultaneously being imple- and may be unable to provide insightful comments even
mented by any current or potential competitors, and when after he has adopted it. Thus, for certain product catego-
these other firms are unable to duplicate the benefits of ries, there is a limited amount of consumer pioneers, and
this strategy” (p. 102). Also according to Barney (1991), they compose a rare resource for firms to perform compe-
to have the potential of sustained competitive advantage, tition.
the resource must have four attributes: value, rareness,
imperfect imitability, and non-substitutability. To argue Imperfect Imitability
that consumer pioneers are a potential source of sustained
competitive advantage, I first examine whether it has To be a potential source of sustained competitive
these attributes. advantage, the source must be imperfectly imitable. That
is, firms that do not possess it cannot obtain it. Barney
Value (1991) proposed how firm resources can be imperfect
imitable. One possibility is that the resource is socially
A firm’s resource is considered valuable if it exploits complex. Are consumer pioneers such a socially complex
opportunities in a firm’s environment. Consumer pio- resource? The answer is positive if a firm can create long-
neers are such a valuable resource in two ways. First, as term and stable relationships with them. However, it is not
mentioned above, consumer pioneers are willing to pro- an easy task. It requires unique pursuing strategies for the
vide valuable evaluations and suggestions for products, firm to hold a pool of its own loyal consumer pioneers, and
especially new products, to firms. These evaluations and this complex relationship is hard to be mimicked by other
suggestions help firms accurately estimate their product firms. Moreover, if a firm can create the reputation among
and market and plan future strategy. Since the comments consumer pioneers that it is a superior relationship holder,
from consumer pioneers are representative of average this imperfect imitability will be durable.
consumers to a certain degree, firms may identify what

Non-Substitutability Lower-level learning, also called adaptive learning (Senge
1990) or single-loop learning (Argyris 1977), occurs
A potential resource of sustained competitive advan- within a given organizational structure when organiza-
tage cannot be strategically substituted. The substitutes tional contexts are well understood and reflect the
for this resource may be valuable but are neither rare nor organization’s assumptions about its environment and
imperfectly imitable. As to consumer pioneers, as long as itself. In contrast, higher-level learning, also called gen-
the firm can maintain a stable relationship with them, they erative learning (Senge 1990) or double-loop learning
are non-substitutable. One reason is that consumer pio- (Argyris 1977), tries to adjust overall rules and norms
neers are mobile and time-effective. For a particular brand rather than specific activities or behaviors. This type of
the consumer pioneer pool is unique and willing to serve learning is a more cognitive process and the associations
the firm who locate them first. Moreover, even competi- resulting from it have long-term effects and impacts on the
tors can locate them later, often the optimal timing (e.g., organization as a whole. Obviously, learning from con-
introductory period) has passed, and the consumer pio- sumer pioneers is such a higher-level learning. In the
neer pool is no longer effective for the product. Of course, three-stage process of learning that includes information
this argument is based on some assumptions. That is, acquisition, information dissemination, and shared inter-
consumer pioneers usually have limited effort and are pretation (Slater and Narver 1995), learning from con-
easier to obtain brand loyalty for the firm that cares them sumer pioneers is crucial in the stage of information
most. For example, a consumer pioneer who serves acquisition for innovative firms. When organizations per-
Microsoft may be hard to serve Macintosh simultaneously, form this kind of learning from consumer pioneers, they
not only due to his preference for Microsoft but also due should not only focus on the openness to such external
to his limited time or expertise. “learning partners” (Slater and Narver 1995), but also
create an effective interface and long-term relationship
In summary, possessing these four attributes – value, with them, thus establishing a superior organizational
rareness, imperfect imitability, and non-substitutability – culture and reputation among customers. This type of
consumer pioneers can be considered a potential resource higher-level learning can be considered market-based
of sustained competitive advantage. Thus, the first hy- organizational learning.
pothesis is,
Market-Based Organizational Learning and Market-
H1: Consumer pioneers are a potential resource of sus- Based Absorptive Capacity
tained competitive advantage for firms.
As Sinkula (1994) stated, market-based organiza-
But this poses another question: how can this re- tional learning is different from other types of learning in
source generate sustained competitive advantage? Does five ways. First, it focuses on external information (e.g.,
every firm have the ability to generate this resource to customers and suppliers) and it is not as explicit as most
sustained competitive advantage? The answer lies on internally focused organizational learning. Second, it
organizational learning. As Cravens (2000) addressed, results in the fundamental bases of competitive advan-
firms can benefit from organizational learning which tage. Third, it pays more attention to the observation of
helps them quickly respond to opportunities and threats, competitors. Fourth, the information it acquires and stores
reduce time to develop new products, improve existing in organizational memory is difficult to access. Finally,
products and services, and accelerate new product adop- market-information is equivocal. From these five at-
tion. In the next section, I will examine the mediating role tributes, we can see that learning from consumer pioneers
of organizational learning in the relationship between is such a market-based organizational learning. Con-
consumer pioneers and sustained competitive advantage. sumer pioneers provide “access to a greater number of
information sources, force the development of mecha-
CONSUMER PIONEERS AND nisms that facilitate the sharing of information, and offer
ORGANIZATIONAL LEARNING alternative perspectives to the meaning of critical infor-
mation that could lead to generative learning” (Slater and
Organizational Learning Narver 1995, p. 70). Furthermore, this kind of “informa-
tion-driven” (Goldstein and Zack 1989) and “market-
Foil and Lyles (1985) defined organizational learn- driven” (Jaworski, Kohli, and Sahay 2000) learning is
ing as “the development of insights, knowledge, and especially important to innovative firms because they face
associations between past actions, the effectiveness of barriers of high learning costs for average consumers in
those actions, and future actions” and adaptation as “the their innovation adoption. Two dimensions underlying
ability to make incremental adjustments as a result of innovation – technology and market – suggest that devel-
environmental changes, goal structure changes, or other oping new products needs not only technology that is
changes” (p. 811). They also classified organizational different from prior technology, but also fulfilling key
learning into two levels: lower- and higher-level learning. customer needs better than existing products (Chandy and

Tellis 1998). Thus, even though the primary goal of the experience difficulties in innovation learning (Engelland
innovative firms is to drive the market to a new direction, and Alford 2000), the information provided by consumer
identifying which direction is most likely to be successful pioneers will help the firm match average consumers’
based on consumers’ needs is at least as, if not more, expectations to increase their willingness to learn, make
critical as new technology development. How does one product quality or features less ambiguous, and finally
identify the right direction? In this article, the answer lies lower their learning costs and diminish the entry barriers.
in market-based absorptive capacity from consumer pio- Thus, I argue that higher-level organizational learning,
neers. i.e., market-based organizational learning and absorptive
capacity, serves as a mediator in the relationship between
Absorptive capacity, also called innovative capabil- consumer pioneers and sustained competitive advantage.
ity, is defined as the ability of a firm to recognize the value
of new, external information, assimilate it, and apply it to H2: The resource of consumer pioneers can generate
commercial ends (Cohen and Levinthal 1990). As Cohen sustained competitive advantage by the firm’s higher-
and Levinthal (1990) addressed, “absorptive capacity level organizational learning (i.e., market-based or-
refers not only to the acquisition or assimilation of infor- ganizational learning and market-based absorptive
mation by an organization but also to the organization’s capacity).
ability to exploit it” (p. 131). Thus, the ability to exploit
market-based knowledge and information can be seen as CONCEPTUAL FRAMEWORK AND
market-based absorptive capacity, which is a special form OVERVIEW OF THE METHOD
of higher-level organizational learning. This exploration
ability, which will eventually determine the firm’s inno- Employing the knowledge transfer theory, the con-
vative capability, is crucial to the firm’s success. An ceptual model illustrates the knowledge transfer mecha-
organization’s market-based absorptive capacity depends nism which emphasizes that firms learn from what con-
mostly on its direct interface with the external environ- sumer pioneers learn and generate the tacit knowledge to
ment (e.g., customers, suppliers, competitors). Since in- sustained competitive advantage (Figure 1).
novative firms face more barriers of high learning costs
from consumers than non-innovative firms, they espe- There are two moderators in the model. One is envi-
cially benefit from working with consumer pioneers to ronmental uncertainty, including market and technology
discover effective ways to lower learning costs for aver- uncertainty. Prior research shows that under high environ-
age consumers when develop new products (e.g., radical mental uncertainty, customer orientation has a positive
product innovation) or upgrade existing products (e.g., influence on firms’ innovation because the development
incremental innovation). As average consumers usually of a new product in a highly uncertain environment
FIGURE 1
Conceptual Framework of Relationship Among Consumer Pioneers, Higher-Level
Organizational Learning, and Sustained Competitive Advantage
Product
Environ- Innovative
mental Level
Uncertainty
Higher-Level
Organizational
Learning
Consumer Sustained
Pioneers Competitive
Advantage

creates the need for more market scanning and network- H4: The higher the innovative level of the product, the
ing with users to identify customer needs (Gatignon and more the likelihood that the resource of consumer
Xuereb 1997). Thus the third hypothesis is: pioneers can generate sustained competitive advan-
tage for the firm via higher-level organizational learn-
H3: The higher the environmental uncertainty, the more ing.
the likelihood that the resource of consumer pioneers
can generate sustained competitive advantage for the Method
firm via higher-level organizational learning.
The possible method will be a survey among key
Another moderator is the innovative level of the informants in innovative product industries. These infor-
product. As I stated before, consumers’ learning costs are mants include firms’ gatekeepers or knowledge manag-
higher for radical products than for incremental products ers, who monitor the environment and translate external
(Chandy and Tellis 1998). Therefore, it is especially information into understandable internal knowledge to
important for firms that produce radical products to re- the research group (Cohen and Levinthal 1990). Another
main close with their customers, lower their learning costs survey will be conducted using relevant industry data-
and increase their adoption rates via consumer pioneers. bases to examine the performance of these firms’ R&D.
So the final hypothesis is:
REFERENCES nal of Consumer Research, 12, 74–82.

Goldstein, David K. and Michael H. Zack (1989), “The
Argyris, Chris (1977), “Double Loop Learning in Organi- Impact of Marketing Information Supply on Product
zations,” Harvard Business Review, 55, 115–25. Managers: An Organizational Processing Perspec-
Barney, Jay (1991), “Firm Resources and Sustained Com- tive,” Information Technology & People, 4, 313–36.
petitive Advantage,” Journal of Management, 17, Gregan-Paxton, Jennifer and Deborah R. John (1997),
99–120. “Consumer Learning by Analogy: A Model of Inter-
Chandy, Rajesh K. and Gerard J. Tellis (1998), “Organiz- nal Knowledge Transfer,” Journal of Consumer Re-
ing for Radical Product Innovation: The Overlooked search, 24, 266–84.
Role of Willingness to Cannibalize,” Journal of Hoyer, Wayne D. and Deborah J. MacInnis (1997), Con-
Marketing Research, 35, 474–87. sumer Behavior. Boston: Houghton Mifflin.
Cohen, Wesley M. and Daniel A. Levinthal (1990), “Ab- Jaworski, Bernard, Ajay K. Kohli, and Arvind Sahay
sorptive Capacity: A New Perspective on Learning (2000), “Market-Driven versus Driving Markets,”
and Innovation,” Administrative Science Quarterly, Academy of Marketing Science, 28, 45–54.
35, 128–52. Miller, George A. (1956), “The Magical Number Seven,
Cravens, David W. (2000), Strategic Marketing. Boston: Plus or Minus Two: Some Limits on Our Capacity to
Irwin McGraw-Hill. Process Information,” Psychological Review, 63, 81–
Engelland, Brain T. and Bruce L. Alford (2000), “Con- 97.
sumer Learning and The Creation of Primacy Advan- Mistri, Maurizio (2002), “Consumer Learning, Connec-
tages for Followers,” Journal of Business Strategies, tionism and Hayek’s Theoretical Legacy,” Eastern
17, 145–62. Economic Journal, 28, 301–17.
Foil, C. Marlene and Marjorie A. Lyles (1985), “Organi- Rogers, Everett M. (1963), Diffusion of Innovation. New
zational Learning,” The Academy of Management York: The Free Press.
Review, 10, 803–13. Schiffman, Leon G. and Leslie L.Kanuk (2000), Con-
Gabszewicz, Jean, Lynne Pepall, and Jacques-Francois sumer Behavior. Upper Saddle River: Prentice Hall.
Thisse (1992), “Sequential Entry with Brand Loyalty Senge, Peter M. (1990), The Fifth Discipline. New York:
Caused by Consumer Learning-by-Using,” Journal Doubleday.
of Industrial Economics, 40, 397–416. Sinkula, James M. (1994), “Market Information Process-
Gatignon, Hubert and Jean-Marc Xuereb (1997), “Strate- ing and Organizational Learning,” Journal of Mar-
gic Orientation of the Firm and New Product Perfor- keting, 58, 35–45.
mance,” Journal of Marketing Research, 34, 77–90. Slater, Stanley F. and John C. Narver (1995), “Market
Goering, Patricia A. (1985), “Effects of Product Trial on Orientation and the Learning Organization,” Journal
Consumer Expectations, Demand, and Prices,” Jour- of Marketing, 59, 63–74.

Yun (Serah) Ye
The University of Arizona
320 McClelland Hall
1130 E. Helen
P.O. Box 210108
Tucson, AZ 85721
Phone: 520.621.9179
FAX: 520.621.7483
E-Mail: yye@email.arizona.edu

WILL CONSUMERS PREFER GLOBAL OR LOCAL BRANDS?
THE ROLE OF IDENTITY ACCESSIBILITY IN CONSUMER
PREFERENCE FOR GLOBAL VERSUS LOCAL BRANDS
Yinlong Zhang, The University of Texas at San Antonio, San Antonio
Lawrence Feick, University of Pittsburgh, Pittsburgh
Vikas Mittal, University of Pittsburgh, Pittsburgh
SUMMARY words, being global can induce better responses from

consumers sometimes, but being local can induce better
Globalization is becoming the buzzword of this cen- responses in other situations.
tury. Business executives and marketing academics are
discussing this term to explore the profound changes that Three experiments were proposed to investigate the
will happen in marketing practice. Many marketers now- effect of accessibility of global versus local identity con-
adays face decisions on whether to market their products struct in consumer preference between global and local
globally or locally. One aspect of this global versus local brands. Study one manipulates the relative accessibility of
decision is whether to emphasize their market brands as global versus local identity, inducing effects in consumer
global or local ones. For example, a recent trade article preference for global or local brands. Study two measures
(Marketing News 2004) reported that many new soft the relative accessibility of the consumer’s global and
drink companies are making decisions on whether to local identity, and obtains the same pattern of results as
stress their brands as global, like Pepsi or Coke, or as local Study one on consumer preference. The first two studies
based on what can be proven to better their chances for offer convergent evidence supporting our hypothesis:
success in markets that are both increasingly globalized When a global identity is activated, consumers tend to
and localized. value the global brands more highly than the local brands;
when the local identity is activated, consumers tend to
From a theoretical perspective, this decision asks value the local brands better than the global brands. Study
whether emphasizing a brand as global or local will three manipulates the differential versus integration modes
induce positive responses from consumers. Marketing to replicate the results similar to the one manipulating the
literature offers contradictory empirical answers for this accessibility of local versus global identity. This will, in
question. For example, Steenkamp, Batra, and Alden turn offer the initial evidence that the effect of global
(2003) found that labeling a brand global instead of local versus local identity on consumer preferences runs through
can increase its appeal to customers even if the objective the differential versus integration ways of thinking. Impli-
product quality is the same. While Zambuni (1993) found cations from these three studies on globalization and
the opposite: consumers like it more when a brand is accessibility-diagnosticity are discussed. Through these
emphasized as a local one than when it is emphasized as studies we make a key theoretical contribution to the
a global one. Along the same line, De Mooij (1998) literature of globalization, since we offer a powerful
argued that localization strategy seems to be more attrac- theoretical explanation for the inconsistent findings in
tive than the globalization strategy for consumer prod- past studies. Our results are consistent with the recent
ucts. These inconsistent findings on global versus local development in the literature of cross-cultural psycholo-
suggest that some boundary conditions moderate the main gy. Our studies are also consistent with the recent devel-
effect of global or local on consumer preference. In other opment in the accessibility diagnosticity framework.

Yinlong “Allen” Zhang
The University of Texas at San Antonio
6900 North Loop 1604 West
San Antonio, TX 78249
Phone: 210.458.6331
FAX: 210.458.6335
E-Mail: yzhag@utsa.edu

WHY DO CONFERENCE GOERS RETURN? A MODEL OF INTENTIONS
TO ATTEND AND RECOMMEND
Annie H. Liu, Loyola Marymount University, Los Angeles
Mark P. Leach, Loyola Marymount University, Los Angeles
Robert D. Winsor, Loyola Marymount University, Los Angeles
SUMMARY a cost-benefit analysis [perceived value assessment], or

an analysis of how well expectations were met [perceived
Abstract satisfaction assessment]). Underlying these hypotheses is
our premise that the high degree of elaboration associated
In the face of competition for limited vacation and with these assessments makes attitudes more easily acces-
leisure time, meeting planners and membership associa- sible and highly diagnostic.
tions have traditionally been interested in how to attract
new members and enhance attendance at organizational We were further interested in members’ intentions to
conferences. This study empirically investigates mem- recommend a conference to others. By evaluating the
bers’ intentions to attend organizational conferences us- relationship between attitudes regarding a conference and
ing an accessibility-diagnosticity framework. Findings referral intentions for past attendees, and the relationship
suggest that post-attendance attitudes are the most impor- between word-of-mouth information and attendance in-
tant factors predicting intentions to (1) attend a future tentions for non-attendees, this study will provide an
conference and (2) recommend the conference to others. empirical test of whether or not a reinforcing feedback
Furthermore, what one has heard about the conference cycle exists for conference attendance.
(i.e., the buzz) and one’s perceived congruence with the
values of the organization are also found to be important Model, Methodology, and Results
when predicting attendance intentions.
Our base model incorporates three antecedent vari-
Purpose of the Research and Hypotheses ables predicting attendance: perceived organizational
congruence, word-of-mouth information, and attitude to-
The purpose of this research is to investigate mem- ward the conference. In addition, this model includes
bers of professional organizations regarding their inten- relationships between perceived organizational congru-
tions to attend conferences. Toward this goal, we examine ence and attitude, and word-of-mouth information and
three types of information that might be relevant to the attitude. An expanded model includes intentions to rec-
formation of conference intentions. These are: (1) the ommend, and adds three direct relationships between
member’s attitude toward the conference, (2) information perceived organizational congruence and referral inten-
acquired through word-of-mouth (i.e., the conference tions, word-of-mouth information and referral intentions,
buzz), and (3) the member’s perceived congruence with and attitude toward the conference and referral intentions.
the values of the organization. We further utilize the
accessibility-diagnosticity framework to provide a theo- A sample of 800 members of an organization for
retical rationale for why the relative importance of these retired employees of a government agency was randomly
three information types might differ between attendees drawn from the population of approximately 8,000 orga-
(i.e., those members who have previously experienced nizational members, and mailed a questionnaire. The total
one or more of the organization’s conferences) and non- number of useful and complete responses obtained was
attendees. Specifically, it is hypothesized that for attend- 448, representing a 58 percent response rate.
ees, attitudes toward the conference will be more diagnos-
tic with regard to intentions to attend future conferences. Structural equation methodology was used to test the
For non-attendees, conversely, their attitudes will be less hypothesized relationships. Results indicated that for the
diagnostic, and therefore other information sources will sub-sample that had attended a recent conference, inten-
be accessed and utilized. tion to attend a future conference was directly related only
to one’s attitude toward the conference. In addition,
In addition, it is hypothesized that for non-attendees, attendees’ attitudes toward the conference were found to
attitudes will be formed primarily based on what they have be related to perceived organizational congruence and to
heard about the conference, and from their general feel- information they received from word-of-mouth. The di-
ings about the organization. For attendees, attitudes will rect relationship between one’s perceived organizational
instead be formed upon more elaborative judgments (e.g., congruence and intention to attend was not found to be

statistically different from zero. Neither was the relation- tion, and their attitude toward the conference. In addition,
ship between word-of-mouth information and intention to one’s perceived organizational congruence and the infor-
attend. Thus, for those who have attended, the effects of mation acquired through word-of-mouth were found to be
perceived congruence and word-of-mouth information related to one’s attitude toward the conference. Again, the
appear to be fully mediated by one’s attitude. direct relationship between the word-of-mouth informa-
tion and intention to attend was not found to be significant,
For the sub-sample without a recent conference expe- suggesting that this relationship is fully mediated by one’s
rience, intention to attend a conference was directly re- attitude toward the conference.
lated to one’s perceived congruence with the organiza-

Annie H. Liu
College of Business Administration
Loyola Marymount University
One LMU Drive
Los Angeles, CA 90045
Phone: 310.338.3039
FAX: 310.338.3000
E-Mail: aliu@lmu.edu

INFLUENCE OF OTHER CUSTOMERS: A SCALE DEVELOPMENT
E. Deanne Brocato, The University of Texas at Arlington, Arlington
Susan B. Kleiser, The University of Texas at Arlington, Arlington
SUMMARY to create service environments where the customers that

enter the environment are compatible. It makes intuitive
Services literature has recognized the role customers sense that customers will “gravitate toward those service
play as an integral part of the service environment (Baker environments with which they are most compatible” (Mar-
1987; Bitner 1992; Grove and Fisk 1997). Other custom- tin and Pranter 1989, p. 7). In line with compatibility, the
ers in many instances can be viewed as part of the service visual characteristics (appearance, dress, gender, ethnicity)
environment (Martin and Pranter 1989), since they are of other customers can have an influence on the service
often required to share the same facility (Baker 1987). environment. Age, income, and social class have been
Drawing on inference theory (Huber and McCann 1982), suggested as characteristics that should influence percep-
other customers within a service environment provide tions of the environment (Baker 1987). Other patrons’
cues, which in turn are used by consumers to make behaviors can have profound effects on customers. Grove
assessments of quality (Baker, Parasuraman, Grewal, and and Fisk (1997) found in an exploratory study at a theme
Voss 2002). “The appearance, behavior, and number of park, behaviors of other customers affected consumer
other customers and contact personnel can clearly affect both positively and negatively. For example, other cus-
the way consumers perceive the service firm” (Baker tomers cutting in line and shoving during waiting for
1987, p. 79). The importance of service personnel has service created dissatisfying service encounters and wait-
been documented in several studies (Parasuraman, ing in line with polite patrons led to satisfying encounters.
Zeithaml, Berry 1988; Bitner 1992; Baker 2002), but In another exploratory study by Martin (1996), general
there appears to be a gap in the extant literature concerning factors of behaviors were found to influence patrons’
the influence of other customers on the perceptions of evaluations of service satisfaction.
service facility quality and ultimately on service evalua-
tions, such as satisfaction. The lack of attention devoted to Scale Development
the influence of other customers is surprising since many
times “the number and duration of such c-c [customer to Scale items for the dimensions of compatibility,
customer] encounters exceed that of employee-customer visual cues and behavior were generated drawing from
encounters by an order of magnitude” (Martin and Clark services literature that has explored customer-to-customer
1996, p. 347). Furthermore, the presence of other custom- interactions (Martin and Pranter 1989; McGrath and Otnes
ers in a service environment may influence, albeit directly 1995; Grove and Fisk 1997). Forty-one (41) items were
or indirectly, one’s satisfaction (Martin and Pranter 1989). developed in order to capture the full conceptual domain
In addition, frequent interactions among customers can of the construct. After item purification 11 items remained
lead to either a more favorable or less favorable service in the model representing the three dimensions. Evidence
experience depending on the type of interaction (e.g., was found for convergent and discriminant validities.
friendly patrons, crying children, cursing, etc.). This pa- Also, all Cronbach’s alpha reliability coefficients ex-
per sets forth to create the Other Customer Influence ceeded the .7 lower limit (Hair et al. 1998) with the
(OCI) scale to measure the influence that other customers dimension reliability estimates ranging from .74 to .83.
have on one’s evaluation of facility quality and ultimately
on satisfaction with the service experience. Nomological validity was assessed by investigating
the hypothesized paths between the three dimensions of
Dimensions of Other Customer Influence OCI and the facility quality construct. The model fit was
assessed and found to be sufficient: (GFI .90; AGFI .86;
Drawing on services literature (Martin and Pranter CFI = .98; RMSEA = .065). The chi-square value was
1989; Grove and Fisk 1997; Martin and Clark 1996; significant (χ2 = 278.40, 128 df, p < .00), however this was
McGrath and Otnes 1995; Baker 1987; Baker et al. 2002) expected due to the sample size (Marsh, Balla, and
three dimensions of the Other Customer Influence (OCI) McDonald 1988). Positive links between compatibility
construct are posited. These dimensions are compatibil- and behavior to facility were found to be significant (p <
ity, visual cues and behavior. We define compatibility as: .05) as indicated by the positive gamma (γ) parameters.
capable of existing or performing in harmonious, agree- The parameter linking visual cues and facility quality was
able, or congenial combination with another or others. not found to be significant. Finally, the path from facility
Martin and Pranter (1989) address the need for managers quality to satisfaction was significant.

Discussion, Limitations, and Future Research of the service environment provides the justification for
this new measure. This measure is a starting point for
One study cannot provide sufficient evidence of a future investigations into the construct of OCI. Each
valid measure. However, this study lays the foundation dimension should be investigated and expanded as more
for the continuing process of construct validation. The empirical evidence is gathered. Continual refinement is
evidence provided indicates the OCI scale is a valid needed for all measures and this measure is no different.
measure. This measure provides marketers with the abil- The complexity of the construct necessitates the need for
ity to determine how other patrons’ social factors influ- future research and investigation. References available
ence evaluations of facility quality and service satisfac- upon request.
tion. The recognition of other customers as an integral part

E. Deanne Brocato
The University of Texas at Arlington
Box 19469
Arlington, TX 76019–0469
Phone: 817.272.6765
FAX: 817.272.2854
E-Mail: brocato@uta.edu

BEYOND JUST BEING THERE: AN EXAMINATION OF THE IMPACT OF
ATTITUDES, MATERIALISM, AND SELF-ESTEEM ON THE QUALITY
OF HELPING BEHAVIOR IN YOUTH VOLUNTEERS
Elten Briggs, University of Oklahoma, Norman
Tim Landry, University of Oklahoma, Norman
Charles Wood, University of Tulsa, Tulsa
Todd Arnold, Oklahoma State University, Stillwater
SUMMARY H2a: The negative relationship between materialism

and the decision to volunteer is mediated by atti-
Young people represent a strong and growing source tude towards the task.
of volunteers for not-for-profit organizations (NPO), and
represent an important focus for NPO marketing efforts. H2b: The negative relationship between materialism
Therefore, it is in the long-term interests of NPOs that they and goal setting is the mediated by attitude towards
develop a better understanding of teenage volunteerism. the task.
Despite the importance of the topic, research concerning
the volunteer sector in the marketing literature has been H3a: The relationship between self-esteem and the deci-
scant historically (Bendapudi, Singh, and Bendapudi 1996; sion to volunteer is mediated by attitude towards
Fisher and Acherman 1998; Wymer and Starnes 2001), the task.
and no marketing studies have focused on the challenges
faced by charitable organizations in their efforts to attract H3b: The relationship between self-esteem and goal
and solicit quality help from youth volunteers. setting is the mediated by attitude towards the task.
The present study attempts to bridge this gap by Questionnaires were administered by mail to a na-
examining the determinants of helping behavior in a tional sample of youth who had received marketing mate-
sample of youth volunteers involved in a task requiring rials asking them to participate in a fundraising effort for
significant self sacrifice. We hypothesize that attitudes an international relief organization that involved fasting –
toward this task will mediate the effect of other individual i.e., not eating solid food for 30 hours. The respondents’
variables on the quality of the helping behavior received attitudes towards the organization and towards the
by the NPO. Based on the conceptualization of helping fundraising task were measured by instructing the respon-
behavior forwarded by Bendapudi, Singh, and Bendapudi dent to “Please mark an “X” on the space that best
(1996), we consider two distinct dimensions of helping describes your honest attitudes or feelings about each of
behavior – help vs. no help and token help vs. serious help. the following.” The fasting task and the name of the
The first of these dimensions is captured by the variable organization were each followed by a five-point scale
participation, which indicates whether or not the respon- anchored by “negative” and “positive.” Materialism
dent chose to take part in the fast; and the second of these (MAT) and Self-Esteem (SES) were both measured using
dimensions is captured by the variable goal-setting, which established scales. In order to test our hypotheses we
indicates whether or not the individual set a fundraising followed the procedure outlined by Barron and Kenney
goal. We examine three primary independent variables of (1986) for testing mediation. This test entails three steps:
interest that have been theorized to play key roles in (1) regress the independent variable on the mediator, (2)
determining volunteer behavior and that should be espe- regress the independent variable on the dependent vari-
cially relevant for better understanding teens’ charitable able, and (3) regress both the independent and mediator on
participation: (1) their attitudes towards the organization, the dependent variable. Four of the six anticipated rela-
(2) their level of self-esteem, and (3) their level of mate- tionships were supported: H1a, H1b, H2a, and H3b. A
rialism. Our hypotheses are as follows: task was consistently a significant predictor of both help-
ing behavior dependent variables.
H1a: The relationship between attitude towards the or-
ganization and the decision to volunteer is medi- Our findings indicate that for NPOs marketing to
ated by attitude towards the task. youth for involvement in volunteer projects, even if teens
think highly of an organization, the choice of the task is
H1b: The relationship between attitude towards the or- crucial to their participation and their commitment to
ganization and goal setting is the mediated by fundraising. For NPOs recruiting teens that are likely to be
attitude towards the task. higher in materialism, the nature of the task or activity

needs to be interesting and compelling enough to get them relationship between helping behavior quality and self-
to join in. However, once they have joined the activity, esteem becomes more pronounced upon the decision to
attitudes toward the fundraising task do not appear to be participate in a volunteer task, but in our context, self-
enough for them to follow-through with actual fundraising esteem was not predictive of the initial decision to volun-
activities. Finally, our results seem to imply that the teer.

Elten Briggs
University of Oklahoma
307 West Brooks, Room 10A
Norman, OK 73019
Phone: 405.325.4675
FAX: 405.325.7688
E-Mail: ebriggs@ou.edu

A CROSS-CULTURAL EXAMINATION OF THE RELATIONSHIP
BETWEEN MATERIALISM AND INDIVIDUAL VALUES
William Kilbourne, Clemson University, Clemson
Marko Grünhagen, Southern Illinois University at Edwardsville, Edwardsville
Janice Foley, University of Regina, Regina
SUMMARY incentives were given. The final sample was 46 percent

female and 54 percent male.
Interest in materialism and its implications is appar-
ent as far back as the early Greek philosophers. Pythago- Following Roberts, Manolis, and Tanner (2003),
ras, for example, required that students relinquish their materialism is represented as a second order factor model
personal possessions before entering his school. Rudmin with success, centrality, and happiness as first order latent
and Kilbourne (1996) provide an historical view of differ- constructs. Second, the model suggests that self-transcen-
ent attitudes toward materialism from the ancient world to dence values will be inversely related to materialism, and
the modern. Marketing scholars recently returned to an self-enhancement values will be positively related to
assessment of materialism in contemporary societies. materialism. The model will be tested across three cul-
While some studies have remained critical of consump- tures, Germany, Canada, and the U.S., using the proce-
tion practices, referring to materialism as a “dark side” dures suggested by Steenkamp and Baumgartner (1998).
variable (Mick 1996), others have suggested that materi-
alism might have positive aspects as well (Holt 1997). Both the measurement model and the structural mod-
Despite the many examinations of materialism, it remains el were tested for invariance across the samples. In these
unclear how it relates to other aspects of life. tests, five statistics were used with standard cutoff criteria
as suggested by Hair et al. (1998). In addition, successive-
Burroughs and Rindfleisch (2002) argue that the ly constrained models were subjected to sequential Chi
nature of materialism needs to be clarified, and that more square difference tests to determine if the constrained
researchers are beginning to examine it in the context of model fit the data as well as the baseline model.
other life goals and values. While there have been many
definitions of materialism, what they have in common is The analysis required that the structural model pro-
that they reflect the use of consumption to acquire more posed be tested. This was done as suggested by Byrne
than instrumental or use value in the things purchased. (1999), maintaining the invariances established in the
Collectively, the definitions suggest that the individual measurement models in the initial structural model. Be-
seeks a relationship with objects through which she or he cause the measurement model for materialism was invari-
is enhanced in some way. ant, it was included as a fully constrained measurement
model across the three countries. The values measurement
Moschis and Chruchill (1978) and Richins (1987) model was fully invariant except for the covariances that
have measured materialism as an attitude structure focus- were partially invariant with the German sample left free
ing on the meaning of possessions to the individual. The to vary. The final structural model was tested first for
model developed here depicts materialism as an attitude overall fit across countries and satisfied all criteria used.
structure that is influenced by the individual values of In addition, the sequential Chi Square Difference Test
self-enhancement and self-transcendence as used by yielded a p-value of 0.74 indicating that the constrained
Schwartz (1994). model predicted as well as the original model. We con-
cluded that the structural model was invariant across
Methods countries as predicted.
The sample for the study consisted of university The path coefficients for the structural model were
students from Canada, Germany and the U.S. There were consistent with the hypotheses established. The coeffi-
168, 139, and 97 respondents respectively from the three cient from self-transcendence to materialism was -0.21
countries. Canadian students completed and returned the (p < 0.01) indicating the negative relationship hypothe-
questionnaire in class, and U.S. and German students sized. The result was consistent across the three countries.
completed the questionnaire outside of class and returned The coefficient from self-enhancement to materialism
it the next week. All participation was voluntary and no was 0.70 (p < 0.01) demonstrating the positive relation-
ship hypothesized. This result also held across the three
samples.

Conclusions that the materialism scale is a measure of attitudes toward
consumption rather than a true value labeled as material-
There are two main conclusions to be drawn from the ism. The hypothesized causal sequence was individual
study. The results of the analysis indicate that the reduced values having an effect on attitudes toward materialism.
form of the materialism scale produces a valid second The results indicated that the proposed causal model had
order factor model of materialism. The latent constructs an acceptable fit on all criteria. In addition to this, the
are happiness, centrality, and success and the model was causal model was demonstrated to be invariant across all
shown to be invariant across countries. three countries. The relationship between self-transcen-
dent values and materialism was negative, and the rela-
The second objective or the study was to determine tionship between self-enhancement values and material-
the causal sequence for values and materialism assuming ism was positive as hypothesized.

William Kilbourne
Clemson University
Clemson, SC 29634
Phone: 864.656.5296
FAX: 864.656.0138
E-Mail: kilbour@clemson.edu

M.E.A.L. TIME: HOW NUTRITIONAL DISCLOSURE AFFECTS GENDER
EVALUATIONS OF FAST FOOD MENU ITEMS
Kenneth W. Bates, University of Arkansas, Fayetteville
Kyle A. Huggins, University of Arkansas, Fayetteville
SUMMARY show that consumers may be able to distinguish differ-

ences between relatively healthy and non-healthy items,
Obesity has become a major health concern for many but they are generally unaware of the large degree to
citizens in the United States, and it is an issue, that which many items are unhealthy. This, in conjunction
according to the CDC, has reached epidemic proportion with the decline in purchase intention for unhealthy items
(Manson and Bassuk 2003). It has been projected by the when nutrition information is provided, suggests benefits
United States Department of Health and Human Services from nutritional disclosure in dinner house restaurants.
that obesity will soon overtake tobacco as the leading
cause of preventable death in the United States and will While these are important findings for dinner house
offset many of the medical advancements made in dis- restaurants, we believe the fast food industry is an equally
eases such as cancer, diabetes, and heart disease (U.S. important arena for exploration. The National Restaurant
DHHS 2001; Manson and Bassuk 2003). Due to these Association projects that $123.9 billion will be spent on
recent developments many policy makers, public-interest quick service dining in 2004 (National Restaurant Asso-
groups, and concerned citizens have turned their focus ciation 2004). While Burton and his colleagues (2004)
upon the restaurant industry. In the last thirty years, the assessed the effect of nutrition disclosure on purchase
amount of money spent on food purchases outside-the- intention and choice, we propose that significant gender
home has risen 20 percent and accounts for almost half of differences will be found within fast food item evalua-
American’s total yearly expenditures on food (Lin, Frazão, tions and choice.
and Guthrie 1999). The correlation between increases in
restaurant patronage and obesity has created concern There has been a substantial amount of research
about the healthiness of restaurant menu items, and propo- stressing the importance of physical appearance and at-
sitions have been made that Congress should extend the tractiveness, especially the importance that it holds for
Nutrition Labeling and Education Act (NLEA) to include females (e.g., Cash, Winstead, and Janada 1986; Bar-Tal
restaurant menu items. The Menu Education and Labeling and Saxe 1976; Richins 1991). According to Hirschman
Act (MEAL) of 2003 in Congress requires restaurant (1987), women’s personal value is much more likely to be
chains with twenty or more outlets to list key nutrition determined by physical attractiveness while males will be
information (DeLauro 2003), and the FDA has initiated evaluated on a wider range of attributes. In 1992, Myers
preliminary discussions about possible national standards and Biocca found that watching only 30 minutes of
for the provision of nutrition information for restaurant television programming and advertising can alter a young
foods (Mathews and Leung 2003). Any ruling that man- woman’s perception of her body. Empirical studies have
dates the provision of nutrition information at restaurants also found that women are more concerned about their
will have profound financial, tactical, and strategic impli- physical appearance than men (Burton, Netemeyer, and
cations for retailers in the restaurant industry. Lichtenstein 1995); young women frequently compare
themselves with idealized advertisement models, and the
Given the importance of this issue to public policy amount of comparison is negatively correlated with satis-
makers and restaurant industry managers, our research faction of one’s physical appearance (Richins 1991).
extends recent findings related to nutrition disclosure in Therefore, in light of this stream of literature, it seems
table service restaurants to the fast food industry. Burton appropriate to investigate gender effects within the pa-
et al. (2004) found that consumers significantly underes- tronage of quick service restaurants.
timate the nutritional content of items that are higher in
calories, fat, and sodium while estimations of relatively Overall, fast food results model that of dinner house
healthier items were far closer to actual nutrient levels. restaurants as respondents’ estimations of calorie, fat, and
When item nutrition levels were actually provided, re- sodium levels in fast food items are generally inconsistent
spondents significantly decreased in their purchase inten- with actual levels. As items become higher in unhealthy
tion for items inconsistent with nutrient estimates (rela- attributes, consumers significantly underestimate objec-
tively unhealthier items) and either showed no effect or tive nutrient levels. Of concern to the MEAL proposition,
increased purchase intention for items consistent with consumers are also unable to predict, with any increased
nutrition estimates (relatively healthier items). Results accuracy, the levels of fat or sodium in an item when

caloric content alone is provided. While consumers may condition (provided information on calories, fat, and
know that higher calories are positively related to higher sodium), males and females both significantly decrease
saturated fat and sodium levels, they do not appear to their frequency of choosing these items. Although males
make accurate predictions of these levels from calories still choose the unhealthy items 47 percent of the time
alone. compared to women at only 5 percent, this provides
empirical support for requiring nutrition disclosure. Due
Findings also indicate that males generally visit fast to the preference exhibited by males for unhealthier items
food establishments more often than females, have more and the significant effect of nutrition condition on item
positive attitude towards unhealthy items, and choose content favorability and purchase intentions, the level of
them more frequently. As expected, when no nutrition nutritional disclosure could potentially have a large im-
information is provided, males select unhealthy items pact upon the male gender. References available upon
nearly 70 percent of the time as opposed to women at only request.
a quarter of the time. However, under the full nutrition

Kenneth W. Bates
Marketing and Logistics Department
Sam M. Walton College of Business, 302 BADM
University of Arkansas
Fayetteville, AR 72701
Phone: 479.575.4055
FAX: 479.575.8407
E-Mail: kbates@walton.uark.edu

UNLOCKING VALUE THROUGH CUSTOMER EDUCATION
Thorsten Hennig-Thurau, Bauhaus-University of Weimar, Germany
Peter C. Honebein, University of Nevada, Reno
Benoit Aubert, Grenoble Ecole de Management, France
SUMMARY the product’s consumption and, ultimately, its disposal.

The overall customer value of a product is the totality of
Vargo and Lusch (2004) argue that it is the customer values realized across these different stages.
who eventually determines the value of a product by using
it, not the producer by manufacturing and distributing it. Second, the customer value delivered by a product
Specifically, the customer “must learn to use, maintain, over the different phases is the result of a joint effort of the
repair, and adapt the appliance to his or her unique needs, producer and the consumer. Both parties’ actions and
usage situation, and behaviors” (Vargo and Lusch 2004, competence are indispensable for generating customer
p. 11). The question we address in this research is derived value in each phase. In the joint value creation framework,
directly from this shifting of the focus of value production manifest products carry the potential for a certain level of
from the producer to the customer: How can the producer customer value, with this potential being innovated and
ensure that there is not only value potential embedded in implemented by the producer in its laboratories and facto-
its goods, but also that this value potential will be un- ries, respectively. The level of value production can be
locked appropriately by the customer during product increased by an active role of the customer in these phases.
usage?
Customer Education Model
In this research, by drawing upon the literature of
customer-as-coproducer from service research, we present As illustrated in the joint value creation framework,
the idea of a joint value creation process between produc- customers hold the key to unlocking post-purchase value
ers and consumers for the context of consumer goods. We through the preparation, use, and disposition of products.
then illustrate how the use of customer education, referred To aid the customer, companies offer supporting activi-
to as a company’s measures and actions that are targeted ties, such as customer education, which ensure customers
at the development of post-purchase related customer have the necessary skills to perform key value-driven
knowledge and skills, can impact key company outcomes, tasks. We offer a conceptual model that predicts the
such as customer satisfaction, trust, and loyalty. effects of unlocking value through customer education on
product satisfaction, company satisfaction, customer trust
Joint Value Creation in the firm, and customer loyalty (see Figure).
Organizations should endeavor to increase the com- From a conceptual standpoint, our research suggests
petence of customers that is needed by the customer to that customer education can be a catalyst for increasing
stimulate full use of a product during consumption. We customer loyalty. Through the acquisition of skills, we
suggest a joint value creation framework should serve as expect customers to use products more intensely, which
the theoretical basis for the customer education concept leads to greater product performance and increased prod-
since it demonstrates how customer value is created. uct satisfaction. Additionally, we believe customer edu-
cation also influences the customer’s satisfaction and trust
The framework is founded on two basic axioms. with the company. The combination of customer satisfac-
First, it is presumed that the generation of customer value tion, company satisfaction, and company trust is what
represents a multi-stage process, which starts with the ultimately leads to customer loyalty. Future research
generation of the product idea, continues with the devel- should examine the link between customer education and
opment and production of the product, and finalizes with customer loyalty.

FIGURE

Thorsten Hennig-Thurau
Department of Marketing and Media Research
Bauhaus-University of Weimar
Bauhausstr. 11
99423 Weimar
Germany
Phone: +49.3643.58.3822
FAX: +49.3643.58.3791
E-Mail: tht@medien.uni-weimar.de

ORGANIZATIONAL LEARNING AND DYNAMIC MARKETING
CAPABILITIES: IMPLICATIONS FOR ORGANIZATIONAL
PERFORMANCE
Linda M. Foley, University of Mississippi, University
Douglas W. Vorhies, University of Mississippi, University
Victoria D. Bush, University of Mississippi, University
SUMMARY to both academics and practitioners. For example, why are

some firms better at adapting to change than others? What
Increasingly, the dynamic capabilities view is being business processes and routines must be in place to allow
utilized as a favored approach to understanding a firm’s firms to adequately adjust to modifications in consumer
competitive advantage. The dynamic capabilities view preferences, economic, or other environmental condi-
builds on the previous resource based view of the firm by tions? How do firms generate continual and successful
supplementing its strengths while simultaneously recog- innovative thinking? How does this internal reconfigura-
nizing its weaknesses (Priem and Butler 2001; Eisenhardt tion to adapt to external changes affect overall business
and Martin 2000; Teece, Pisano, and Schuen 1997). This performance, from both a perspective of financial out-
emerging capabilities literature realizes that the posses- comes and positions of advantage? And, finally, how do
sion of rare, imperfectly imitable, and valuable resources specific marketing functions such as customer relation-
is an important step to build competitive advantage. ship and brand management play into outcomes like
However, dynamic capabilities also address the develop- customer satisfaction and brand equity from a dynamic
ment and deployment of these resources and how resourc- capabilities perspective?
es are integrated together within a firm (Barney, Wright,
and Ketchen 2001; Day 1994). The dynamic capabilities This paper attempts to answer the above questions by
“view recognizes that the deployment of resources through integrating dynamic capabilities into a single unified
these organizational processes may better explain firm framework. In order to do so, we classify and subdivide
performance variations than absolute resource levels in the various levels of organizational routines and process-
driving firm performance” (Ray, Barney, and Muhanna es into lower and higher order capabilities constructs.
2004). Since the addition of the word “dynamic” to the term
“dynamic capabilities” implies learning through recon-
Due to its inherent advantage, researchers have stud- figuration, development, and integration, it was necessary
ied this important concept (e.g., Teece, Pisano, and Schuen to integrate the organizational learning literature into the
1997; Day 1994). However, despite all the recent atten- framework presented here.
tion, little research has been conducted with specific
regard to dynamic capabilities and marketing. Since mar- This paper contributes to academia by tying market-
keting processes are constantly undergoing adaptations ing capabilities to the financial performance of the firm,
that parallel the changes in consumer preferences and the which is a topic which has had numerous calls for research
environment as a whole, it is extremely important to (e.g., Srivastiva, Shervani, and Fahey 1998) including a
understand marketing processes in a dynamic context. recent edition of the Journal of Marketing devoted to
understanding this relationship. Additionally, the frame-
A better understanding of marketing capabilities may work was also developed based on qualitative interviews
lead to the ability to answer some fundamental questions with key informants in marketing organizations.
REFERENCES gerial Representations of Competitive Advantage,”

Journal of Marketing, 58 (April), 31–44.
Barney, Jay, Mike Wright, and David J. Ketchen, Jr. Eisenhardt, Kathleen M. and Jeffrey A. Martin (2000),
(2001), “The Resource-Based View of the Firm: Ten “Dynamic Capabilities: What Are They?” Strategic
Years After 1991,” Journal of Management, 27, Management Journal, 21, 1105–21.
625–41. Priem, Richard L. and John E. Butler (2000), “Is the
Day, George S. and Prakash Nedungadi (1994), “Mana- Resource-Based ‘View’ a Useful Perspective for

Strategic Management Research?” Academy of Man- Srivastava, Rajendra K., Tasadduq A. Shervani, and Liam
agement Review, 26 (1), 22–40. Fahey (1998), “Market-Based Assets and Sharehold-
Ray, Gautam, Jay B. Barney, and Waleed A. Muhanna er Value: A Framework for Analysis,” Journal of
(2004), “Capabilities, Business Processes, and Com- Marketing, 62 (1), 2–18.
petitive Advantage: Choosing the Dependent Vari- Teece, David J., Gary Pisano, and Amy Shuen (1997),
able in Empirical Tests of the Resource-Based View,” “Dynamic Capabilities and Strategic Management,”
Strategic Management Journal, 25 (1), 23–38. Strategic Management Journal, 18 (7), 509–33.

Linda M. Foley
School of Business Administration
University of Mississippi
University, MS 38677
Phone: 412.401.2662
FAX: 412.481.3160
E-Mail: LHorton945@yahoo.com

ADOPTING RFID TECHNOLOGY: DOES THE MANAGER’S
ATTITUDE MATTER?
Gilbert N. Nyaga, Michigan State University, East Lansing
Roger J. Calantone, Michigan State University, East Lansing
Thomas J. Page, Michigan State University, East Lansing
ABSTRACT to address this gap by underscoring the importance of

managers’ attitude toward RFID adoption.
This paper explores the mediating role of manager’s
attitude, subjective norms, and perceived behavioral con- LITERATURE REVIEW
trol in the adoption of RFID technology. Moderating
variables are also examined. Strategic and theoretical Radio-Frequency Identification (RFID) Technology
implications are discussed. The examined antecedents are
posited to have significant influence on RFID adoption. RFID consists of a passive radio-frequency tag with
a printed antenna and a radio-frequency (RF) emitter/
INTRODUCTION reader. The tag emits a signal using energy derived from
the RF emitter/reader. The tag stores detailed product
Advances in technology have made the need for information and is attached to a product or pallet, with
companies to implement new systems that improve busi- each tag specifying a unique product identification code
ness process performance a continuous imperative. How- (Rutner et al. 2004). Currently the RFID signal range is a
ever, decision makers in organizations are not always few meters, but with continued advances in technology,
willing to implement or adopt new technologies or to do the range is likely to increase tremendously. Use of RFID
so early. This unwillingness maybe due to shortage of enables more accurate inventory management, fewer stock-
funds, lack of appreciation of the technology’s potential, outs, and less product loss throughout the supply chain.
or even outright negative attitude toward new technology. As a result, most retailers including Wal-Mart, Target,
This study seeks to understand how managers’ attitude and Albertsons as well as the Department of Defense have
toward a new technology, radio-frequency identification imposed deadlines for their suppliers to begin shipping
(RFID), influences their decisions to implement, delay, or RFID-tagged pallets and cases (Bednarz 2004).
reject its adoption. RFID is the latest technology byword
in supply chain management. Manufacturers, wholesal- RFID can be deployed in anything that can be tagged,
ers, retailers, and third party firms are seeking to leverage which means that RFID can be used in very diverse
the advantages brought about by the RFID technology. industries and data management systems. For instance, in
Yet, there are marked differences in the deployment of the the tire industry, RFID chips that are smaller than the size
technology among firms, especially with respect to the of a rice grain can be embedded into new tires, which
attitude of managers mandated to evaluate and implement enables association of the tire with a specific vehicle, store
RFID adoption. a 12-character coding for a number required by the U.S.
Department of Transportation, and indicate when and
There exists a significant body of research on the where the tire was manufactured (Faber 2002). Both big
reasons why people accept or reject new technological and small firms are implementing RFID technology. For
systems (Davis et al. 1989). Most studies are premised on example, a fairly small company, Beaver Street Fisheries
theories and models borrowed from social psychology is using RFID to track cases and pallets of fish and more
(Davis 1986; Swanson 1987). In this study, the theory of exotic fare-including alligator and turtle meat at the com-
reasoned action (TRA) and its two variants, the theory of pany’s seafood-processing and packing plant (Sullivan
planned behavior (TPB) and the technology adoption 2004).
model (TAM) are used. Extant literature shows that stip-
ulations of these theories have been applied, directly or In spite of the demonstrable benefits, some managers
indirectly, in examining the impact of an individual’s are less than enthusiastic about implementing RFID tech-
attitude on behavioral intention, and subsequently ob- nology. One of the reasons is that it is a costly investment.
served behavior (Eagly and Chaiken 1993). Most studies Generally, the cost of implementing RFID ranges from
on technology adoption tend to focus on end users (Davis $100,000 to several millions (Sullivan 2004). A market
et al. 1989) while ignoring the role of manager’s involved study by The Aberdeen Group shows that most of the
in implementation of those technologies. This study seeks polled company officials believe that deploying RFID

will add costs and significantly erode their profit margins individual’s belief about the extent of difficulty or ease of
(Cooke 2002). Other concerns include lack of RFID performing the behavior (i.e., implementing RFID). It is
standardization and security/privacy issues. Managers are a measure of a person’s perception of his or her control
concerned that some RFID deployments may not be over factors necessary to carry out the behavior (Hill et al.
compatible with those of supply chain partners. Indeed, 1996), and is expected to reflect past experiences as well
big firms such as Kimberly-Clark, Target, and Wal-Mart as anticipated impediments and obstacles (Doll and Ajzen
have been pressing for adoption of standards on what is 1992). Examining managers’ perceived behavioral con-
included in RFID chip and how readers and tags commu- trol as an antecedent of intention to adopt new technology
nicate (Bacheldor and Sullivan 2004). can help better understand RFID adoption.
Theories TAM is an adaptation of TRA, which aims at provid-

ing a basis for tracing the impact of external factors on
The theory of reasoned action (TRA) (Ajzen and internal beliefs, attitudes, and intentions (Davis et al.
Fishbein 1980) is a very general model that has been 1989) and is specifically tailored for modeling user accep-
subjected to several adaptations. These include Davis’ tance of information systems (Phillips et al. 1994). Ac-
(1986) technology adoption model (TAM) and the theory cording to TAM, the key determinants of acceptance
of planned behavior (TPB) (Ajzen 1980). These theories behavior are perceived usefulness and perceived ease of
have been shown to provide a robust analysis of the use. Perceived usefulness refers to the prospective user’s
attitude-behavior intention relationship. In this study, subjective probability that using a specific application
antecedents of attitude toward technology adoption pos- system will increase his/her job performance in an orga-
tulated in TAM, and antecedents of behavioral intention nizational context. Perceived ease of use (EOU) reflects
posited in TRA and TPB are integrated to develop a model the potential difficulties in transferring and utilizing the
of RFID adoption. new technology both by the adopting firm and individuals
expected to use it (Phillips et al. 1994). TAM stipulates
TRA stipulates that an individual’s performance of a that people form intentions to perform behaviors that they
behavior is determined by his or her behavioral intention have positive affect. For instance, people form intentions
(BI) to perform the behavior, and that BI is jointly deter- toward using computer systems based largely on their
mined by the person’s attitude and subjective norms cognitive appraisal of how the computer system will
concerning the behavior in question (Ajzen and Fishbein improve their performance. TAM offers a base for exam-
1980; Davis et al. 1989). BI is a measure of strength of ining antecedents of managers’ attitude toward deploying
ones intentions to perform a specified behavior while RFID.
attitude refers to the degree to which a person has favor-
able or unfavorable evaluation of the behavior in ques- CONCEPTUAL MODEL AND HYPOTHESES
tion. Subjective norm (SN) is a person’s perception that
people who are important to him/her think that he/she The preceding literature review highlights several
should or should not perform the behavior (Ajzen and variables that have a bearing on a manager’s attitude and
Fishbein 1980; Hill et al. 1996). Extant literature on behavioral intention with respect to implementing RFID
adoption identifies several salient referents including top in their organizations. Each of the three models, TRA,
management, supervisors, the organization’s MIS depart- TPB, and TAM, provides some bases for understanding
ment, technology experts, and peers. Attitude toward a technology adoption. Thus, this study integrates stipula-
behavior is determined by an individual’s salient beliefs tions of the three models to develop the model shown in
about consequences of performing the behavior (i.e., Figure 1.
adopting RFID technology) multiplied by the subjective
values or evaluations of those consequences (i.e., impact As argued in TAM, perceived usefulness and per-
of RFID technology). Beliefs refer to an individual’s ceived ease of use are stipulated to influence managers’
subjective probability that performing the target behav- attitude toward RFID technology adoption (Davis et al.
ior, in this case deploying the RFID technology, will result 1989; Adams et al. 1992; Phillips et al. 1994). In the case
in a given consequence. TRA has been applied widely and of RFID adoption, perceived usefulness refers to per-
provides a useful framework for this study. ceived RFID utilities for the organization. It is conceptu-
alized in terms of costs and benefits of RFID technology
TPB is an extension of TRA, which seeks to address compared to current technology (i.e., technology in use
limitations in TRA with respect to behaviors over which that is to be replaced). For instance, if a firm has bar coding
people have incomplete volitional control (Ajzen 1985; technology in place and wishes to adopt RFID, the man-
Doll and Ajzen 1992). TPB stipulates that besides attitude ager will compare the costs of implementing RFID (actu-
and subjective norms (as stipulated in TRA), another al, purchase cost, installation costs, employee training
factor, perceived behavior control is predictive of behav- costs, etc.) and expected benefits (i.e., asset management,
ioral intentions. Perceived behavioral control refers to an error reduction, supply chain collaboration, etc.) against

FIGURE 1
Conceptual Model
Perceived usefulness
H1 Manager’s attitude
of RFID
toward adopting RFID H4
Perceived ease of use Behavioral intention
H2 Manager’s subjective
and implementation H5 about adopting
of RFID norms toward adopting RFID
RFID
H6
Normative beliefs and H3
motivation to comply Manager’s perceived
with referent behavioral control on
adopting RFID H7a,b,c
Moderators
♦ Manager’s experience
♦ Firm size
the costs and benefits of bar coding. If the gap between Managers’ normative beliefs and motivation to com-
these is significantly large (i.e., RFID is more effective ply with referents are stipulated to influence subjective
than bar coding), then the manager can justifiably seek norms and subsequently behavior intention (Ajzen and
RFID implementation. Whether a firm adopts RFID tech- Fishbein 1980; Davis et al. 1989). Referents expected to
nology or not is likely to be influenced by the manager’s exert influence on managers adopting RFID include sup-
cognitive evaluation of the gains from the implementa- ply chain partners, senior managers, consultants, etc. For
tion. Thus, it is hypothesize that: instance, recommendations by highly reputed consultants
as well as advice by experienced executives held in high
H1: Perceived usefulness has a direct and positive effect esteem by a manager may motivate the manager to under-
on the manager’s attitude toward RFID adoption. take a specific business function, in this case, deploying
RFID technology. Furthermore, if supply chain partners
Perceived ease of use is conceptualized in terms of and competitors are deploying RFID, then managers will
extent of difficulty, time, and learning costs associated most likely follow suit. Thus, it is hypothesized that:
with changing from the current system to RFID technol-
ogy. If RFID deployment is perceived to be cumbersome H3: Manager’s normative beliefs and motivation to com-
and difficult, or inconveniencing to supply chain partners, ply with a referent have direct effects on his/her
then it is expected that managers might be hesitant to subjective norms toward RFID adoption.
implement it. However, if the deployment is perceived to
be less involving, then managers might be inclined to The manager’s attitude toward implementation of
readily implement it. Moreover, the cost of training em- RFID is posited to be a major determinant of whether or
ployees on how to use RFID can be high and time not he/she executes RFID deployment. Attitudes develop
consuming. Learning costs can also include the negative from beliefs that people hold about the object of the
performance implications of technology changeover and attitude (Doll and Ajzen 1992). It is expected that if
potential losses in the process of replacing the current managers strongly believe that implementing RFID in
system with RFID. These costs could influence manag- their organization will culminate in improved process
ers’ attitude toward RFID adoption. Thus, it is hypothe- performance and better coordination with supply chain
sized that: partners, they are likely to develop favorable attitudes
toward RFID. However, if their RFID evaluation is neg-
H2: Perceived ease of use has a direct and positive effect ative, then we expect that their attitude toward RFID
on the manager’s attitude toward adopting RFID. implementation will be negative. This is expected to

impact on their behavioral intentions and subsequently experience with a practice differentially affects their atti-
decisions to either implement or not implement the RFID. tudinal and normative beliefs, which in turn influences
Thus, it is hypothesized that: their intention to continue or introduce a practice. Doll and
Ajzen (1992) concluded that experience with behavior
H4: Manager’s attitude toward RFID technology has a increased the overall predictiveness of attitude, subjective
direct and positive effect on intentions to adopt RFID. norm, and perceived behavioral control on behavior in-
tention.
Subjective norms have been shown to influence be-
havioral intention (Ajzen and Fishbein 1980; Hill et al. Firm size has also been stipulated as a moderator. In
1996; Karahanna et al. 1999). In the case of RFID adop- their international study of attitudes toward adoption,
tion, it is expected that manager’s subjective norms as Bunker and McGregor (2002) suggest that contextual
determined by normative beliefs and motivation to com- variables such as firm size and business type are the most
ply with referents will have a strong influence on technol- influential factors. Indeed, they report that skill-based
ogy deployment. These norms may arise from available (experience) factors have lesser influence on adoption
information (i.e., trade journals, business magazines, con- than contextual factors (i.e., firm size, business type).
sultant reports, etc.) or desire to conform to expectations Sillence et al. (1998) also reported size of business as an
of senior managers, experts, and/or supply chain partners. influential factor in adopting e-commerce. In small firms
Thus, it is hypothesized that: faced with resource constraints, RFID implementation
may not be the highest priority project. However, in large
H5: Manager’s subjective norm toward RFID has a direct firms with greater resource endowment, RFID implemen-
and positive effect on intention to adopt RFID. tation may be easily accomplished. Moreover, the influ-
ence of trading partners may be more intense (i.e., Wal-
Perceived behavior control, according to Ajzen Mart’s requirement that its top 100 suppliers adopt RFID).
(1991), can influence whether people choose to pursue an
outcome, their degree of preparation, the effort they From these studies, it can be argued that managers’
expend, their perseverance, as well as the thoughts and past experience in implementing new technology and firm
emotions experienced during the task. Thus, the extent to size will increase the predictive power of antecedents of
which manager’s view themselves as capable of imple- behavioral intention (attitude, subjective norm, and per-
menting RFID, both in terms of mandate to oversee RFID ceived behavior control). In other words, the three ante-
implementation and possession of requisite skills, will cedents will more strongly predict intention to implement
influence their behavioral intention. Moreover, as Doll RFID among managers with experience in implementing
and Ajzen (1992) state, beliefs dealing with presence or new technologies than those with no prior experience, and
absence of required resources and opportunities to per- in bigger firms than in small firms. Thus, it is hypothesized
form a specific behavior are perceived to dictate per- that:
ceived behavioral control. It is expected that the more
resources and opportunities a manager has, the greater H7: The influence of (i) attitude, (ii) subjective norms,
should be their perceived behavioral control over RFID and (iii) perceived behavioral control on intention to
adoption. Consequently, greater perceived behavioral adopt RFID technology increases with:
control is expected to lead to greater intention to adopt
RFID. Thus, it is hypothesized that: a. Increases in manager’s experience in new tech-
nology adoption
H6: Manager’s perceived behavioral control on RFID
deployment has a direct effect on intention to adopt b. Increases in firm size
RFID.
METHOD
Moderators of Attitude-Behavior Relationship
To test the proposed model, data can be collected
Past studies employing TPB have shown that situa- using a survey questionnaire directed to senior logistics
tional and personal factors moderate the relationship officials involved in the new technology implementation
between attitude, subjective norm, perceived behavioral process. The study sample can be selected from member
control, and behavior intention (Doll and Ajzen 1992; Hill list of appropriate industry associations such as the Coun-
et al. 1996; Ajzen 2001). For instance, in a study of cil of Logistics Management. Mail surveys can be effec-
competitive benchmarking intentions among managers, tive in collecting required data. To measure the constructs
Hill et al. (1996) found out that attitude has stronger effect stipulated in the study, tested and validated measures in
for more experienced managers than less experienced extant literature could be adapted. For perceived useful-
managers. They posit that managers’ experience or lack of ness and perceived ease of use of RFID, Davis’ (1989)

technology acceptance scale could be adapted with mod- RFID’s implementation is expected to attract inputs from
ifications to make them relevant to RFID deployment. diverse quarters including industry pundits, supply chain
Measures of attitude and subjective norms could be adapt- partners, managers’ peers, and top echelons of firm’s
ed from TRA scale (Ajzen and Fishbein 1980) while management hierarchy. However, extant literature has
perceived behavioral control construct measures could be not given much attention to this fact. This study therefore
adapted from Ajzen’s (1985) TPB scale. Behavioral in- raises an issue of practical and theoretical import, which
tention construct measures have been developed in all calls for further investigation.
three theories (Ajzen and Fishbein 1980; Ajzen 1985;
Davis 1989). The scales entail asking the respondents to Third, the study implicitly extends the theory of
indicate the extent to which they agree or disagree with reasoned action to supply chain research. In examining an
statements concerning the subject of interest on a seven- issue that has widespread impact on the supply chain, the
point Likert scale anchored with 1 = “Strongly Disagree” study serves to open new frontiers for TRA, TAM, and
and 7 = “Strongly Agree” (Ajzen and Fishbein 1980; TPB applications and research. Future research is likely to
Ajzen 1985; Davis 1989). Measures of moderator vari- add great value by investigating organizational and sup-
ables include information about the respondent such as ply chain issues within the framework of expectancy
number of years as a manager (experience) and company models. Moreover, the study takes into account all three
information such as annual revenue, number of employ- theories, thereby developing a more comprehensive as-
ees, etc (firm size). Firms could also be grouped as small, sessment of attitude-intention linkage within the context
medium, or large in size. of new technology adoption. Such an integrative view of
expectancy models is likely to give better insights to
To analyze the data, structural equation modeling future research.
(SEM) approach could be used. SEM is a comprehensive
statistical approach to testing hypotheses about relations Fourth, the study raises the issue of factors moderat-
among observed and latent variables, and combines as- ing the impact of manager’s attitudes toward new technol-
pects of multiple regression and factor analysis to estimate ogy adoption on behavioral intentions. Firm size and
a series of interrelated dependence relationships simulta- manager’s experience are only a start. There could be
neously (Hair et al. 1998). SEM provides a robust way of other moderating variables that future research can ex-
addressing the relations between constructs and subse- plore. For instance, how does centralization versus decen-
quently enables a better understanding of issues in ques- tralization mediate the attitude-intention relationship? By
tion. understanding the moderating effects of different vari-
ables, researchers and practitioners alike will be able to
STRATEGIC AND THEORETICAL identify conditions in which specific actions are most
IMPLICATIONS appropriate. This in turn will have both managerial and
theoretical implications.
There are several strategic and theoretical implica-
tions from the preceding arguments. First, the study raises CONCLUSION
the issue of manager’s attitude toward RFID application,
and by extension technology adoption for firms engaged Scholars and practitioners alike concur that new
in supply chain relationships. Past studies have tended to technology adoption is a process characterized by many
ignore the role of manager’s attitude, yet, in the final challenges for both the users and implementers. The study
analysis, the manager in charge of new technology de- shows that the role of managers involved in technology
ployment must be convinced that RFID implementation is adoption ought to be given greater consideration when
appropriate, feasible within a given timeframe, and ur- examining technology adoption. In addition, the study
gent. This is an issue of great importance especially with shows that an integrative assessment of propositions from
the expanding deployment of RFID in supply chain oper- various expectancy models greatly help in the under-
ations. By examining the role of managers’ attitude, standing of technology adoption process. Moreover, as
greater insights are likely to emerge that could enable a Davis et al. (1989) suggest, conditions under which atti-
better understanding of not only RFID adoption, but also tudes mediate the belief-intention link ought to be inves-
other new technology adoption situations. This could tigated further. This study is a positive step in responding
result in better planning of marketing efforts. to Davis et al.’s challenge.
Secondly, the study reinforces the role of normative The issues raised in this study have important impli-
beliefs in technology adoption. RFID deployment has cations for academics and researchers. For academicians,
strategic and operational implications on firm’s partners the study raises important questions about the mediating
in the supply chain. Furthermore, being a recent develop- role of manager’s attitude in technology adoption. Indeed,
ment in supply chain collaboration and management, the study’s focus on technology adoption from the point

of view of the manager’s attitude will be of interest to marketing programs. The study has implications on how
consumer behavior researchers. For practitioners, the to design new technology adoption programs and mea-
issues raised offer an opportunity for a better assessment sures to increase usage (i.e., number of companies adopt-
and planning of new technology implementation and ing RFID).
REFERENCES nal, 36 (6), 60–63.

Fishbein, Martin and Icek Ajzen (1975), Belief Attitude
Adams, D.A., R.R. Nelson, and P.A. Todd (1992), “Per- Intention and Behavior: An Introduction to Theory
ceived Usefulness, Ease of Use, and Usage of Infor- and Research. Reading, MA: Addison-Wesley.
mation Technology: A Replication,” MIS Quarterly, Hair, Joseph F., Rolph E. Anderson, Ronald L. Tatham,
16 (2), 227–47. and William C. Black (1998), Multivariate Data
Ajzen, Icek and Martin Fishbein (1980), Understanding Analysis, 5th ed. Upper Saddle River, NJ: Prentice
Attitudes and Predicting Social Behavior. Engle- Hall.
wood Cliffs, NJ: Prentice-Hall. Hill, Matt, Leon Mann, and Alexander J. Wearing (1996),
____________ (1985), “From Intentions to Actions: A “The Effects of Attitude, Subjective Norm, and Self-
Theory of Planned Behavior,” in Action-Control: Efficacy on Intention to Benchmark: A Comparison
From Cognition to Behavior, J. Kuhl and J. Beck- Between Managers with Experience and No Experi-
mann, eds. New York: Springer. ence in Benchmarking,” Journal of Organizational
____________ (1991), “The Theory of Planned Behav- Behavior, 17, 313–27.
ior,” Organization Behavior and Human Decision Karahanna, Elena, Detmar W. Straub, and Norman L.
Process, 50, 179–211 Chervany (1999), “Information Technology Adop-
Bacheldor, Beth and Laurie Sullivan (2004), “RFID Kick- tion Across Time: A Cross-Sectional Comparison of
Start,” InformationWeek, (May 24), 20–22. Pre-Adoption and Post-Adoption Beliefs,” MIS Quar-
Bednarz, Ann (2004), “Manufacturers Ready RFID Roll- terly, 23 (2), 183–213.
outs,” Network World, 21 (16), 40–. Phillips, Lisa A., Roger Calantone, and Ming-Tung Lee
Bunker, D. and R. MacGregor (2002), “The Context of (1994), “Intentional Technology Adoption: Behav-
Information Technology and Electronic Commerce ior Structure, Demand Uncertainty, and Culture,”
Adoption in Small/Medium Enterprises: A Global Journal of Business & Industrial Marketing, 19 (2),
Perspective,” Eighth Americas Conference on Infor- 16–28.
mation Systems, 1675–84. Plouffe, Christopher R., Mark Vandenbosch, and John
Cookie, James A. (2002), “Companies Focusing on Min- Hulland (2001), “Intermediating Technologies and
imal RFID Compliance,” Logistics Management, 43 Multi-Group Adoption: A Comparison of Consumer
(5), 24. and Merchant Adoption Intentions Toward a New
Davis, Fred D. (1986), “A Technology Acceptance Model Electronic Payment System,” Journal of Product
for Empirically Testing New End-User Information Innovation Management, 18, 65–81.
Systems: Theory and Results,” Sloan School of Man- Rutner, Stephen, Mathew A. Waller, and John T. Mentzer
agement, Doctoral dissertation, Massachusetts Insti- (2004), “A Practical Look at RFID,” Supply Chain
tute of Technology. Management Review, 8 (1), 36–41.
____________ (1989), “Perceived Usefulness, Perceived Sheppard, Blair H., Jon Hartwick, and Paul R. Warshaw
Ease of Use, and User Acceptance of Information (1988), “The Theory of Reasoned Action: A Meta-
Technology,” MIS Quarterly, 13 (3), 319–40. Analysis of Past Research with Recommendations
Doll, J. and Icek Ajzen (1992), “Accessibility and Stabil- for Modifications and Future Research,” Journal of
ity of Predictions in the Theory of Planned Behav- Consumer Research, 15, 325–43.
ior,” Journal of Personality and Social Psychology, Sullivan, Laurie (2004), “Reaching Down the Supply
63, 754–65. Chain,” InformationWeek, Issue 981, (March 22),
Eagly, Alice H. and Shelly Chaiken (1993), The Psychol- 49–51.
ogy of Attitudes. Fort Worth, TX: Harcourt Brace. Swanson, E. Burton (1987), “Information Channel Dis-
Faber, Michael J. (2002), “RFID: The Next Tool for position and Use,” Decision Science, 18, 131–45.
Managing Records?” Information Management Jour-

Gilbert N. Nyaga
Phone: 517.353.6381
FAX: 517.432.1112
E-Mail: nyaga@bus.msu.edu

A SYNERGISTIC MODEL FOR NEW PRODUCT SUCCESS
Russell Adams, University of Texas – Pan American, Edinburg
SUMMARY the different aspects of product success into one model it

will be easier to recognize the critical factors of success
For decades researchers have ascribed to Rogers’ and and facilitate the focus of future research on these aspects.
Porter’s contention that new products are one of the The paper also explains the appropriateness of using
strategic keys for a company’s survival (Porter 1980; percolation (Weisbuch and Solomon 2000) as an alternate
Rogers 1962), much research has been conducted in this view of diffusion theory that reflects what occurs in
field which is well summarized by Mahajan, Muller, and consumer markets.
Bass (1990). Companies are continually developing new
products to remain competitive in their markets. Given the The model views product success, measured by dif-
substantial investment involved and the relatively low fusion, as the intersection of two important paths; produc-
success rates of new product introduction, fifty percent of tion, and marketing. From the production side, network
new product introductions fail within the first year (Zirger effects are created by both installed user base and licens-
and Mandique 1990), and given that new products can ing relationships, this then expands the content value of
count for a third of a firms profits (Booz et al. 1982) the hardware leading to greater diffusion. The marketing
research that discovers ways to increase this success rate side involves the human or consumer element. With
is critically important. Several aspects of successful New appropriate marketing and the positive attitude (which
Product Development (NPD) and adoption have been can also be derived from the installed user base), positive
studied independently, such as diffusion, strategic part- WOM will synergistically combine with the network to
nerships, Word-of-Mouth (WOM), and installed user result in product diffusion (percolation). The comprehen-
base as well as the effects of marketing, but these factors sive model is then applied to the current platform wars for
have not been modeled in conjunction. computer gaming. Currently Sony, Microsoft, and Nin-
tendo are vying for dominance in the console game market
The goal of this paper is not to create a definitive with Sony’s Playstation 2 the current leader. The value of
model for guaranteeing the success of new product launch- the model becomes apparent as the factors of success for
es, but to shift the theoretical lens to allow researchers to Sony reflect the synergies that are developed in the model.
pursue avenues that have yet to be studied. By combining References available upon request.

Russell Adams
University of Texas – Brownsville
80 Fort Brown
Brownsville, TX 78520
FAX: 956.983.7558
E-Mail: radams@utb.edu

INCOME ELASTICITY OF HOUSEHOLD’S DEMAND FOR
COMMUNICATION AND IT PRODUCTS: GLOBAL
MEASUREMENT AND ITS MARKETING
IMPLICATIONS
Min Lu, University of British Columbia, Vancouver
Yanbin Tu, University of Connecticut, Storrs
ABSTRACT goods or necessary goods by households, who exhibit

different demand patterns from other categories of con-
This paper investigates the income elasticity of house- sumers such as private firms and government agencies.
hold’s demand for communication and IT (CIT) products We investigate the income elasticity of household’s de-
in twenty-three OECD countries during 1989 to 2001. We mand for communication and IT products in the OECD
find that CIT products are luxury goods to the households countries ranging from 1989 to 2002, and find that com-
in these countries. We discuss the marketing implications munication and IT products are luxury goods to house-
of our findings, and provide some marketing recommen- holds in OECD countries. We also discuss the marketing
dation. implications of our findings in the paper.
INTRODUCTION The rest of the paper is organized as follows: Brief

Literature Review; Hypothesis and Testing Results; A
Business firms engaged in the manufacturing and Discussion of our Findings; and Concludes the paper with
sales of communication and IT products are facing a Limitations of this work.
dynamic and challenging marketplace. In such an envi-
ronment, the capacity to attract and retain consumers, and BRIEF LITERATURE REVIEW
maintain reasonable profit margins, is critical to the busi-
ness success. The idea and practice of differentiating and For a given product, income elasticity and price
segmenting customers are quite common in the technolo- elasticity of demand are expected to change over time as
gy sector. Some communication and IT products vendors the product moves through its life cycle. Early marketing
such as Dell divide their customers into categories of researchers about product life cycle theory including
home, small business, and large business. Other vendors Clifford (1965), Levitt (1965), Cox (1967), Polli and
such as Amazon target the markets in different countries. Cook (1969), and Scheuing (1969) describe in detail the
However, these differentiations and segmentations are four major stages in a product’s life cycle: introduction,
typically based on differences in the type of product or growth, maturity, and decline. New products diffusion
service each segment desires. Theoretically or empirical- models that incorporate price changes have been devel-
ly, we need to offer additional information that can be oped by Robinson and Lakhani (1975), Bass (1980),
used to support and refine their differentiating and seg- Jeuland (1981), and Kalish (1985). From this literature,
menting strategy. One perspective is to measure whether price elasticity is expected to increase as a product moves
one commodity is necessary or luxury to different groups through its life cycle. Two reasons are for this result. First,
of customers. This represents an important research topic, consumers will collect more information about the prod-
because understanding the nature of income demand uct, especially regarding availability, deals, prices. and
elasticity is very useful to developing effective pricing promotions, in the later stages of the life cycle (Tellis
and versioning strategies for communication and IT prod- 1988; Tellis et al. 1988). Second, the “early adopters”
ucts vendors. According to economic theory, the income during the introduction stage tend to have higher incomes
elasticity of demand is used to describe how many per- and are more likely to exhibit lower price elasticities,
centage points the demand for an item will change when accordingly lower income elasticities, than late adopters
income adjusts by one percent. An income elasticity of (Onkvisit et al. 1989; Nagle 1987; Simon 1979). So, we
demand greater than one suggests that the good is luxury can say that price elasticity of demand is likely to be more
to the consumer. An income elasticity of demand between negative in the later stages of the product life cycle.
zero and one implies a necessary good to him, while a
value less than zero suggests that the good is an inferior A better understanding of income elasticity and/or
good to him. price elasticity of demand in combination with the product
life-cycle theory can enable more effective marketing
The objective of this research is to determine whether strategies. For instance, if a product is necessary to some
communication and IT products are considered luxury consumers, they are more likely to be early adopters and

more likely to purchase the product during the introduc- income elasticity of demand is higher. We set up the
tion and growth stages of the product life cycle. In addi- following hypothesis.
tion, these consumers are less price-sensitive so a price
premium can be charged. Luxurious users, on the con- Hypothesis (H): Communication and IT products are
trary, do not tend to have an “urgent” need for the product, luxury goods to households.
and are more likely to buy the product during the growth
or maturation stages. Luxurious users are also more sen- In order to measure the income elasticity of demand,
sitive to prices, that is, lower prices likely to spur demand empirically we perform the following regression: In Q =
and higher prices likely to discourage demand. α + β In M + ε where Q is the quantity of good, M is income
or expenditure and ε is the error term. Taking the deriva-
Economics and marketing scholars have conducted tive with respect to M, we obtain β = (cQ / cM)
some empirical research on the income elasticities of a (M / Q)which is the estimated value of the income elastic-
number of goods and services, such as energy (Bohi ity of demand for Q.
1981), agricultural products (Klonaris and Hallam 2003;
Mdafri and Brorsen 1993) healthcare (Matteo and Matteo We collect the data of annual household final con-
1998; Getzen 2000; Freeman 2003) and information goods sumption expenditure, annual expenditure on communi-
(Bakos and Brynjolfsson 1999, 2000). Lu et al. (2004) cation products and annual expenditure on audio-visual
measure whether computers and packaged software are photographic and information processing equipment
necessary or luxury to private firms, households and (called IT products) which are all at 1995 prices, from
government agencies in USA. They find that such IT 1989 to 2001 in OECD countries (OECD 2002). We
products are necessary goods to private firms, luxury conduct regressions for communication products and IT
goods to households and inferior goods to government products respectively. Table 1 shows the regression re-
agencies. This paper is related to Lu et al.’s research. sults for communication products in twenty-three OECD
However, the difference of this study form theirs is as countries. From the table, we clearly see that the estimated
follows: (1) we investigate communication and IT prod- coefficient for each country is greater than one (from the
ucts, a broader range of commodities, which include highest 8.686 for Japan to the lowest 1.577 for Iceland),
computers and packaged software; (2) we conduct a which suggest communication products are luxury goods
global measurement of income elasticity of communica- to the households in these countries. The t-values for these
tion and IT products across twenty-three countries, which estimated coefficients indicate that they are significant at
provide more robust testing results; (3) this study only one percent level. Most important, the R-squares for most
focuses on the households. of countries (17 out of 23) are over 0.90 (the average R-
square is 0.927), which means that, generally, the regres-
RESEARCH HYPOTHESIS AND sions have a high level of goodness-of-fit.
TESTING RESULT
Table 2 presents the regression results for IT products
The demand for a commodity is said to be normal if in seventeen OECD countries. Except for Germany, all
demand does not fall when income rises, and inferior if the estimated coefficients for other countries are greater
demand falls when income rises. We define the income than 1 (from the highest 6.221 for Demark to the lowest
elasticity of demand for x as η = (cx / cM) where M is the 1.576 for Korean), which also indicate that IT products are
income of consumer. η is interpreted as the percentage by luxury goods to the households in these countries. All the
which the demand for good x changes when income estimated coefficients, except for Germany and Greece,
changes by one percent. A commodity, say x, is a luxury are significant at one percent level. The R-squares for 13
good if η < 1, a necessary good if 0 < η < 1, or an inferior out of 17 countries are over 0.90, and the average R-
good if η < 0. square is 0.906, which mean that the regressions for IT
products have a high level of goodness-of-fit as well.
While some parts of communication and IT products German, the outlier in the sample, has the income elastic-
will be necessary goods to private firms (see Brynjolfsson ity of 0.8141, which suggests IT products are necessary
and Yang 1996; Dewan and Min 1996; Lu et al. 2004), we goods to German households. But, the testing for it is not
expect that overall communication and IT products will be significant below three percent level.
considered luxury goods by households, since the main
function of such products are entertainment, and commu- The testing results for both communication and IT
nication, and rarely critical computing to households. products support the hypothesis (H) in the study, that is,
Following this reasoning process that private firms will that overall communication and IT products are luxury
tend to be early adopters of communication and IT prod- goods to the households in OECD countries. Since most
ucts, as their income elasticity of demand is quite low, we of OECD countries are developed countries, we can
expect that households, on the other hand, will tend to be generalize our conclusion to that communication and IT
later adopters of communication and IT products, as their products are luxury goods to the households globally. The

TABLE 1
Regression Results for Communication Products
Income # of
Country Elasticity t-Value p-Value Observation R-Square
Australia 3.115 21.914 8.77x10-10 12 0.980

Austria 4.095 15.702 2.25 x10-8 12 0.961
Belgium 5.048 11.340 0.000345 6 0.985
Canada 2.075 20.254 4.67x10-10 13 0.974
Czech 5.175 5.09392 0.002235 8 0.812
Demark 2.535 6.360 5.37x10-5 13 0.786
Finland 4.710 4.523 0.001103 12 0.672
France 5.821 36.213 6.13x10-12 12 0.992
Germany 6.063 15.337 3.24x10-7 10 0.967
Greece 4.683 14.423 5.09x10-8 12 0.954
Iceland 1.557 9.731 2.04x10-6 12 0.905
Ireland 2.445 47.216 4.29x10-12 11 0.996
Italy 6.186 17.563 2.14x10-6 13 0.966
Japan 8.686 7.795 2.72x10-5 11 0.871
Korean 4.752 11.785 1.4x10-7 13 0.927
Mexico 3.752 6.767 4.9x10-5 12 0.821
Netherlands 4.723 13.012 0.00098 5 0.983
Norway 3.510 12.548 4.71x10-6 9 0.957
Poland 4.860 8.363 0.003585 5 0.959
Spain 3.128 23.737 1.8x10-5 6 0.993
Sweden 3.301 4.823 0.016986 5 0.886
UK 3.021 26.507 1.35x10-10 12 0.986
USA 2.195 44.806 7.38x10-6 12 0.995
TABLE 2
Regression Results for IT Products
Income # of
Country Elasticity t-Value p-Value Observation R-Square
Austria 3.051 11.049 6.33x10-7 12 0.924

Belgium 2.755 17.151 6.78x10-5 6 0.987
Canada 3.434 34.224 1.59x10-12 13 0.990
Demark 6.221 20.673 3.75x10-10 13 0.975
Finland 3.564 14.974 3.56x10-8 12 0.957
France 5.061 35.897 6.69x10-12 12 0.992
Germany 0.8141 2.598 0.032 10 0.458
Greece 5.354 4.404 0.012 6 0.829
Iceland 2.798 14.058 6.51x10-8 12 0.952
Ireland 2.096 6.422 0.000122 11 0.821
Italy 3.454 10.699 3.75x10-7 13 0.912
Korean 1.576 17.314 2.5x10-9 13 0.965
Mexico 1.602 4.893 0.00063 12 0.705
Norway 2.832 14.297 1.95x10-6 9 0.967
Spain 2.220 41.896 2.99x10-5 5 0.998
UK 4.246 21.250 1.19x10-9 12 0.976
USA 4.655 82.128 1.75x10-15 12 0.999

underlying reason is that if such products are considered household processes a computer. Intuitively, these two
as luxury goods by the households in developed countries, figures further demonstrate that computers are luxury
they must be perceived as luxury goods by the households goods to households in OECD countries.
in developing countries as well since the developing
countries have a low GDP per capital and an associated This study provides some interesting insights into
living standard. communication and IT products markets. We find evi-
dence in the global setting to suggest that communication
DISCUSSION AND IMPLICATIONS and IT products are luxury goods to households. Vendors
OF FINDINGS of communication and IT products can take advantage of
these findings by strategically setting prices for the house-
Why are communication and IT products luxury hold. Vendors should set low prices of communication
goods to the households? Unlike private firms, who use and IT products for households for two reasons. First, as
communication and IT products for the purposes of com- these products are considered to be luxury goods, a high
puting function and competitive advantage, households price will result in a sharp drop in demand.
usually use such products for entertainment, communica-
tion and convenience. It is common to state that game Second, as growth in the household sector is a precur-
software, digital camera, audio or video systems are sor to growth in the overall market, a drop in current
luxury goods to households. For other IT products such as demand will also entail a reduction in future sales (see Lu
computers, households need them to function these “lux- et al. 2004). As Lu et al. (2004) point out that during the
ury” goods. Therefore, computers and associated applica- introductory stage of the product life cycle for computers
tion software are the extension of such luxury goods to the and packaged software the consumer base will tend to be
households. Figure 1 shows the percentage of households comprised predominantly of private firms, and once the
with access to a home computer in OECD countries in growth stage begins, marketing and advertising efforts
2000. The percentage varies from 11.1 for Mexico to 65 should shift to cultivate demand within the household
for Demark. For U.S., 51 percent households use comput- sector. Our findings in the global setting further confirm
er at home in 2000. Figure 2 shows that more and more that vendors of communication and IT products should
people buy and use personal computers in USA during make more marketing efforts towards the household mar-
1997 to 2001, which tells us that even in USA, not every kets when their products evolve into the growth stage. As
FIGURE 1
Households With Access to a Home Computer in 2000
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0
Denmark
Switzerland
Sweden
Korea
United
Canada
Australia
Germany
Finland (1)
Japan
United
Belgium (3)
Portugal
Austria
Ireland
Turkey (4)
Mexico
France
Spain
Italy
Notes: Beginning of 2002; 2March 2001–April 2002 (financial year) instead of 2001; 31999 instead of 2000;
1
4
Households in urban areas only; 5For 1999, households in urban areas with more than 15,000 inhabitants
only.

FIGURE 2
Personal Computers in USA (Per 1000 People)
700
600
500
400
300
200
100
0.0
1997 1998 1999 2000 2001
the life cycle of communication and IT products become some marketing recommendation of using price discrim-
shorter and shorter, this recommendation is significant to ination to households and switching their market target to
the communication and IT industry. households as their products evolve to the growth stage of
product life cycle.
CONCLUSIONS, LIMITATIONS, AND
FUTURE RESEARCH The work presented here is not without limitations.
We use aggregate data in our analytical framework. The
This paper investigates the income elasticity of house- conclusions reached in this paper may not be valid across
hold’s demand for communication and IT products in 23 all types of communication and IT products. Some parts
OECD countries during 1989 to 2001. We find that of communication and IT products, taken cell phone as an
communication and IT products are luxury goods to example, might be necessary goods to the households in
households in these countries, which suggests that such some countries. Therefore, using micro-data to investi-
products are luxury goods to the households globally. We gate these types of nuances is a worthwhile area for future
also discuss the marketing implications of our findings research.
and provide communication and IT products vendors with
REFERENCES Brynjolfsson, E. and S. Yang (1996), “Information Tech-

nology and Productivity: A Review of the Litera-
Bakos, Y. and E. Brynjolfsson (1999), “Bundling Infor- ture,” Advances in Computers, 43, 179–214.
mation Goods: Pricing, Profits, and Efficiency,” Clifford, D.K. (1965), “Leverage in the Product Life
Management Science, 45 (12), 1613–30. Cycle,” Dun’s Review and Modern Industry, 85 (5),
____________ and ____________ (2000), “Bundling 62–68.
and Competition on the Internet,” Marketing Sci- Cox, W.E. (1967), “Product Life Cycles as Marketing
ence, 19 (1), 63–82. Models,” Journal of Business, 40 (10), 375–84.
Bass, F.M. (1980), “The Relationship Between Diffusion Dewan, S. and C. Min (1996), “The Substitution of
Rates, Experience Curves, and Demand Elasticities Information Technology for Other Factors of Pro-
for Consumer Durable Technological Innovations,” duction: A Firm Level Analysis,” Management Sci-
Journal of Business, 53, Part 2 (July), 51–68. ence, 43 (12), 1660–75.
Bohi, Douglas R. (1981), Analyzing Demand Behavior: Freeman, D.G. (2003), “Is the Health Care a Necessity or
A Study of Energy Elasticities. Baltimore, MD: The a Luxury? Pooled Estimates of Income Elasticity
Johns Hopkins University Press. from U.S. State-Level Data,” Applied Economics,

35, 495–502. 155–63.
Getzen, T. (2000), “Health Care is an Individual Necessi- Nagle, P. (1987), The Strategy and Tactics of Pricing.
ty and a National Luxury: Applying Multilevel Deci- Englewood Cliffs: Prentice-Hall, Inc.
sion Models to the Analysis of Health Care Expendi- OECD (2002), National Accounts of OECD Countries.
tures,” Journal of Health Economics, 19, 259–70. Volume II. 1989–2000.
Jeuland, A.P. (1981), “Parsimonious Models of Diffusion Onkvisit, Sak and John J. Shaw (1989), Product Life
of Innovation, Part B: Incorporating the Variable of Cycles and Product Management. Connecticut: Quo-
Price,” Working Paper, University of Chicago. rum Books, Greenwood Press, Inc.
Kalish, S. (1983), “Monopolistic Pricing with Dynamic Polli, R. and V. Cook (1969), “Validity of the Product Life
Demand and Production Cost,” Marketing Science, 2 Cycle,” Journal of Business, 42 (10), 385–400.
(Spring), 135–60. Robinson, B. and C. Lakhani (1975), “Dynamic Price
Klonaris, S. and D. Hallam. (2003), “Conditional and Models for New-Product Planning,” Management
Unconditional Food Demand Elasticities in a Dy- Science, 21, 1113–22.
namic Multistage Demand System,” Applied Eco- Scheuing, E.E. (1969), “The Product Life Cycle as an Aid
nomics, 35, 503–14. in Strategy Decision,” Management International
Levitt, T. (1965), “Exploit the Product Life Cycle,” Har- Review, 9 (4/5), 111–24.
vard Business Review, 43 (11), 81–94. Simon, H. (1979), “Dynamics of Price Elasticity and
Lu, M., Y. Tu, and S. Thompson (2004), “Computers and Brand Life Cycles: An Empirical Study,” Journal of
Packaged Software: Necessary or Luxury Goods?” Marketing Research, 16, 439–52.
Working Paper, OPIM, Business School, University Tellis, G.J. (1988), “The Price Elasticity of Selective
of Connecticut, Storrs. Demand: A Meta-Analysis of Econometric Models
Matteo, L.D. and R.D. Matteo (1998), “Evidence on the of Sales,” Journal of Marketing Research, 25 (No-
Determinants of Canadian Provincial Government vember), 331–41.
Health Expenditures: 1965–1991,” Journal of Health Tellis, G. and C. Fornell (1988), “The Relationship Be-
Economics, 17, 21–228. tween Advertising and Quality Over the Product Life
Mdafri, A. and B.W. Brorsen (1993), “Demand for Red Cycle: A Contingency Theory,” Journal of Market-
Meat, Poultry, and Fish in Morocco: An Almost Ideal ing Research, 15, 64–71.
Demand System,” Agricultural Economics, 9 (2),

Yanbin Tu
Department of Operations and Information Management
Business School
University of Connecticut
Storrs, CT 06269
Phone: 860.486.6485
FAX: 860.486.4839
E-Mail: yanbin.tu@business.uconn.edu

COMPUTATIONAL ESTIMATION OF PARTIONED PRICES: ANOTHER
HEURISTIC MOVES INTO THE MARKETING NEIGHBORHOOD
William J. Jones, University of Kentucky, Lexington
Devon S. DelVecchio, University of Kentucky, Lexington
Terry L. Childers, University of Kentucky, Lexington
SUMMARY This research addresses three key questions exam-

ined via three studies. First, what varieties of computa-
Economic choice models originally conceived con- tional estimation strategies are consumers likely to em-
sumers as rational agents making optimal choices. Later, ploy when revising a discounted price? Second, given
consumer researchers would present evidence of a con- computational estimation strategies how do interactions
sumer who is rational within the bounds set by the desire between the numerical value of the price and discount
to maximize outcomes given an acceptable level of mental influence consumer choice? Specifically, we examine
effort (e.g., Payne, Bettman, and Johnson 1993; Simon situations in which price and discount combinations sug-
1978). In such a “satisficing” mode, consumers rely on gest a particular computational estimation strategy and
heuristics to reduce processing demands. One of the most how the effort expended in using this strategy affects
widely cited heuristics is anchoring and adjustment. An- choice share. Third, does the affect generated from em-
choring and adjustment has often been used to explain the ploying more or less effortful computational estimation
integration of price information (e.g., Morwitz, Greenleaf, procedures increase or decrease choice share for the
and Johnson 1998; Yadav and Seiders 1998). This may be brand. In or initial study, we examine strategies that
a popular heuristic among consumers since it requires consumers use when estimating the value of a discount for
little mental effort. However, anchoring and adjustment 144 price-discount combinations. Study 2 examines choice
typically fails to provide highly accurate estimates (e.g., behavior for a discounted brand and how price-discount
Jacowitz and Kahneman 1995). combination s favoring particular computational estima-
tion strategies increase or decrease share for the brand
In this study we introduce computational estimation relative to other brands. Study 3 largely replicates Study 2,
as an alternative explanation to consumers’ use of parti- but measures affect and confidence generated through the
tioned prices. Dowker et al. (1996) define computational estimation process.
estimation as a procedure in which elements of simple
computation and approximation skills are combined to Because of its higher accuracy, lower effort niche, we
devise strategies for manipulating numbers. The strate- expect that consumers employ computational estimation
gies in computational estimation include, but are not strategies in a variety of situations. Thus, our point-of-
limited to, rounding, using known numbers (e.g., trans- view is consistent with a bounded rationality view of
forming $106 X .23 to $100 X .25), distributing one consumer decision-making, but one that supposes greater
number into two or more (e.g., 15% is broken into 10% + mental effort and accuracy than the predominant heuris-
5%), and use of fractions (e.g., 80 X 25 = 80 X ¼ X 100). tic. This view is supported by results that differ from those
Thus, computational estimation procedures enable con- that would be predicted by anchoring and adjustment.
sumers to evaluate price information in a more sophisti- However, additional studies that directly contrast anchor-
cated manner than other heuristics, but without investing ing and adjustment and computational estimation in terms
the large amount of mental energy needed to perform of consumers’ price estimates and choice are warranted.
exact calculations.

William J. Jones
Gatton School of Business and Economics
University of Kentucky
Lexington, KY 40507
Phone: 859.275.2962
FAX: 859.257.3577
E-Mail: William.jones@uky.edu

THE EFFECTS OF MAGNITUDE REPRESENTATION ENCODING
INTERFERENCE AND ORDER OF PRICE EXPOSURE IN
COMPARATIVE PRICE ADVERTISING
Keith S. Coulter, Clark University, Worcester
Robin A. Coulter, University of Connecticut, Storrs
SUMMARY affects their perceptions (Monroe 2003). Because con-

sumers typically examine the items in a list from top to
Research has demonstrated that consumers often rely bottom, prices at the top of the list become externally-
on the non-conscious, automatic processing of price in- supplied reference prices that influence the perceptions of
formation when choosing among brands. The numerical prices that occur lower on the list. Buyers who initially see
cognition literature suggests that one way in which nu- high prices perceive subsequent lower prices as less
merical stimuli (and hence prices) may be non-consciously expensive (and hence of greater value) than they would if
represented and encoded in memory is in terms of magni- they initially see low prices (Slonim and Garbarino 1999).
tude representations. Magnitude representations are judg-
ments of relative “size” arrayed in analog format along a Thus, in the case of the aforementioned price lists
left-to-right oriented mental number line, and may reflect consisting of standard/sale prices in different size fonts,
either the exact value (e.g., 8), or an approximation of the we also consider that these prices may occur in either
exact value (e.g., “large”) of a number (Dehaene 1992). descending (e.g., $20–18, 16–14,12–10) or ascending
Research further suggests that the magnitude representa- (e.g., $4–2, 8–6,12–10) series. Thus, relative to a particu-
tion that sustains the processing of numeric value may be lar target item that appears at the bottom of the lists and
highly related to the underlying magnitude code that hence is typically examined last (i.e., the item priced at
sustains the processing of physical stimuli (Dehaene and $12–10 in the example above), initial reference prices
Akhavein 1995). Thus, interference may ensue if the would be higher in the descending series, and lower in the
magnitude representation associated with the numeric ascending series.
value of a number (e.g., large) is inconsistent with the
magnitude representation associated with the physical In the case of the descending order series, both
size or appearance of that number (e.g., small). standard and sale initial reference prices would be greater
than a target sale price appearing at the bottom of the list.
Consider, then, a set of price lists containing com- Thus we hypothesize that a (small) congruent physical
paratively high (standard) and low (sale) prices for several size magnitude representation associated with that target
items within a particular product category. The lower sale price should reinforce the perception that the target
prices are all displayed in either smaller font, such that the price is actually lower or smaller (and hence a better buy)
numerical value and physical size dimensions are congru- than the higher reference price(s). Both size/value con-
ent (e.g., $12–10), or larger font such that the numerical gruence and price-order effects are expected to occur.
value and physical size dimensions are incongruent (e.g., Conversely, a (large) incongruent physical size magni-
tude representation associated with the target sale price is
$12–10). In the case of the incongruent lists, we hypoth- expected to interfere with the perception that the target
esize that the non-corresponding magnitude representa- price is lower or smaller than the higher reference price(s).
tions should interfere with consumers’ ability to encode Therefore in this latter instance, price-order effects should
the sale prices as lower or smaller than the standard high be diminished.
prices. As a result, the differences between the standard
high prices and the low sale prices should be perceived as Method
less, and the more favorable price and/or value assess-
ments typically associated with the comparatively low Hypotheses were tested in a 2 (congruent/incongru-
sale prices should be weakened or reduced. The reduced ent) x 2 (ascending/descending order) experiment (N =
value assessments should lead to a decrease in demand for 120). Subjects were shown a list of (6) brands in table
each of the low-priced items. The opposite effects are format containing high (standard) and low (sale) prices
expected in the case of the congruent physical-size/nu- and brief product descriptions. In the congruent (incon-
meric value lists. gruent) conditions the higher standard prices appeared in
larger (smaller) font than the lower sale prices. Brands
Research has also shown that the order in which were listed in either ascending or descending order of
buyers are exposed to multiple prices (i.e., in list format) price, with a fictitious target brand always appearing at the

bottom of the list. After viewing the stimulus price list, value congruency interacts with price-presentation order
subjects were exposed to a brief “filler” infomercial. They such that price-order effects are eliminated under incon-
then completed a paper and pencil questionnaire. gruent magnitude representation conditions. Processing
check results supported our contention that subjects’
Results processing of price information was implicit, automatic,
and non-conscious, rather than explicit, conscious, and/or
Findings indicated that congruent magnitude repre- inferential. Results also indicated that differing degrees of
sentations result in more favorable value perceptions, attention caused by the size of the larger/smaller fonts
lower price judgments, and increased purchase likeli- could not have accounted for dependent variable results.
hood. Findings also demonstrated that numerical size/ References available upon request.

Keith S. Coulter
Clark University GSOM
950 Main Street
Worcester, MA 01610–1477
Phone: 508.793.7749
FAX: 508.793.8822
E-Mail: kcoulter@clarku.edu

INFLUENCES ON WHAT CONSUMERS KNOW AND WHAT THEY
THINK THEY KNOW REGARDING THE PERSUASIVE ASPECTS
OF PRICING-RELATED SELLING TACTICS
Jay P. Carlson, The Graduate College of Union University, Schenectady
William O. Bearden, University of South Carolina, Columbia
David M. Hardesty, University of Miami, Coral Gables
SUMMARY knowledge or self-assessed perceptions regarding what

consumers think they know. To this point, no research has
An array of pricing-related selling tactics (i.e., tactics attempted to investigate directly influences on objective
used by sellers to generate favorable price perceptions and subjective knowledge or any moderators of the rela-
regarding their brands, stores, and/or offerings) exists in tionship between objective and subjective knowledge
the marketplace. Prior research examining pricing-related regarding consumer persuasion knowledge of marketer
selling tactics has focused on assessing consumer re- pricing-related selling tactics. To address this objective,
sponse to individual tactics (e.g., partitioned pricing – the present paper describes the results of a study that
Morwitz, Greenleaf, and Johnson 1998; pennies-a-day – investigated issues related to both objective and subjec-
Gourville 1998; price bundling – Yadav and Monroe tive knowledge of the persuasive aspects of pricing-
1993). Research that has studied consumer price knowl- related selling tactics.
edge has done so in the context of individual prices (e.g.,
Dickson and Sawyer 1990; Mazumdar and Monroe 1992; Hypothesized Relationships
Monroe and Lee 1999; Vanhuele and Dreze 2002). Also,
consumer persuasion knowledge, as advanced by Friestad Preference for numerical information (Viswanathan
and Wright (1994), has been investigated in the domains 1993), need for cognition (Cacioppo and Petty 1982), and
of advertising (Boush, Friestad, and Rose 1994) and experience with pricing-related selling tactics were all
personal selling (Campbell and Kirmani 2000). However, predicted to be positively related with objective knowl-
consumer persuasion knowledge in the domain of pricing, edge. The personal relevance of pricing-related selling
more specifically, for marketer pricing-related selling tactics, frame of reference (i.e., the beliefs about the
tactics, has not been examined. knowledge possessed by significant others in an individ-
ual’s life – Radecki and Jaccard 1995), experience with
Alba and Hutchinson’s (2000) review of knowledge pricing-related selling tactics, and objective knowledge
calibration supports the contentions of prior researchers were all hypothesized to be positively related with subjec-
(e.g., Brucks 1985; Park, Mothersbaugh, and Feick 1994) tive knowledge. Lastly, experience was predicted to mod-
that objective and subjective knowledge are two impor- erate the relationship between objective knowledge and
tant but distinct knowledge constructs that differentially subjective knowledge such that this relationship would be
impact information search and consumer evaluations. weaker when experience was high relative to when expe-
Moreover, prior research examining the relationship be- rience was low.
tween objective and subjective knowledge has not always
found consistent results, and researchers argue that a Study Overview
better understanding of the relationship is needed (Cow-
ley and Mitchell 2003). The current research examined The eight hypotheses were tested from data collected
both objective knowledge and self-perceived knowledge in a survey research study involving a sample of 191
regarding the persuasive aspects of marketer pricing- undergraduate students. The data were collected in two
related selling tactics in order to better understand the time periods, separated by one week. Measures of objec-
relationship between the two knowledge constructs. tive knowledge, subjective knowledge, and experience
were collected in phase one, with preference for numeri-
The objective then of the current research was to cal information, need for cognition, personal relevance,
identify and test a set of hypothesized factors that influ- and frame of reference collected in phase two.
ence two knowledge constructs as they relate to knowl-
edge regarding the persuasive aspects of marketer pric- Results and Discussion
ing-related selling tactics (cf., Park et al. 1994; Radecki
and Jaccard 1995). The first of these constructs was Briefly, the results provided support for several hy-
objective knowledge or accurate information stored in pothesized antecedents of both objective and subjective
long-term memory. The second construct was subjective knowledge and suggest that experience is a key moderator

of the objective knowledge – subjective knowledge rela- knowledge, while the relationship between frame of ref-
tionship in the domain of pricing-related selling tactics. erence and subjective knowledge received mixed support.
Specifically, need for cognition and experience with pric- Lastly, the relationship between objective knowledge and
ing-related selling tactics were positively and significant- subjective knowledge was found to be positive and signif-
ly related with objective knowledge, while the relation- icant when experience was low, but was non-significant
ship between preference for numerical information and when experience was high. This result suggests that
objective knowledge received mixed support. In addition, consumers might be susceptible to overconfidence when
personal relevance, experience, and objective knowledge experience is high but objective knowledge is low. Refer-
were positively and significantly related with subjective ences available upon request.

Jay Carlson
The Graduate College of Union University
Schenectady, NY 12308
Phone: (518) 388.6738
FAX: (518) 388.6754
E-Mail: carlsonj@union.edu

CREATIVITY IN ADVERTISING: PURCHASE INTENT AND BRAND
ATTITUDE EFFECTS
Brian D. Till, Saint Louis University, St. Louis
Daniel W. Baack, Saint Louis University, St. Louis
SUMMARY H1: Creative advertising will lead to higher levels of

purchase intent for the advertised brands.
Creativity is a very important component of advertis-
ing. With the industry focus on creative advertising, there H2: Creative advertising will lead to more positive atti-
is surprisingly little research looking directly at the effec- tudes for the advertised brands.
tiveness of award-winning advertising. There are only a
handful of studies that focus specifically on this topic. Methodology
First, the research by Kover et al. (1995) investigated the
link creative advertising and consumers’ responses to that For the creative commercials, we used Communica-
advertising. They found that student defined “creative” tion Arts award winners. Communication Arts bases its
advertisements were linked to higher purchase intent. The awards primarily on ad creativity as assessed by a panel of
second study, Ang and Low (2000), investigated the distinguished advertising professionals. Forty advertise-
relationship between advertising creativity and subjects’ ments were randomly selected from a pool of winners
affective responses. The authors found that creative ads from three recent years to form the sample. To create the
were consistently perceived as more favorable, and, to a pool of control advertisements, advertisements were sam-
lesser degree, resulted in a more favorable view of the pled during prime time television during four randomly
brand and increased purchase intent. The third study, selected days of the week. From this pool, forty were
Stone et al. (2000), attempted to link creative commercials randomly selected. We inserted the commercials selected
to likeability. The study found that liked commercials into two television programs (Dream Living and Ground
were more often judged as creative. Two recent studies, Forces). We used two programs embedded with a differ-
Till and Baack (2004a) and Till and Baack (2004b) ent set of 20 commercials (½ creative, ½ control) for intra-
explore the effectiveness of creative advertisements by study replicability. Therefore, our study used four two and
linking exposure to creative advertising (measured as a half minute pods with five thirty second commercials
winning an advertising award) to increased unaided recall each. This resulted in a total of ten minutes of advertising
of both the brand advertised and the features of the (twenty commercials) per program. Subjects were 69
commercial. Till and Baack (2004a) finds that creative undergraduate students enrolled in business courses at a
advertisements led to greater unaided brand and feature Midwestern university. Three weeks before watching the
recall. Till and Baack (2004b) replicated their first study, television program, subjects complete a questionnaire
but used a one-week delay measure of recall. They found measuring the attitude and purchase intentions regarding
that brand names and commercial features are better the brands advertised. After watching the television pro-
recalled for creative commercials, but only in the unaided gram, the subjects filled out the same questionnaire previ-
recall task. ously completed. To measure purchase intent and brand
attitude, subjects were asked to rate on a seven-point,
All of the above studies, while significant, leave bipolar adjective scale purchase intent and brand attitude.
some issues for further investigation. First, for most of the Both of these measures were based on previous studies
studies, the grouping of advertisements into creative and (e.g., Chapman and Aylesworth 1999; Till and Busler
non-creative groups was based on student ratings. Stu- 2000).
dents may not be the most appropriate judges of the
creativity of an advertisement (White and Smith 2001; Results and Discussion
Kover et al. 1995). This potential problem with student
raters cast particular doubt on studies of the affective The unit of measure for the analysis is the change in
effectiveness of creativity advertising. To further investi- purchase intent and brand attitude between pre and post
gate this issue, this study explores the link between advertisement exposure (statistics available upon request).
creative advertising and brand attitude and purchase in- For both programs, a paired t-test analysis revealed no
tent, but uses professionals, not students, to rate creativity. significant differences between control and creative ad-
Specifically, the following hypotheses are investigated: vertisements for brand attitude and purchase intent.

This research focused on the effect of creativity on ingrained and less influenced by a single ad exposure. It
both purchase intent and attitude toward the brand. The also possible that differences in findings between this and
results find suggest that creative advertisements, at least previous research is due differences in the types of judges
with limited viewing, do not appear to affect purchase used to evaluate advertising as creative. While past stud-
intent or attitude toward the brand. These results cast ies have found creative advertisements to be effective
doubt on past studies finding that creative advertising led using similar measures (e.g., Ang and Low 2000; Kover
to increased affective effectiveness (Ang and Low 2000; et al. 1995; Stone et al. 2000), they used students to assess
Kover et al. 1995; Stone et al. 2000). It is possible that creativity. We feel that the use of professionals to judge
these results reflect the difficulty in affecting consumer the advertising is a far more valid measure of “creativity”
brand attitude and intent to purchase with one commercial and may account for the different results of the present
exposure. Unlike measures of recall, as used in Till and research versus earlier studies. References available upon
Baack (2004a) and Till and Baack (2004b), brand attitude request
and purchase intent of known brands is likely more solidly

Daniel Baack
Saint Louis University
3674 Lindell Boulevard
St. Louis, MO 63108–3397
Phone: 314.977.3810
FAX: 314.977.3897
E-Mail: baackdw@slu.edu

WHERE’S THE AFFECT? AN INVESTIGATION OF THE EFFECT OF
THREE ADVERTISING SCALES ON ATTITUDE TO THE AD
Arjun Chaudhuri, Fairfield University, Fairfield
ABSTRACT ad. For instance, Chaudhuri and Buck (1995a; see also
Chaudhuri 2004) showed that reptilian affects are not
A natural viewing situation is used to examine the related to liking for the ad while prosocial affects are
efficacies of three scales that purport to account for positively, strongly and significantly related. Burke and
emotional and rational responses to advertising. With Edell (1989) showed that upbeat, warm and negative
attitude to the ad as the dependent variable, the three scales feelings differ in their effects as well.
are tested using a large number of subjects, a random and
large number of ads and a sufficiently large number of One problem with the operationalization of this last
independent observations (each subject was exposed to a approach has been that researchers who use the Edell and
single ad only). The results indicate differences in the Burke (1987) feelings scale tend to shorten the scale (53
prediction of attitude to the ad among the three scales. items) and adapt it to their purposes (as an example, see
Surprisingly, the most parsimonious of the three scales MacInnis, Rao, and Weiss 2002). In general though,
achieves the greatest predictive ability. measurement scales are developed with the understand-
ing that they should be used in their entirety and not
INTRODUCTION abridged according to the purposes of a particular piece of
research. Thus, there would seem to be a need for a shorter
The purpose of this study is straightforward. Can a scale which still captures all the multidimensional aspects
short scale for assessing the emotional and rational reac- of qualitatively different affects. The CASC scale offers
tions to advertisements predict the attitude to these ads as such an alternative with only 16 items which purport to
well or better than two scales that are well known but are measure four different affective and cognitive dimen-
considerably longer in length? If so, the implications of sions. One of the goals of this study was to test if the CASC
the study for practical purposes of data collection are also scale had the same predictive power of the feelings and
obvious. Respondent fatigue will be lowered and better judgment scales. If so, then it presents a viable and
quality of responses may be expected with a substantially perhaps more practical alternative to the other two scales.
smaller scale that also addresses the multidimensional
properties of responses to advertising. Although there has been considerable work, as cited
above (see also Batra and Ray 1986; Holbrook and West-
Accordingly, this study compared the “feelings” scale wood 1989 and others), that suggests that a range of
(53 items) developed by Edell and Burke (1987) and the emotional effects are evoked by advertising, no single
“judgments” scale [25 items taken from the Reaction accepted theoretical paradigm has emerged that guides
Profile for TV commercials (Wells, Leavitt, and McCon- research in this area. Individual studies have used a variety
ville 1971)] also used by these authors (Burke and Edell of theoretical frameworks, such as Plutchik’s (1980)
1989; Edell and Burke 1987; Burke and Edell 1986) with typological approach (Holbrook and Westwood 1989)
the relatively short (16 items) CASC (Communication and Mehrabian and Russell’s (1974) dimensional ap-
Analytic and Syncretic Cognition) scale developed by proach (Olney, Holbrook, and Batra 1991). None of these
Chaudhuri and Buck (1995a, 1995b, 1998). The three approaches, however, are clearly anchored in the struc-
scales were examined in terms of their ability to predict ture of the brain. The theory of the “triune brain,” as
attitude to the ad. adapted in the CASC scale, described below, bases itself
on knowledge of the human brain.
Affect has been operationalized in different ways in
consumer research and advertising studies. One way has THE CASC SCALE
been to view affect as an overall, global reaction such as
“liking” for an ad (Haley and Baldinger 1991; Herr and This scale was developed in the area of communica-
Page 2004; Mitchell and Olson 1981). Another way has tion sciences (Chaudhuri and Buck 1995a) and subse-
been to assess affect as an amalgam of a set of qualitatively quently refined and published in the social psychology
different types of affects that may be evoked by an ad. This and consumer research literature (Chaudhuri and Buck
has been prompted by evidence (Chaudhuri and Buck 1995b, 1998). It claims, on the basis of confirmatory
1995a; Edell and Burke 1987) that different dimensions of factor analysis and repeated testing, to measure both
affect can have strikingly different effects on liking for the emotional and rational (syncretic and analytic cognitions)

responses to advertising with a short list of 16 items which The general form of the scale is “Did the ad make you feel/
load on four different factors. It has been tested at both the think. . . .” In all there are sixteen items, four in each
individual level of subjects and at the aggregate level dimension. The prosocial dimension comprises of the
using ads as the units of analysis. It has been tested for both questions “did the ad make you feel happy/proud/hopeful/
print and television media. Its results have been replicated a sense of affiliation.” The individualistic dimension
across different subjects and different ads. The CASC contains the items “did the ad make you feel angry/afraid/
scale also has the virtue of being grounded in a widely disgusted/irritated.” The reptilian dimension consists of
accepted theory based upon scientific evidence on the the items “did the ad make you feel sexy/aggressive/
workings of the human brain as described next. envious/a sense of power.” The analytic dimension is
based on brand differentiation and asks questions on
The scale uses the theory of the triune brain (McLean whether the ad made the respondent think of the “facts
1973, 1990) as its conceptual basis. According to McLean, about the brand/pros and cons of the brand/arguments for
the human brain has developed in a series of stages and using or not using the brand/differences between the
today there are essentially three brain structures that are brand and its competitors” (Chaudhuri and Buck 1995b,
interconnected and represent a “triune brain.” The first of 1998). Thus, the CASC scale measures both emotional
these structures is the reptilian brain (reticular formation, and rational dimensions of responses to ads. The emotion-
basal ganglia, and midbrain) that governs such basic al items measure a spectrum of qualitatively different
behavior as reproduction, aggression, territoriality, etc. affective responses that are relevant to advertising while
The second brain structure is the paleomammalian forma- the rational items refer to the processing of the elements
tion (limbic system) found in mammals which guides in an ad which differentiate between brands. Brand differ-
prosocial behavior associated with the preservation of the entiation strategies have repeatedly been found to be the
species and also agnostic behavior associated with the single most important executional factor that influences
preservation of the individual. The third, and most recent, ad effectiveness (Stewart and Furse 1986; Stewart and
structure is the neomammalian formation (neocortex and Koslow 1989) and Holbrook (1987) specifically men-
thalamic structures) and its primary functions are in the tions such strategies among the more desirable effects of
realm of higher order cognitive processing, including advertising.
verbal communication, language, ideas, problem solving,
complex learning, and memory (Chaudhuri and Buck THE FEELINGS SCALE
1995b, 1998).
The feelings scale was developed by Edell and Burke
The reptilian counterpart of the human brain is still (1987) in trying to understand the role of feeling respons-
considered to influence us in certain basic behavior such es to an ad (as distinct from thoughts about an ad) in
as the struggle for power, adherence to routine, imitation, promoting advertising effects. They took their cue from,
obeisance to precedent and deception. These are some of among others, Zajonc (1980) in trying to establish that
the same behaviors that are espoused in advertising – affective responses are valid, reliable, and unique evalu-
“never let them see you sweat” (deception), “did you DQ ations of objects that are independent of cognitive evalu-
today?” (adherence to routine), “caviar for the power ations. Moreover, they asserted and established that ads
hungry” (power), etc. While the reptilian brain programmes were capable of simultaneously evoking both positive and
stereotypical behaviors, the old mammalian portion of the negative feelings among viewers (unlike positive and
brain, or the limbic system, functions in the subjective negative cognitive evaluations) and that such feelings
experience of feelings and desires. In general, circuits were as important in understanding and explaining adver-
involving the septal area are involved in prosocial feelings tising effects.
associated with behaviors conducive to the preservation
of the species, nursing, maternal care, parenting, and play. Edell and Burke (1987) used 69 items which were
In contrast, circuits involving the amygdala are concerned drawn from a larger pool of items in a pretest prior to their
with individualistic feelings such as fear, anger and dis- study. They reported that three factors with high internal
gust, which are associated with self preservation and self consistency between the factors were extracted from a
protection (Chaudhuri and Buck 1995b, 1998). factor analysis of the items. They described these factors
as upbeat (“active/confident/good/lively, etc.”), negative
The CASC scale was developed within the frame- (“angry/bad/dull/sad, etc.”) and warm (“affectionate/calm/
work of the triune brain and contains multidimensional kind/sentimental, etc.”). These three factors, the three
elements which conform to the mammalian brain (proso- factors from the judgments scale (Burke and Edell 1986)
cial and individualistic dimensions), the reptilian brain and prior attitude to the brand were used to predict attitude
(reptilian dimension) and the neomammalian brain (ana- to the ad. It was found that the upbeat and warm feelings
lytic dimension). CASC is a seven point paper and pencil scales did not add uniquely to the explanation of variance
scale anchored at two endpoints by “not at all” and “a lot.” in attitude to the ad but that they were jointly significant.

The authors concluded that, overall, the feelings dimen- subjective responses by individuals to ads and it is not a
sions contributed importantly and differently from the scale which assesses the objectively existing elements in
judgments scale. Edell and Burke (1987) also reported an ad such as the use of scenery, actors, music, etc. Such
that in a follow up study they reduced the list of items to reliably verifiable elements in an ad are often used in
56 items based on the redundancy noted between the content analysis of ads as in the method described by
items. Stewart and Furse (1986). The items in the judgments
scale constitute evaluations of ads and, thus, may be better
Examination of the items in the feelings scale shows viewed as responses that are a result of the cognitive
that the three factors in this scale have some important processing of ads and which are considered opinions of
differences with the factors in the CASC scale. First, the the overall effectiveness of ads while the feelings scale
four items in the reptilian dimension of the CASC scale measures affective responses to ads which are more spon-
are not represented at all in the feelings scale. Second, the taneous and less considered. Note, however, that the
negative dimension of the feelings scale has three of the judgments scale is different from the analytic dimension
four items in the individualistic dimension of the CASC of the CASC scale which measures the cognitive process-
scale with the notable exception of the “fear” item. Third, ing of brand differentiation (“did the ad make you think of
the feelings scale possesses three of the four items in the the differences between the brand and its competitors,”
prosocial dimension of the CASC scale with the exception etc.) in an ad while the judgments scale asks for a cogni-
of the “sense of affiliation” item in the CASC scale. tive evaluation of the ad in terms of overall adjective based
Further, in the feelings scale, these three items are not all judgments about the ad. Thus, the analytic dimension of
on the same factor as in the CASC scale. Fourth, in spite the CASC scale should have a different and unique effect
of having four dimensions, the CASC scale is far shorter on attitude to the ad over and above the effect of the
(sixteen items) than the feelings scale (53 items). Finally, judgments scale.
the analytic dimension (four items) of the CASC scale is
not represented in either the feelings scale or in the In any case, controlling for both affective and cogni-
judgments scale as discussed below. tive responses to ads provides for better model specifica-
tion in the present study. Thus, the judgments scale was
THE JUDGEMENTS SCALE also included in the study.
The “judgments” scale was used by Burke and Edell STUDY GOALS
(1986) as an adjective based attitude to the ad scale in
order to capture ad reactions over time in a naturally The following research goals formed the motivation
occurring situation. These authors used the scale to test for the study. Attitude to the ad was chosen as the depen-
television commercial wearout – i.e., to see if subjects’ dent variable on which to test the predictive ability of the
evaluation of ads declined as levels of exposure increased. three scales since these scales all purport to measure
They used thirty of the items from the Reaction Profile for responses to advertising and not necessarily to brands or
TV commercials developed by Wells, Leavitt, and McCo- companies. Further, attitude to the ad has been found to be
nville (1971) and four additional items from other re- an important concept in determining the persuasiveness of
search on attitude to the ad in order to compile the scale. advertising messages (MacKenzie and Lutz 1989; Mue-
The scale was found to be reliable (internal consistency) hling and McCann 1993).
and to have a stable three factor structure consisting of the
34 items. Notably, it was found in this study that ratings 1. Which scale is better overall at predicting attitude to
for the scale did not change over time while the ratings of the ad?
an overall attitude to the ad scale did. Thus, this scale is
different from an overall attitude to the ad scale. 2. Which dimensions of the three scales predict attitude
to the ad?
In a later study, Edell and Burke (1987) reported
reducing the scale to 25 items based on the redundancy 3. Are the dimensions of these scales stable under the
among some of the items. These authors referred to the following conditions?
three factors of the scale as the “judgments” scale and they
found that this scale was qualitatively different from the • When there is a natural viewing situation. While
feelings scales also used in their study. These authors also the feelings and judgments scales were devel-
emphasized that the scale measures judgments about the oped in a naturally occurring setting, the CASC
characteristics of an ad such as Humorous, Informative, scale was not.
Gentle, Valuable, etc. while the feelings scale measures
responses that are properties of the individual (happy, sad, • With a random sample of ads. None of the scales
etc.) and not a characteristic of the ad itself. This view may were developed under this condition.
be contested since the judgments scale also measures

• With a large sample of ads. Only the CASC scale RESULTS
used over a hundred ads.
In order to investigate the structure of the three scales
• With a large sample of subjects. None of the the data in this study were analyzed using Principal
scales used a large sample of individual subjects Components Analysis. This is appropriate since all three
(not exposures). scales report the use of exploratory factor analysis in the
development of the scales.
• When each observation is independent of the
others. All of the scales were developed using The CASC scale in this study was found to comprise
subjects who responded to multiple ads. This of three factors with an eigenvalue greater than one. A
causes autocorrelation, since the units of obser- fourth factor with an eigenvalue of .93 was also extracted
vation are not independent of each other, result- but not considered further since the cutoff for all the scales
ing in biased estimates. was taken to be the customary eigenvalue of one. Howev-
er, the presence of this fourth factor should be noted since
METHODS the CASC scale is supposed to have four dimensions.
Varimax rotation of the remaining three factor structure
Data Collection extracted 54 percent of the variance in the items. All
sixteen items loaded higher than .50 on one of the three
The data were collected in a natural viewing situa- factors and not higher than .40 on any other factor. With
tion. Four hundred and forty eight undergraduate subjects the exception of the fourth factor which narrowly missed
at a private university in the northeast were interviewed in the cutoff, the factor structure of the CASC scale seems to
their dorm rooms while watching television. Respondents be somewhat stable with three of the four expected factors
were asked to tune into one of the major television performing as expected for the set of randomly selected
networks that they normally watched and to watch some ads. All the four items for the prosocial dimension of the
program content. During the first commercial break one scale loaded on the first factor along with all the four items
of the commercials in the pod was randomly chosen and in the reptilian dimension. Similarly, all the four items for
at the end of the commercial the respondent was asked to the analytic and individualistic items loaded on the second
switch off the television and to fill in a questionnaire about and third factors respectively. The items were summed
the commercial just viewed. according to the factors they loaded on.
As a result of asking one subject to react to one ad The feelings scale, which was expected to comprise
only, a data set of 448 independent observations was of three factors, resulted in eleven factors for the 53 items
obtained for the measures discussed next. used in this study and explained 64.8 percent of the
variance in the items. Moreover, the varimax rotated
Measures factor solution failed to converge. Examination of the
initial factor matrix showed that only the first three factors
The questionnaire contained items from all the three had loadings greater than .50 and, accordingly, only these
scales and also on an attitude to the ad scale. The CASC three factors were used in further analysis. Burke and
scale had 16 items all of which were taken from Chaudhuri Edell (1989) also state that these items should result in
and Buck (1998). The feelings scale consisted of the 53 three factors. Thus, only these 39 items were used in
items described by Burke and Edell (1989) which loaded further analysis and they were summed according to the
greater than .5 on a factor in that study. The judgments factors they loaded on. Thirty one of the items loaded
scale had the 25 items refined and used by Burke and Edell greater than .5 on the first factor; 6 items on the second
(1989) and Edell and Burke (1987). The instructions to the factor and 2 items on the third factor.
subjects reproduced the instructions provided by Chaudhu-
ri and Buck (1998) and Edell and Burke (1987) for the The 25 items in the judgements scale loaded on five
three scales.1 factors which together explained 60.8 percent of the
variance in the items. Burke and Edell (1989) reported that
The attitude to the ad scale was measured as the sum these 25 items loaded on only three factors in their study
of the responses to a seven point semantic differential but this was not the case with regard to the data in the
scale with the following five items: pleasant/unpleasant, present study. However, the original formulation of this
unfavorable/favorable, unlikeable/likeable, good/bad, and scale (Wells, Leavitt, and McConville 1977; Wells 1964)
negative/positive. These items have been widely used to reported six dimensions and, thus, the five factors were
measure attitude to the ad in previous research (Muehling accepted as valid dimensions to investigate further. Ac-
and McCann 1993). cordingly, the items which loaded greater than .5 on each

of the factors were summed on their respective factors. All attitude to the ad and all three of the dimensions in the
25 items loaded greater than .5 on one of the five factors CASC scale featured in this set of explanatory variables.
and did not cross load higher than .50. There were a
minimum of four items on each factor. DISCUSSION
The three scales were next analyzed with regard to Of the three scales in the analysis described above,
their effects on attitude to the ad. Multiple regression was the CASC scale emerges as the most successful in predict-
used for this analysis. First, the feelings scale was used to ing consumer attitudes to advertising. First, it is the only
predict attitude to the ad. All three dimensions of the scale in which all the dimensions made unique and signif-
feelings scale together accounted for 36.8 percent of the icant contributions to attitude to the ad. In fact, the
variance in attitude to the ad and only two of the three individualistic dimension of the CASC scale made the
factors were significantly (p. < .05) related to attitude to highest contribution of all the dimensions in the three
the ad. Second, the judgments scale was used to predict scales. The prosocial dimension made the third highest
attitude to the ad. The five dimensions of the scale ac- contribution and the analytic dimension made a smaller
counted for 45 percent of the variance in attitude to the ad but unique and significant contribution. Second, even
and only four of the five factors in the scale were signif- with only 16 items it accounted for more of the variance
icantly (p. < .05) related to the dependent variable. Last, (44.2%) in attitude to the ad than the 39 items used from
the CASC scale was used to predict attitude to the ad. The the feelings scale (36.8%)and almost as much of the
three dimensions of the CASC scale accounted for 44.2 variance (45%) as the 25 items in the judgments scale.
percent of the variance in attitude to the ad and all three Last, it accounted for 9.4 percent incremental explanation
dimensions were significantly related. of variance in attitude to the ad over and above the
contributions of the other two scales. Moreover, its factor
structure seems to be fairly stable, in spite of the fact that
To test the incremental variance in attitude to the ad, the fourth dimension of the scale narrowly missed the
if any, accounted for by the CASC scale over and above arbitrarily set eigenvalue cutoff point of one. Three of the
the variance accounted for by the feelings and judgments four dimensions had eigenvalues higher than one and
scales, a two step procedure was used. In the first step, all these explained more than half of the variance in the 16
the eight dimensions of the feelings and judgments scales items.
were used as independent variables with attitude to the ad
as the dependent variable. This resulted in an R square of The feelings scale was the least successful of the three
.484. Next, the three dimensions of the CASC scale were scales with only one factor making a unique and signifi-
also introduced into the equation. This resulted in an R cant contribution to attitude to the ad over and above the
square of .578. Thus, the increase in R square, or the unique contributions of the CASC and judgments scales.
additional variance explained in attitude to the ad by the Further, at least in this study, the factor structure of the
CASC scale was .094 or 9.4 percent. In this final step, only scale was not as expected. Instead of the three factors
one of the three dimensions in the feelings scale and two which were expected, eleven factors with an eigenvalue
of the dimensions in the judgments scale were significant greater than one emerged. The judgments scale fared
(p. < .05) predictors of attitude to the ad. However, all the better with all 25 items loading on five factors, more or
three dimensions of the CASC scale were significantly less according to the specifications of the original formu-
(p. < .05) related to attitude to the ad. Thus, all the lation of the reaction profile for television commercials
dimensions of the CASC scale uniquely contributed to the (Wells, Leavitt, and McConville 1971). The judgments
prediction of attitude to the ad and the scale as a whole also scale also made a sizeable contribution to the explanation
uniquely and substantially contributed to the explanation of variance in attitude to the ad and two of the dimensions
of variance in attitude to the ad over and above the of the scale made unique and significant contributions.
variance accounted for by the feelings and judgments
scale. Overall, it would seem from this study that the 16
items of the CASC scale deserve the attention of those
The significant (p. < .05) standardized coefficients attempting to test the effectiveness of television commer-
(beta) in the last step were .209 (prosocial), .097 (analytic) cials in terms of consumer attitudes. The CASC scale
and -.332 (individualistic) for the three factors in the offers the ability to measure both emotional and rational
CASC scale. The only dimension of the feelings scale responses to advertising in a practical and fairly stable
which was significant (p. < .05) had a beta of -.122. The format. Moreover, it appears to contain unique elements
only two significant dimensions of the judgments scale that are not in the other two scales. Further, even with a
had beta weights of .242 and -.164. Thus, only six of the shorter number of items, it predicts attitude to the ad as
eleven independent variables in the equation made unique well if not better than the other two scales. The CASC
and significant (p. < .05) contributions to the prediction of scale does not claim to measure the entire range of human

emotional experience. However, the simplicity of the sample of subjects and ads. Albeit, the study used only
CASC scale makes it useful for conducting large scale student subjects and that may have been a shortcoming.
testing of emotional and rational responses and advertis- However, since the subjects were viewing the programs
ing effectiveness. The scale can be used to aggregate that they wanted to watch, they were probably the target
responses and assign scores to ads or it can be used to audiences for these commercials and, thus, users or poten-
assess individual level processing of ads. Further, it can be tial users of the products and services in the commercials.
used for ads in both print and electronic media and for a
variety of consumer products. And, finally, the scale can Further testing of the CASC scale versus the other
be used to test ads with vastly different advertising strat- two scales in this study is warranted. First, all three scales
egies since it has previously been validated in a sample of should be tested among “real world” respondents as
240 ads which contained sex, humor, animation, patrio- pointed out above in a limitation of the present study.
tism, status, fear appeals, family situations, celebrities, Future studies should also test the efficacies of all three
animals, typical spokespersons, product demonstrations, scales against important outcome variables other than
comparisons, price appeals, health appeals, etc. (Chaudhuri attitude to the ad. For instance, it may be expected that the
and Buck 1995b). CASC scale will perform better than the other scales with
regard to the formation of attitude to the brand since the
Finally, the CASC scale performed well in the natural analytic dimension of the CASC scale specifically mea-
and “real world” setting of the present study with a large sures responses to brand differentiation within an ad.
ENDNOTE ____________ and ____________ (1998), “CASC – A

Scale for Measuring Emotional and Rational Re-
1
The three scales in the study are not reproduced here due sponses to Advertising,” Zietschrift Fur Sozial Psy-
to space constraints but may be obtained from the chologie, 29 (2), 194–206.
sources cited in this paper. Edell, Julie A. and Marian C. Burke (1987), “The Power
of Feelings in Understanding Advertising Effects,”
REFERENCES Journal of Consumer Research, 14 (December), 421–
33.
Batra, Rajeev and Michael L. Ray (1986), “Affective Haley, Russell I. and Allan L. Baldinger (1991), “The
Responses Mediating Acceptance of Advertising,” ARF Copy Research Validity Project,” Journal of
Journal of Consumer Research, 13 (September), Advertising Research, 31 (2), 11–32.
234–49. Herr, Paul M. and Christine M. Page (2004), “Asymmet-
Burke, Marian C. and Julie A. Edell (1986), “Ad Reac- ric Association of Liking and Disliking Judgments:
tions Over Time: Capturing Changes in the Real So, What’s Not to Like?” Journal of Consumer
World,” Journal of Consumer Research, 13 (June), Research, 30 (March), 588–601.
56–65. Holbrook, Morris B. (1987), “Mirror, Mirror on the Wall,
____________ and ____________ (1989), “The Impact What’s Unfair in the Reflections on Advertising?”
of Feelings on Ad-Based Affect and Cognition,” Journal of Marketing, 51, 95–103.
Journal of Marketing Research, 26 (February), 69– ____________ and R.A. Westwood (1989), “The Role of
83. Emotion in Advertising Revisited: Testing a Typolo-
Chaudhuri, Arjun (2004), “Testing the Independence of gy of Emotional Responses,” in Cognitive and Affec-
Affect Using Prosocial and Reptilian Feelings,” in tive Responses to Advertising, P. Cafferata and Alice
Marketing Theory and Applications, William L. Cron M. Tybout, eds. Lexington, MA: Lexington Books.
and George S. Low, eds. Chicago, IL: American MacInnis, Deborah J., Ambar G. Rao, and Allen M. Weiss
Marketing Association, 15, 286–92. (2002), “Assessing When Increased Media Weight
____________ and Ross Buck (1995a), “Affect, Reason, of Real-World Advertisements Helps Sales,” Jour-
and Persuasion: Advertising Strategies That Predict nal of Marketing Research, 39 (November), 391–
Affective and Analytic-Cognitive Responses,” Hu- 407.
man Communication Research, 21 (3), 422–41. MacKenzie, Scott and Richard J. Lutz (1989), “An Empir-
____________ and ____________ (1995b), “An Explo- ical Examination of the Structural Antecedents of
ration of Triune Brain Effects in Advertising,” in Attitude Toward the Ad in an Advertising Pretesting
Advances in Consumer Research, Frank R. Kardes Context,” Journal of Marketing, 53 (April), 48–65
and Mita Sujan, eds. Ann Arbor, MI: Association for MacLean, Paul D. (1973), A Triune Concept of the Brain
Consumer Research, 22, 133–38. and Behavior. Toronto: University of Toronto Press.

____________ (1990), The Triune Brain in Evolution: Plutchik, R. (1980), Emotion: A Psychoevolutionary Syn-
Role in Paleocerebral Functions. New York: Plenum thesis. New York: Harper and Row.
Press. Stewart, David W. and D.H. Furse (1986), Effective
Mehrabian, A. and J.A. Russell (1974), An Approach to Television Advertising: A Study of 1000 Commer-
Environmental Psychology. Cambridge, MA: MIT cials. Lexington, MA: Lexington Books.
Press. ____________ and S. Koslow, (1989), “Executional Fac-
Mitchell Andrew A. and Jerry C. Olsen (1981), “Are tors and Advertising Effectiveness: A Replication,”
Product Attribute Beliefs the Only Mediator of Ad- Journal of Advertising, 18, 21–32.
vertising Effects on Brand Attitude?” Journal of Zajonc, Robert B. (1980), “Feeling and Thinking: Prefer-
Marketing Research, 18 (August), 318–32. ences Need No Inferences,” American Psychologist,
Muehling, Darrel D. and Michelle McCann (1993), “At- 35, 151–75.
titude Toward the Ad: A Review,” Journal of Current Wells, William D. (1964), “EQ, Son of EQ, and the
Issues and Research in Advertising, 15 (2), 25–58. Reaction Profile,” Journal of Marketing, 28 (Octo-
Olney, Thomas J., Morris B. Holbrook, and Rajeev Batra ber), 45–52.
(1991), “Consumer Responses to Advertising: The ____________, Leavitt Clark, and Maureen McConville
Effects of Ad Content, Emotions, and Attitude To- (1971), “A Reaction Profile for TV Commercials,”
ward the Ad on Viewing Time,” Journal of Consum- Journal of Advertising Research, 11 (6), 11–17.
er Research, 17, 440–53.

Arjun Chaudhuri
Charles F. Dolan School of Business
Fairfield University
Fairfield, CT 06824
Phone: 203.2544000, Ext. 2823
FAX: 203.254.4105
E-Mail: achaudhuri@mail.fairfield.edu

MARKETING COMMUNICATION AND COMPANY BRAND ATTITUDE
Marc Weinberger, University of Massachusetts, Amherst
Dale Taoping Tzeng, University of Massachusetts, Amherst
Paul Bottomley, Cardiff Business School, United Kingdom
Harlan Spotts, Western New England College, Springfield
ABSTRACT corporate reputation, as well as foundation work in brand-

ing, negativity, and branding and media effects.
The value of strong stakeholder brand attitudes to-
ward companies has become increasingly apparent to Branding
practitioners and academic researchers. Less apparent is
the role of the media in the formation of corporate brand The concept of corporate branding has resonance
attitudes. This research draws on key literature from with several disciplines that deal with business and orga-
branding and legitimacy theory to examine the interplay nization theories, but its popularization can be traced to
between the volume and valence of publicity and volume product branding research in the early 1980’s. There is
of advertising as well as the moderating effect of brand considerable agreement that the corporate brand has im-
strength in the formation of company brand attitudes. We portant value that must be nurtured and protected.
employ a unique dataset from three proprietary sources
providing company advertising, publicity and company The parallel but related trend to corporate branding
brand attitudes for a set of 18 technology firms over a 42 has been the emergence of integrated marketing commu-
month time period. The outcome reinforces recent asser- nication with its recognition that product, service, or
tions about the importance of publicity compared to company reputations are formed by many controllable
advertising and also the growing recognition of company and uncontrollable forces. These forces may include adver-
brand strength in tempering media effects. tising, but publicity and public relations may have an
equal or greater impact on brand perceptions.
INTRODUCTION
Public relations activity related to branding has often
The importance of branding and brand equity for been treated by marketers as the step-child of advertising,
products, companies and organizations has become al- but that thinking has been changing in recent years. Rance
most axiomatic over the past decade. Rather than debate Crain (2002), Editor in Chief of Advertising Age, argues
the importance of strong brands, research has focused on that “the landscape of marketing is changing more quickly
the creation, maintenance, preservation and worth of and dramatically than I ever could have imagined, and its
brands. This research focuses on the association between new realities will require public relations to shoulder more
marketing communications (advertising and publicity) of the load. Kitchen (1996) argues that the role of adver-
and company brand attitudes. tising in corporate branding may be diminishing with
marketing public relations (MPR) and corporate public
BACKGROUND relations (CPR) on the rise.” In their book The Fall of
Advertising and Rise of PR, Al and Laura Ries (2002)
Research by Aaker and Jacobson (1994, 2001) and assert that public relations has become the most powerful
Jacobson and Aaker (1987) examined various aspects of marketing-services discipline.
company valuation and branding. Based on firms in
computer-related industries, they concluded that “. . . Legitimacy Theory
company brand attitude, a component and indicator of
brand equity, has value relevance. Not only are changes in The conceptual basis for the importance on company
brand attitude associated with stock return, but this asso- branding is “Legitimacy Theory,” a field that began with
ciation is also incremental to information contained in foundation work by Parson (1960) and Weber (1978). It
accounting measures” (p. 492). Brand attitude is a major has been extended into the organization and strategy
component of brand equity, thus a follow-up question to literature in management (for a review see Suchman
Aaker and Jacobson’s analysis is “what drives brand 1995). “Legitimacy is a generalized perception or as-
attitude?” Our research examines marketing communica- sumption that the actions of an entity are desirable, proper,
tions related antecedents of brand attitude; specifically, or appropriate within some socially constructed system of
volume and valence of publicity, and, the volume of norms, values, beliefs, and definitions” (Suchman 1995,
advertising for firms in the technology industry. The study p. 574). Accordingly, one type of legitimacy is pragmatic
draws on conceptual work from legitimacy theory and legitimacy whose task is to Conform to Demands (respond

to needs, co-opt constituents, build reputations), Select summarize three key findings from social cognition that
Markets (locate friendly audiences), Recruit Friendly Co- involve the impact of volume. The first is a familiarity
Optees, and Advertise (advertise product, advertise image). effect that increased exposure is associated with greater
familiarity and stronger liking (Harrison 1977; Zajonc
Intermediaries that distribute publicity in the media 1968). Research by Lee and Labroo (2004) provides
and advertising by the firm are all potential sources that evidence that when processing fluency of the target is
legitimate the organization. Pollock and Rindova (2003) enhanced by prior exposures, a more favorable attitude is
use the term infomediary to describe the media as a funnel observed (Anand and Sternthal 1991; Bornstein 1989).
that channels and shapes stakeholder knowledge and Second, research by Hawkins and Hoch (1992) suggests
opinion. that simple repetition increases acceptance of a statement.
Finally, the amount of available information about an
The media may play a pivotal role in the branding activity reduces perception of riskiness (Heath and Tver-
process, a view supported conceptually (Gray and Balmer sky 1991). Pollock and Rindova (2003) state, “all else
1997) and empirically (Fombrun and Rindova 1998). To being equal, the combined effects of increased familiarity,
wit, Fombrun and Shanley (1990, 240) observed: “The acceptance, and reduced perceptions of risk can generate
media themselves act not only as vehicles for advertising legitimacy benefits for a firm that receives a higher
and mirrors of reality reflecting firms’ actions, but also as volume of media coverage” (p. 633). Of course, all else is
active agents shaping information through editorials and usually not equal and so the evidence for a volume of
feature articles.” Speaking from the “Legitimacy” per- coverage effect is not universal.
spective, Deephouse (2000) argues that media reputation
is a strategic resource for firms. “A firm’s reputation is In an advertising context, Lodish and his co-authors
produced by the interactions of the firm with its stakehold- concluded that ad weight is not enough to explain overall
ers, and, by information about the firm and its actions sales effects (Lodish et al. 1995). Smaller or newer brands,
circulated among stakeholders, including specialized in- however, do benefit from greater advertising weight.
formation intermediaries (Fombrun 1996; Logsdon and MacInnis, Rao, and Weiss (2002) examined a sample of
Wartick 1995).” He further argues, that “a positive repu- mature brands that did benefit from greater volume of
tation assists the legitimation process is important for advertising and a set that did not. The difference in effects,
competitive advantage because it signals stakeholders however, were driven by message type not the volume of
about the attractiveness of the firm who are then more spending.
willing to contract with it” (Fombrun and Shanley 1990;
Weigelt and Camerer 1988). There are two general effects related to volume of
exposure and valence that may result from the framing of
Pollock and Rindova (2003) examine the connection a message as positive or negative (sometimes referred to
between the media and its influence on market analysts of as tone). Reddy, Swaminathan, and Motley (1998) dem-
IPO’s. In their social constructionist framework, social onstrated that the valence of a critic’s (infomediary’s)
structures like the media enhance the flow of credible reviews are important for Broadway shows. Shows that
information that helps reduces uncertainty in market ex- received extremely low reviews did not last beyond one or
changes. These social structures include specific infome- two weeks before closing. In a study of movie reviews,
diaries or critics such as financial analyst reports, restau- Basuroy, Chatterjee, and Ravid (2003) concluded that
rants reviews, books, movies, and plays (Cameron 1995; both positive and negative reviews are correlated with
Caves 2000; Deephouse 2000; Rindova and Fombrun revenues; just the negative reviews diminish in their
1999). Pollock and Rindova (2003) point out that such impact over time.
infomediaries legitimate firms or products by influencing
buyer or stakeholder desirability perceptions of an orga- Valence has also been shown to matter for negative
nization’s activities. They directly test this connection information from critical sources such as Consumers
between media and legitimacy. “The legitimate organiza- Union when directed at services versus tangible goods
tion [is perceived] not only as more worthy, but also as (Weinberger and Brown 1978). In a political context the
more meaningful, more predictable, and more trustwor- valence of “Ad Watch” coverage of candidate advertising
thy” (Suchman 1995, p. 571). had a significant impact on voter preferences (Min 2002).
Communication Volume and Valence In a publicity context, message valence is an aspect of

media coverage that influences audiences by virtue of the
There are important findings from social cognition choice of topics covered and the interpretation or spin
about the impact of information on impression formation placed on the coverage. “In particular, framing events and
(Fiske and Taylor 1991) and judgment (Heath and Tver- issues in positive or negative terms provides audiences with
sky 1991) that have a bearing on the media effects of visible public expressions of approval or disapproval of
volume and valence of coverage. Pollock and Rindova firms and their actions” (Pollock and Rindova 2003, p. 634).

Negativity and Impression Formation holders temper the media message based on prior experi-
ence formed through direct and/or indirect communica-
Denigrating information about people, products, com- tion with an organization. The effect of new information
panies, or other products abound in the media. The impact is either enhanced or diminished based on this prior
of negative and positive reports on impression formation perception. For the overall sample of companies in
of political candidates (Golan and Wanta 2001), products Wartick’s study, the valence, or tone of coverage, influ-
(Weinberger, Allen, and Dillon 1981; Weinberger and enced change in corporate reputation. For good reputation
Romeo 1989), and job candidates (Bolster and Springbett firms, volume of media exposure was most highly corre-
1961) are well known. The mere mention of an issue on lated with level of reputation. In contrast, valence was
the news makes a story important, possibly memorable, most important for influencing reputation for firms with
and perhaps worthy to be passed along to others (Altheide poor reputations. It is possible that the mere exposure for
1977). Information in the media takes on a form of social companies with neutral or positive reputation are rein-
proof (Rao et al. 2001) leading to information cascades forced by more exposure. Negative coverage, however,
(Bikhchandani, Hirschleifer, and Welch 1992) and avail- may be discounted similar to consumers committed to a
ability cascades (Kuran and Sunstein 1999). The role as brand seem to discount disconfirming brand information.
“infomediary” may result in opinion leadership because
of its inherent social proof. Firms with poor reputation may not have this shield
of protection and may be vulnerable to negative media
Under conditions of uncertainty audiences may coverage. This is the protection that legitimacy theory
consciously or unknowingly imitate media statements suggests positive reputation may provide. Perhaps it eases
rather than form independent judgments (Bikhchandani the uncertainty that stakeholders have while a poor repu-
et al. 1992). Perhaps this helps explain why advertising tation may leave it vulnerable to negatively valenced
weight is more important for new products or small brands messages from the media.
than for mature brands. If simple volume or mere expo-
sure effects prevailed, any kind of exposure in the media RESEARCH QUESTIONS
would be expected to have a positive impact on liking of
the organization or message target. However, Lee and The current study examines the relationship expressed
Labroo (2004) suggest that conceptual fluency is trig- in Figure A looking at the relative impact of publicity
gered when negatively valenced information is presented, (volume and valence) and advertising weight (volume) on
resulting in a less favorable attitude when the valence of brand attitude.
other constructs brought to mind are negative. Like prior
experience with a company or brand, the conceptual The research examines the important issue of relative
fluency effect presumes a more elaborate message effect impact of advertising weight and publicity on company
and less automatic positive frequency of exposure effect. brand attitudes (CBA). Further, the effect of the volume of
stories in comparison to the positive or negative framing
Existing Company Perceptions of the issues is investigated. The first research question
addresses this general issue.
Pollock and Rindova (2003) found that volume and
valence of media reports influenced the market for new RQ1 – What is the relative impact of publicity and
IPO’s (high uncertainty product). Volume affected inter- advertising volume and publicity valence on compa-
est and attention while the valence seemed to affect ny brand attitude?
investor preferences. Here the newness of the offering
could explain the inordinate reliance on media reports. A In keeping with current views about the dominance of
study of consumer response to negative publicity (Ahlu- publicity, the expectation is that publicity will be more
walia, Burnkrant, and Unnava 2000) found that consum- closely related to company brand attitudes than advertis-
ers committed to a brand were not affected as severely by ing. Given that stakeholders are business people in tech-
negative publicity. IPO’s by their general nature may have nology firms, it might be expected that publicity domi-
fewer investors committed to the company, thus being nance should be particularly apparent. Based on the mixed
more influenced by media stories. results of past research, it is unclear whether volume or
valence will dominate in the overall sample.
Company perception expressed as strength of corpo-
rate reputation was examined by Wartick (1992), specif- The second research question (RQ2) addresses the
ically looking at both the volume and valence of media matter of prior brand strength.
coverage. His model suggests that a defined stakeholder
(audience) is exposed to media (amount, tone, and recen- RQ2 – Are the effects of RQ1 tempered by higher and
cy) through a filter defined by source credibility, selective lower levels of prior company brand attitude?
perception, and topic relevance. After exposure, stake-

FIGURE A
Publicity (Volume &Valence) Media Communication
and Company Brand Attitude
Publicity (Volume &

Valance)
Level of
Company Brand
Company Brand
Attitude
Attitude
Advertising
Volume
The expectation is that for strong reputation firms, national advertising expenditure information on over
volume will matter more than valence; for weaker firms, 100,000 different brands. Monthly data were aggregated
valence should be more important than volume. In keep- into quarters over the time period of investigation for each
ing with research showing information effects dependent of the companies included in the study.
on stakeholder perceptions (Wartick 1992; Ahluwalia,
Burnkrant, and Unnava 2000), the analysis accounts for Company Brand Attitude Data
stronger and weaker prior company brand attitudes.
Techtel Corporation provided the company brand
DATA AND DATA ANALYSIS attitude data, which was collected via quarterly surveys of
the personal computing and network computing markets.
This study utilizes data on publicity, advertising, and Their panel surveyed approximately 1500 people influen-
company brand attitude for each of 18 firms, requiring tial in purchasing computer software and hardware (esti-
three separate sources. All data cover the period Janu- mated 50% response rate). The Techtel measure of com-
ary 1, 2000 to July 1, 2003 encompassing 42 months or 14 pany brand attitude asked respondents whether they have
quarters. positive, negative, or no opinion of a company. Like
Aaker and Jacobson (2001), we made use of this informa-
Publicity Data tion to develop a company brand attitude measure defined
as Net Positive Opinion (NPO = percent of respondents
CARMA International provided the publicity data, with positive opinion about the firm –percent of respon-
which are based on a Media Analysis Rating System first dents with a negative opinion about the company). A
developed in 1991. The system rates each article in the simple 3 point scale (positive (1)/neutral (0)/negative (-1)
context in which it appears. Using a pre-determined set of has been used in political science (Robinson, Shaver, and
seven criteria, each article is coded and receives a score Wrightsman 1999) and has been shown (e.g., Haley and
between zero (the least favorable) to 100 (the most favor- Case 1979) to provide information similar to measures
able). All codings begin at 50 (the neutral point), with a with more scale points. For this study we focus on Tech-
possible 50 points added or subtracted from this value. tel’s Enterprise Panel consisting of firms also studied by
CARMA conducts daily tracking of the volume of posi- CARMA, including: AT&T, Cisco, Compaq, Computer
tive and negative press coverage as well as the valence Associates, Dell, EMC, Gateway, Hewlett Packard, IBM,
rating for each article appearing in eighteen major maga- Intel, Microsoft, Oracle, Peoplesoft, SAP, Siebel, Sun,
zines and newspapers in the U.S. For this study daily data Sybase, Unisys.
were compiled into quarterly periods to match the compa-
ny brand attitude data. Analysis of Data
Advertising Data The company brand attitude variable (NPO) is col-

lected from the first quarter of 2000 to the second quarter
Data on media spending were collected from of 2003. Since the advertising data and the publicity data
Ad$pender (2004), a multi-media database providing on some of the firms in the Techtel enterprise data set were

not collected during the exact same period, observations Positive Opinion. It appears that publicity valence and
for companies with any missing values were eliminated advertising spending do not significantly influence atti-
from further analysis. This process reduced the initial tudes toward the company.
dataset from 266 to 166 usable and unique observations.
Research Question 2: Are the effects of Research
Analysis first regressed the total number of favorable Question 1 tempered by higher and lower levels of prior
(volfav) and unfavorable (volunf) articles, the overall company brand attitude?
publicity rating (valence) and the total advertising spend-
ing (advol) on net positive opinion (NPO). Since we have Prior research has established that the effects of
both time series and cross-sectional data, the independent publicity vary depending on prior company attitudes.
variables were checked for multicollinearity, heteroske- Thus, a median split of the sample was used to divide firms
dasticity (Breusch-Pagan) and autocorrelation (Durbin- into groups, Stronger versus Weaker Net Positive Opin-
Watson). No multicollinearity or heterskedasticity were ion. The descriptive statistics for each group are presented
observed among the variables; however, autocorrelation in Table 1. As expected, there is a statistically significant
was a problem, which was corrected with a first order difference between the two groups on Net Positive Opin-
autoregressive process. ion with higher mean opinions for Stronger than Weaker
Companies. While the two groups did not differ on the
RESULTS level of favorable publicity volume, there was a signifi-
cant difference in the average volume of unfavorable
Research Question 1: What is the relative impact of publicity. Firms with Weaker opinion had almost twice
publicity and advertising volume and publicity valence on the volume of unfavorable publicity as firms with Stron-
company brand attitude? ger opinion. Similarly, the valence of publicity was more
positive for Stronger than Weaker firms.
Descriptive statistics for the overall sample of firms
are presented in Table 1. The average Net Positive Opin- A Chow test (Greene 1993) was used to test for
ion for firms in the sample was 50.38. Volume of publicity differences in regression models between the two groups.
varied on average with 66 favorable articles to 43 unfavor- The results of this test (F = 32.27, p < .0001) indicated that
able articles per quarter. The average valence of publicity significant differences exist between the two groups of
was slightly positive with a CARMA publicity rating of firms in terms of the relationship between marketing
51.76. Finally, average ad spending was $3,302,000 per communications and Net Positive Opinion. The models
quarter. differ on volume of favorable publicity (t = 2.25, p < .03)
and publicity valence (t = 3.63, p < .001). The other
Results of the regression model were statistically variables in the model were not significant.
significant (see Table 2), explaining approximately 22
percent of the variation in Net Positive Opinion. After A regression model including the four independent
appropriate corrections for serial correlation, Favorable variables was fit for each group of Stronger and Weaker
Publicity Volume was the only variable influencing Net opinion companies. For Stronger companies the volume
TABLE 1
Descriptive Statistics of Variables Used in the Analysis for Overall Sample of Firms, and
Stronger and Weaker Net Positive Opinion Groups of Firms (per quarter)
OVERALL Stronger Weaker Significance
Net Positive Opinion 50.382 62.485 43.582 < .0001
Favorable Publicity Volume 65.666 63.559 68.417 .6718

Unfavorable Publicity Volume 42.937 31.894 57.354 < .003
Publicity Valence 51.763 53.132 49.977 < .0001
Advertising Spending (‘000’s) $3302.47 $3376.83 $3205.39 .8320
Number of Observations 166 94 72
* t-test results based on df = 164.

TABLE 2
Regression of Ad Spending, Volume of Publicity, and Valence on
Company Brand Attitudes
Variable Coefficient t-value Significance
Intercept 32.2088 2.59 0.01

Favorable Publicity Volums 0.084 3.17 0.002
Unfavorable Publicity Volume -0.003 -0.11 0.913
Publicity Valence 0.268 1.20 0.230
Advertising Spending -0.0002 -1.50 0.1364
Regression Significance F = 11.22 Df = 4, 161 P < .0001 R-sq = .2180
of favorable publicity was the only significant influence where opinions are unformed or neutral. Such a result was
on Net Positive Opinion (see Table 3A). This model not found; no positive advertising effects were detected
accounts for over 50 percent of the variance in the depen- here. This may support MacInnis et al. (2002) who con-
dent variable. tend that for mature brands, it is the type of message and
not the volume of ad spending that matters.
Results for the Weaker opinion companies were
markedly different (see Table 3B). The overall regression A second notable result was that publicity did matter,
model accounted for only 22 percent of the variance in Net with the influence of valence and volume nuanced. In the
Positive Opinion. An examination of the individual coef- overall model the volume of positive publicity accounted
ficients revealed Publicity Valence as the only factor for the variation in company brand attitude. We can see,
having significant influence on Net Positive Opinion. It however, that this effect obscures a significant difference
appears that Volume of publicity and advertising spend- in how publicity works. Wartick (1992) found that com-
ing had no significant influence on company brand atti- panies with stronger versus weaker reputations are differ-
tude for those with Weaker reputations. entially affected by the volume and valence of publicity.
In his analysis as in ours, volume of positive publicity
DISCUSSION correlated with higher ratings for firms with better repu-
tations. These firms were not influenced by publicity
The task of this research was to examine marketing valence per se. In the context of the social cognition
communications related antecedents of company brand research discussed earlier, it is possible that companies
attitude, namely the volume and valence of publicity and with prior positive reputations are reinforced by more
the volume of advertising on a set of firms in the technol- exposure, a mere exposure effect. Perhaps these stronger
ogy industry. In several articles noted earlier, Aaker and perceptions are related to the stronger commitment found
Jacobson argued that company brand attitude is a compo- to insulate audiences from negative information (Ahlu-
nent of brand equity. These attitudes are related to stock walia et al. 2000).
returns and provided information beyond normal ac-
counting measures. Our current research adds to insights Firms with weaker prior attitudes are affected by the
about company brand attitude by examining the role valence of publicity in the media, but not the volume of
advertising and publicity play in the support or weakening positive or negative articles. Again this result is consistent
of brand equity. with both Wartick’s and Ahluwalia’s findings that com-
panies with weaker reputations are affected more by
First, the volume of advertising is not related to brand unfavorable information than the amount of information.
attitude for the full sample of 18 firms; nor for the separate It appears that that firms with stronger prior reputations
examination of firms with stronger or weaker net positive are given a limited shield of protection not afforded to
opinion. This result is consistent with conclusions by firms with weaker reputations.
Lodish et al. (1995) that advertising weight is not suffi-
cient to explain sales effects. In particular they suggested Third, firms with stronger and weaker reputations do
that newer and smaller brands gain the most from adver- have some important differences in publicity coverage.
tising. Though our study did not examine newer or smaller As the descriptive statistics reveal, firms with stronger
brands, firms with weaker company brand attitudes may brand attitudes have far fewer negative stories and a more
be expected to gain from advertising like a newer brand positive publicity rating on average. The levels of positive

TABLE 3A
Regression of Ad Spending, Volume of Publicity, and Valence on Company Brand Attitudes
for Firms with Stronger Prior Company Brand Attitudes
Intercept 19.487 0.93 0.353

Favorable Publicity Volume 0.243 5.92 0.0001
Unfavorable Publicity Volume 0.064 1.13 0.260
TABLE 3B
Regression of Ad Spending, Volume of Publicity, and Valence on Company Brand Attitudes
for Firms with Weaker Prior Company Brand Attitudes
Intercept 0.789 0.06 0.950

Favorable Publicity Volume -0.007 -0.30 0.766
Unfavorable Publicity Volume 0.034 1.54 0.128
publicity and advertising between stronger and weaker Finally, the results give credence to the belief that
firms are not different. The important result is that the public relations activity is gaining ascendance relative to
individual regression models run on the firms with stron- advertising in corporate branding. Here we see the unique
ger and weaker brand attitudes still finds a volume effect role of volume and valence of publicity shaping stake-
for the stronger firms and a valence effect for the weaker holder opinions. It supports the view that the media may
firms. Even though firms with stronger company attitudes play a pivotal role in the branding process. Though
have about the same number of positive articles as do beyond the scope of the current research, prior research
firms with weaker company attitudes, the volume of suggests that this media influence effect is magnified
positive publicity continues to perpetuate the strong rep- under conditions of audience uncertainty where informa-
utation. Firms with stronger prior reputations should tion may reduce riskiness perceptions.
strive to enhance the volume of positive publicity they
receive.. No study is without limitations. The effects observed
in this investigation are clearly related to a small set of
At the same time, firms with weaker reputations have firms within the technology industry for a set period of
an average publicity rating lower than the stronger firms; time. An examination of a larger set of firms across a
however, they scored right at the CARMA neutral public- variety of industries may reveal different relationships.
ity rating of 50. It appears that the variation in valence is However, the results reported here are consistent with
influencing firm reputation. This implies that firms with those of other studies focused on different industries and
weaker reputations are not impacted by more articles, using different measures of company brand attitude, pro-
rather it is the ratings or valence of the articles that is most viding a level of convergent validity. The strong effects of
closely related to the level of company brand attitude. publicity found in this study should not be used to elimi-
Fluctuations in the rating valence of stories about these nate advertising. All that is clear from this study is that the
firms is closely related to brand attitudes for these weaker amount of advertising is not related to company brand
reputation firms. attitudes. As MacInnis and her colleagues suggested, it

may simply be that “what” firms say in their ads is more tation should be viewed as “legitimacy theory” suggests,
important than how “loud” they say it. as a strategic resource. This research lays the groundwork
to examine the broader relationship of the direct and
CONCLUSION indirect effects between marketing communications, com-
pany brand attitude, and company value. In doing so we
The conclusion from this analysis is that publicity is will gain greater insight to the dollar value of marketing
a driver of corporate brand attitude and that media repu- communications on company valuation.
REFERENCES Management, 26 (6), 1091.

Fiske, Susan and Susan Taylor (1991), Social Cognition.
Aaker, David A. and Robert Jacobson (1994), “The Fi- New York: McGraw-Hill.
nancial Information Content of Perceived Quality,” Fombrun, Charles and Mark Shanley (1990), “What’s in
Journal of Marketing Research, 31 (2), 191. a Name? Reputation Building and Corporate Strate-
____________ (2001), “The Value Relevance of Brand gy,” Academy of Management Journal, 33 (2), 233.
Attitude in High-Technology Markets,” Journal of ____________ (1996), Reputation: Realizing Value from
Marketing Research, 38 (4), 485. the Corporate Image. Boston: Harvard Business Press.
Ad$pender (2004), Competitive Media Reporting and _____________ and Violina Rindova (1998), “Reputa-
Publisher’s Information Bureau. New York: TNS tion Management in Global 1000 Firms: A Bench-
Media Intelligence/CMR. Marking Study,” Corporate Reputation Review, 1
Ahluwalia, Rohini, Robert E. Burnkrant, and H. Rao (3), 205–15.
Unnava (2000), “Consumer Response to Negative Golan, Guy and Wayne Wanta (2001), “Second-Level
Publicity: The Moderating Role of Commitment,” Agenda Setting in the New Hampshire Primary: A
Journal of Marketing Research, 37 (2), 203. Comparison of Coverage in Three Newspapers and
Altheide, David (1977), Creating Reality: How TV News Public Perceptions of Candidates,” Journalism and
Distorts Events. Beverly Hills: Sage. Mass Communication Quarterly, 78 (2), 247.
Anand, Punam and Brian Sternthal (1990), “Ease of Gray, Edmund R. and John M.T. Balmer (1998), “Manag-
Message Processing as a Moderator of Repetition ing Corporate Image and Corporate Reputation,”
Effects in Advertising,” Journal of Marketing Re- Long Range Planning, 31 (5), 695.
search, 27 (3), 345. Greene, William H. (1993), Econometric Analysis, 2d ed.
Basuroy, Suman, Subimal Chatterjee, and S. Abraham New York: MacMillan.
Ravid (2003), “How Critical Are Critical Reviews? Haley, Russell I. and Peter B. Case (1979), “Testing
The Box Office Effects of Film Critics, Star Power, Thirteen Attitude Scales for Agreement and Brand
and Budgets,” Journal of Marketing, 67 (4), 103. Discrimination,” Journal of Marketing, 43 (4), 20.
Bikhchandani, Sushil, David Hirshleifer, and Ivo Welch Harrison, A. (1977), “Mere Exposure,” in Advances in
(1992), “A Theory of Fads, Fashion, Custom, and Experimental Social Psychology, L. Berkowitz, ed.
Cultural Change as Informational Cascades,” The Vol. 10. Greenwich, CT: JAI Press.
Journal of Political Economy, 100 (5), 992. Hawkins, Scott A. and Stephen J. Hoch (1992), “Low-
Bolster, B.I. and B.M. Springbett (1961), “The Reaction Involvement Learning: Memory Without Evalua-
of Interviewers to Favorable and Unfavorable Infor- tion,” Journal of Consumer Research, 19 (2), 212.
mation,” Journal of Applied Psychology, 45, 97–103. Heath, C. and A. Tversky (1991), “Preferences and Be-
Bornstein, Robert F. (1989), “Exposure and Affect: Over- liefs: Ambiguity and Competence in Choice Under
view and Meta-Analysis of Research,” Psychologi- Uncertainty,” Journal of Risk and Uncertainty, 4, 5–
cal Bulletin, 106 (2), 265. 28.
Cameron, S. (1995), “On the Role of Critics in the Culture Jacobson, Robert and David A. Aaker (1987), “The Stra-
Industry,” Journal of Cultural Economics, 19, 321– tegic Role of Product Quality,” Journal of Market-
31. ing, 51 (4), 31.
Caves, Richard E. (2000), Creative Industries. Cam- Kitchen, Philip J. (1996), “Public Relations in the Promo-
bridge, MA: Harvard University Press. tional Mix: A Three-Phase Analysis,” Marketing
Crain, Rance (2002), “Advertising and PR as a One-Two Intelligence and Planning, 14 (2), 5–12.
Punch,” in Advertising Age, Vol. July 31. Kuran, T. and C. Sunstein (1999), “Availability Cascades
Deephouse, David L. (2000), “Media Reputation as a and Risk Regulation,” Stanford Law Review, 51,
Strategic Resource: An Integration of Mass Commu- 683–768.
nication and Resource-Based Theories,” Journal of Lee, Angela and Aparna Labroo (2004), “The Effect of

Conceptual and Perceptual Fluency on Brand Evalu- lishers.
ation,” Journal of Marketing Research, 41 (May), Rindova, Violina and Charles Fombrun (1999). “Con-
151–65. structing Competitive Advantage: The Role of Firm-
Lodish, Leonard M., Magid Abraham, Stuart Kalmenson, constituent Interactions,” Strategic Management
Jeanne Livelsberger, et al. (1995), “How T.V. Adver- Journal, 20, 691–710.
tising Works: A Meta-Analysis of 389 Real Work,” Robinson, John P., Philip R. Shaver, and Lawrence S.
Journal of Marketing Research, 32 (2), 125. Wrightsman (1998), Measures of Political Attitudes.
Logsdon, Jeanne M. and Steven L. Wartick (1995), “The- New York: Academic Press.
oretically Based Applications and Implications for Suchman, Mark C. (1995), “Managing Legitimacy: Stra-
Using the Brown and Perry Database,” Business and tegic and Institutional Approaches,” The Academy of
Society, 34 (2), 222. Management Review, 20 (3), 571.
MacInnis, Deborah J., Ambar G. Rao, and Allen M. Weiss Wartick, Steven L. (1992), “The Relationship Between
(2002), “Assessing When Increased Media Weight Intense Media Exposure and Change in Corporate
of Real-World Advertisements Helps Sales,” Jour- Reputation,” Business and Society, 31 (1), 33.
nal of Marketing Research, 39 (4), 391. Weber, Max (1978), Economy and Society. Berkeley:
Min, Young (2002), “Intertwining of Campaign News University of California Press.
and Advertising: The Content and Electoral Effects Weigelt, Keith and Colin Camerer (1988), “Reputation
of Newspaper Ad Watches,” Journalism and Mass and Corporate Strategy: A Review of Recent Theory
Communication Quarterly, 79 (4), 927. and Applications,” Strategic Management Journal, 9
Parsons, Talcott (1960), Structure and Process in Modern (5), 443–54.
Societies. Glencoe, IL: Free Press. Weinberger, Marc G. and Stephen W. Brown (1978), “A
Pollock, Timothy G. and Violina P. Rindova (2003), Difference in Informational Influences: Services vs.
“Media Legitimation Effects in the Market for Initial Goods,” Academy of Marketing Science, Journal
Public Offerings,” Academy of Management Jour- (pre-1986), 5 (000004), 389.
nal, 46 (5), 631. ____________, Chris T. Allen, and William R. Dillon
Rao, H., H. Greve, and G. Davis (2001), “Fool’s Gold: (1981), “The Impact of Negative Marketing Commu-
Social Proof in the Initiation and Abandonment of nications: The Consumers Union/Chrysler Contro-
Coverage by Wall Street Analysts,” Administrative versy,” Journal of Advertising, 10 (4), 20.
Science Quarterly, 46, 502–26. ____________ and Jean B. Romeo (1989), “The Impact
Reddy, Srinivas K., Vanitha Swaminathan, and Carol M. of Negative Product News,” Business Horizons, 32
Motley (1998), “Exploring the Determinants of Broad- (1), 44.
way Show Success,” JMR, Journal of Marketing Zajonc, Robert (1968), “The Attitudinal Effects of Mere
Research, 35 (3), 370. Exposure,” Journal of Personality and Social Psy-
Ries, Al and Laura Ries (2002), The Fall of Advertising chology Monograph Supplement, 9, 1–27.
and the Rise of PR. New York: Harper Collins Pub-

Marc Weinberger
Isenberg School of Management
University of Massachusetts
Amherst, MA 01003
Phone: 413.545.5674
FAX: 413.545.3858
E-Mail: marcw@mktg.umass.edu

ORGANIZATIONAL ANTECEDENTS TO AND OUTCOMES OF
MARKETING STRATEGY DEVELOPMENT STYLES:
A CONTINGENCY MODEL
J. Chris White, Michigan State University, East Lansing
Jeffrey S. Conant, Texas A&M University, College Station
Raj Echambadi, University of Central Florida, Orlando
SUMMARY with control and coordination. Therefore, we propose that

the relationship between the number of MSD styles used
Despite growing interest evidenced by recent publi- and implementation capability is curvilinear (an inverse
cations in the marketing strategy literature and by the U-shaped relationship); that is, the number of MSD styles
Fortune 500 business members of Marketing Science used is positively associated with implementation capa-
Institute (MSI), there is a dearth of research on and a need bility and the number of MSD styles used squared is
to improve our understanding of the process of develop- negatively associated with implementation capability.
ing and implementing marketing strategy. Several impor- We propose that the relationship between MSD styles and
tant gaps in the literature have limited our understanding. implementation capability is contingent on firm size,
First, most existing models of strategy-making fail to fully environmental turbulence, and the organization’s com-
capture the complexity and variety of the phenomena, and petitive strategy emphasis. Finally, we argue that a firm’s
overlook the roles top managers and organizational mem- implementation capability will positively impact its per-
bers play in developing strategy. Second, prior research formance and that the relationship between number of
has focused primarily on either antecedents or conse- MSD styles used and performance will be mediated by
quences of strategy-making, but not both (see Menon et al. implementation capability.
1999 for a notable exception). And third, despite evidence
suggesting the performance outcomes of strategy-making The conceptual model we develop has the potential to
may depend on complex contextual interactions, studies make several important contributions to the marketing
of the implications of strategy-making have focused al- strategy literature. First, empirical support would provide
most exclusively on direct financial payoffs, with incon- evidence that multi-dimensional measures are necessary
sistent results. to capture and better understand the complexity and
variety of the strategy development process. Second,
In this paper, we address these gaps by developing a empirical testing of the proposed model can improve the
model that incorporates organizational antecedents and reliability of the strategy-making scales created by Hart
performance outcomes of marketing strategy develop- and Banbury (1994) and adapted to the marketing context
ment (MSD) styles in a contingency framework. The by White et al. (2003), thus incorporating an organization-
organizational antecedents of MSD styles tap the domain wide approach to the study of marketing strategy develop-
of organizational structure, organizational culture, and ment. Third, the propositions put forth in this paper
top management team characteristics. We propose that the provide guidelines for testing a complex framework in
number of MSD styles is positively influenced by organi- which the relationship between the number of MSD styles
zational formalization, an innovative organizational cul- used and implementation capability is contingent on the
ture, and heterogeneity of the top management team, and interaction of strategy development with organizational,
negatively influenced by organizational centralization. environmental, and competitive strategy factors. And
We draw on resource-based theory, the theory of compet- fourth, we believe empirical testing of the proposed model
itive rationality, and the paradox perspective on organiza- would support the notion that organizations with superior
tional effectiveness to discuss the relationship between implementation capability realize significantly greater
the number of MSD styles used by the firm and the ability performance. In summary, this research has the potential
to implement strategy. Specifically, we argue that even to provide managers and researchers alike with a better
though the use of multiple MSD styles results in benefits understanding of the process of developing and imple-
(e.g., causal ambiguity, barriers to imitation, etc.), it may menting marketing strategy.
also result in additional costs, such as those associated

REFERENCES Model and a Test,” Journal of Marketing, 63 (April),
18–40.
Hart, Stuart and Catherine Banbury (1994), “How Strat- White, J. Chris, Jeffrey S. Conant, and Raj Echambadi
egy-Making Processes Can Make a Difference,” Stra- (2003), “Marketing Strategy Development Styles,
tegic Management Journal, 15 (May), 251–68. Implementation Capability, and Firm Performance:
Menon, Anil, Sundar G. Bharadwaj, Phani Tej Adidam, Investigating the Curvilinear Impact of Multiple Strat-
and Steven W. Edison (1999), “Antecedents and egy-Making Styles,” Marketing Letters, 14 (July),
Consequences of Marketing Strategy-Making: A 111–24.

Chris White
N370 North Business Complex
East Lansing, MI 48824–1122
Phone: 517.353.6381
FAX: 527.432.1112
E-Mail: cwhite@msu.edu

COORDINATING MARKETING AND SALES: EXPLORATION OF A
NEGLECTED INTERFACE
Ove Jensen, University of Mannheim, Germany
SUMMARY unit level, and at the local (country) level. However,

conflicts between M&S are not further explored. Dews-
Academic research has studied marketing’s inter- nap and Jobber (2000) conceptually explore the M&S
functional interfaces in detail (Maltz and Kohli 2000; interface in consumer packaged-goods companies.
Ruekert and Walker 1987a). However, it has almost
completely neglected one interface that is very important We identified two studies that provide empirical
in managerial practice: the interface between marketing information on the M&S interface. Strahle, Spiro, and
units (such as product management, communication, Acito (1996) observe that sales managers do not generally
market research) and sales units (such as the field reps). set sales objectives which are consistent with the strategy
As a result of an MSI workshop on marketing’s interfaces, specified by a marketing executive for a particular prod-
Montgomery and Webster (1997, p. 16) noted that “in- uct. Some of the reasons behind this gap are miscommu-
trafunctional conflict within marketing was a more impor- nications and volume-goal differences. Their results un-
tant topic for discussion than we had expected. The most derline the need for more research on communication as
frequently discussed issue was the conflict between sales well as coordination between M&S. Homburg, Work-
and marketing.” man, and Krohmer (1999) compare the influence of five
functional groups over marketing and non-marketing
Given growing recognition that this interface is high- issues: marketing, R&D, operations, and finance. Howev-
ly conflict-laden in practice, our paper prepares the ground er, their study does neither analyze the extent of conflict
for future empirical research. We develop a conceptual between M&S nor mechanisms to cope with it.
framework of the interface and of coordination mecha-
nisms for coping with it. Our framework defines the key Conceptual Framework
constructs and develops hypotheses.
Our conceptual framework comprises three general
Literature Review components: (1) coordination mechanisms, (2) perfor-
mance domains, and (3) moderators and mediators of the
One of the most important findings of our literature coordination-performance relationship.
review is what we did not find: dedicated empirical
research of the M&S interface. The vast majority of 1. We conceptualize two domains of coordination mech-
studies dealing with organizational issues in marketing anisms for M&S units: boundary-reducing mecha-
have not distinguished between marketing and sales sub- nisms and boundary-bridging mechanisms. These
functions or functions. two domains reflect two facets of integration that
have been distinguished in the literature. For exam-
Among those studies that have dealt with the internal ple, Kahn and Mentzer (1998) differentiate between
structure of the marketing function, one group uses design an “interaction” facet, that focuses on information
dimensions such as centralization, formalization, and exchange, and a “collaboration” facet focusing on the
specialization (Dastmalchian and Boag 1990; Ruekert, willingness to work together. Recent work on coordi-
Walker, and Roering 1985). However, this work does not nation mechanisms increasingly acknowledges their
further differentiate the marketing function into market- differential effects on an interfunctional boundary
ing subunits and sales subunits. and complements the classical coordination mecha-
nisms (Galbraith 1973; Van den Ven, Delbequec, and
Very few articles explicitly deal with the distinction Koenig 1976) by mechanisms targeted at interdepart-
between M&S. The pioneering work of Cespedes (1993, mental beliefs and values (Martinez and Jarillo 1989;
1994) highlights the key challenges at the M&S interface Roth, Schweiger, and Morrison 1991; St. John, Young,
as well as instruments for managing the interface. Work- and Miller 1999). We define boundary-reducing
man, Homburg, and Gruner (1998) develop a typology of mechanisms as the “pull” facets of coordination which
the structural location of M&S and identify environmen- foster the willingness to cooperate, thereby diminish-
tal antecedents influencing the configuration chosen. M&S ing the interdepartmental barriers. Among these are
subunits are found at the corporate level, at the business integrative recruiting and HR development as well as

integrative reward systems. On the other hand, we 3. Our moderators and mediators are centered around
define boundary-bridging mechanisms as the “push” the boundary between M&S. By boundary criticali-
facets of coordination which drive two departments ty, we understand the amount of coordination needed
to cross the boundary, but do not decrease the inter- at that interface. The greater the criticality, the more
departmental barrier. These include formalization, traffic occurs at the boundary. Boundary criticality is
integration teams, joint planning, and information constituted by such constructs as task interdepen-
systems. dence and market dynamism. Boundary-creating dif-
ferences refer to the extent of barriers hampering
2. Our framework contains two domains of success. coordination. Among these are differences in goal
The first domain encompasses success at the busi- orientation, time orientation, and skills. The third
ness level. We distinguish between market-level domain, integrativeness of the corporate environ-
measures and financial measures. The second do- ment, pertains to the extent to which M&S are embed-
main represents the integration success at the depart- ded in an intraorganizational context that fosters
mental level. Here, we subsume such constructs as coordination between two units. Most importantly,
coordination efficiency, coordination effectiveness, this includes top management support. References
communication intensity, information quality, and available upon request.
information usage.

Christian Homburg
Director of the Institute for Market-Oriented Management
68131 Mannheim
Germany
Phone: 49.621.181.1555
FAX: 49.621.181.1556
E-Mail: prof.homburg@bwl.uni-mannheim.de

THE EFFECT OF INTERACTIONAL JUSTICE ON THE PERFORMANCE
OF CROSS-FUNCTIONAL PRODUCT DEVELOPMENT TEAMS
Tianjiao Qiu, University of Illinois at Urbana–Champaign, Champaign
Deborah Rupp, University of Illinois at Urbana–Champaign, Champaign
William Qualls, University of Illinois at Urbana–Champaign, Champaign
SUMMARY managers impact their task performance and interper-

sonal citizenship behaviors.
In recent decades, cross-functional product develop-
ment teams (CFPDTs) have become a popular mechanism Interactional justice refers to the extent to which team
for achieving greater interfunctional integration in the members feel they are treated with dignity and respect by
new product development process. Research has demon- their project managers. Interactional justice is particularly
strated that CFPDTs composed of members from various relevant in the study of CFPDT members’ performance
areas such as marketing, product design, process design, because it reflects the interpersonal dimension of the new
engineering, and finance work as an efficient mechanism product development process. The quality of interperson-
for generating innovative new product ideas (e.g., Sethi al treatment received during the product development
2000; Wind and Mahajan 1997). A key characteristic of process from the project manager greatly influences team
CFPDTs is the diversity of functional areas represented in members’ fairness judgment and predicts supervisor-
the team. On one hand, cross-functional diversity pro- related outcomes. Three aspects of supervisor-related
vides for crucial functional inputs during the new product outcomes are examined: team members’ multi-foci com-
development process. Successful cross-functional rela- mitments, their task performance, and their interpersonal
tionships ensure a high level of information sharing and citizenship behaviors.
integration, which in turn leads to innovative new prod-
ucts. On the other hand, cross-functional diversity poses In the paper, we develop and test an integrative model
great challenges to efficient working relationships, re- relating interactional justice perceptions to CFPDT per-
flected in three potential barriers – turf barriers, interpre- formance (see Figure 1). We hypothesize that higher
tive barriers, and communication barriers (Hunt 1995). levels of supervisor-focused interactional justice climate
enhance the interpersonal relationships between team
To benefit from cross-functional diversity, a high members and project managers. Such enhanced relation-
level of interaction and collaboration among CFPDT ships are reflected in team members’ multi-foci commit-
members is necessary. Since each team member in CFP- ments to the team and to the product development project.
DTs is equipped with unique experience with the existing Furthermore, team members’ multi-foci commitments
product technology or manufacturing process, it is essen- serve as a mediator between the perceived interactional
tial that he/she communicates effectively in order to create justice and their task performance as well as their interper-
the synergy that CFPDTs provide. Without such informa- sonal citizenship behavior.
tion sharing and integration, team members cannot devel-
op a common understanding of or a shared goal regarding Members from 50 students’ product development
the product development process. teams participated in the study. Student teams were com-
posed of majors from the College of Engineering and the
One factor that can potentially influence the level of Business school, which emulate the structure of cross-
cooperativeness among CFPDT members is their fairness functional product development teams. The model is
perceptions. Fairness perceptions have a significant im- tested using structural equation modeling. The overall
pact on a wide variety of job-related attitudes and behav- model is strongly supported by the data, with satisfactory
iors (e.g., Cohen-Charash and Spector 2001; Colquitt goodness-of-fit statistics as follows: χ 2(58)=83.42 (p =
2001; Colquitt et al. 2001). The literature shows that .02); RMSEA = 0.047; GFI = 0.937; CFI = 0.985; NFI =
employees make at least three types of fairness judg- 0.951; AGFI = 0.902; and RMR = 0.036. All estimated
ments: distributive, procedural, and interactional. Using beta parameters for the model are significant at 0.05
the reward structures as the key variables, Shikhar and levels.
Vijay (2001) have examined the performance of CFPDT
through distributive and procedural justice perspectives The findings indicate that the relationship between
without mentioning the role of interactional justice. To team members and the project manager is enhanced by
bridge the research gap, we examine how CFPDT mem- higher levels of perceived interactional justice. Such
bers’ interactional justice perceptions about their project enhanced relationship is reflected in team members’ multi-

FIGURE 1
The Conceptual Model of Interactional Justice on CFPDT Members’
Task Performance and Interpersonal Citizenship Behavior
Task
Performance
Interactional
Justice
Person-Focused
Interpersonal
Citizenship
Behavior
Multi-Foci
Commitment
Task-Focused
Interpersonal
Citizenship Behavior
foci commitments towards the team and the on-going contributes to the better understanding of the role of
project. It is further demonstrated that multi-foci commit- interactional justice perception in enhancing interfunc-
ment serves as the partial mediator between the perceived tional communication and cooperation in the new product
interactional justice and both team members’ task perfor- development process.
mance and interpersonal citizenship behavior. The paper

Tianjiao Qiu
Department of Business Administration
339 Wohlers Hall
University of Illinois at Urbana–Champaign
1206 S Sixth Street
Champaign, IL 61820
Phone: 217.333.4240
FAX: 217.244.7969
E-Mail: tqiu@uiuc.edu

CROSS-FUNCTIONAL INTEGRATION AND NEW PRODUCT
PERFORMANCE: A META-ANALYSIS
Tanawat Hirunyawipada, University of North Texas, Denton
Archna Vahie, University of North Texas, Denton
SUMMARY Second, the decisions to include or exclude any

specific characteristics into the samples always impact the
Literature has emphasized the importance of the relation reliability of measures. The results show that the larger the
between cross-functional integration (CFI) and the suc- number of functions integrated, the more difficult it will
cess of new products (e.g., Anderson 1982; Nystrom be for CFI to become effective. The results confirm the
1985; Ruekert and Walker 1987a, 1987b). Many empiri- importance of the inclusion of the R&D function in the
cal and theoretical studies confirm a positive relation CFI team for the success of NPD. The inclusion of the
between CFI and new product performance (e.g., Gupta, marketing function in CFI shows neutral impact on the
Raj, and Wilemon 1986; Griffin and Hauser 1992, 1996; CFI-new product performance relation. This may imply
Song and Parry 1997a, 1997b). Yet other studies show that the importance of the marketing function be made
either non significant or negative effect of CFI on the clear before it is included in the CFI. Future research may
success of new product development (e.g., Zirger and investigate which other function should be integrated for
Hartley 1996; Henard and Szymanski 2001; Gray, Matear, the success of new product. The meta-analysis also shows
and Matheson 2002). These disparate findings in the that the respondents from NPD teams have positive im-
literature make the indispensability of CFI in new product pacts toward effect size. This suggests that the NPD team
development (NPD) a questionable concept. We use meta- members, which are the primary source of data, represent
analysis to provide a clearer answer to the importance of the actual information of the CFI-new product success
CFI for the success of new products. relation better than do the respondents from senior man-
agements. However, this result serves as a warning that
Employing the model-level effect sizes, we found people who are involved in CFI may provide biased
108 correlations between CFI and new product perfor- information about the CFI-new product success relation.
mance. Findings from the meta-analysis show that the Researchers may consider using data from different infor-
relation between CFI and new product success is positive- mants for independent and dependent variables to prevent
ly correlated, but that some key factors moderate this common method bias. Interestingly, the authors find that
relation. This meta-analysis provides three main findings. studies published in the year 2000 and thereafter show
First, attention must be paid to the operationalization of considerably positive relation with effect size. This result
the constructs because the way the dimensions of the sheds some light on why Henard and Szymanski (2001)
construct are defined can impact the findings. We find that found no significant relation between CFI and new prod-
CFI is likely to be comprised of at least two distinct uct success as most of the extant studies they looked at
dimensions: internal integration and inter-functional cli- were published prior to year 2000. This finding supports
mate. Internal integration captures the concept of CFI in our expectation that the increasing needs for diverse
terms of its physical interactivities between functions expertise and experiences can lead to more integration of
whereas inter-functional climate is more likely to capture different functions, and therefore to greater outcomes.
the norm within group or organization that encourages the
trust and relationship among different functions. Inter- The generalization of the results is limited because of
functional climate can increase the capability to resolve the moderating effects of contextual variables. Studies
conflicts among integrated teams. The result from the with respondents from companies across multiple coun-
meta-analysis indicates that CFI has stronger relation with tries do not support the salient relation between CFI and
new product success when the CFI is defined as the new product success. CFI may not be able to impact
supportive climate for collaboration. This issue casts a consistently across countries with different norms and
doubt on whether researchers have crafted enough dimen- cultures. Thus, researchers conducting CFI studies should
sions to reflect the underlying concept of CFI. The results take the context, organizational environment, and culture
also show that the CFI-new product success relation is into consideration. Accordingly, different cultures and
stronger when the new product performance is operation- societies can be studied to create a typology of norms that
alized as product effectiveness than when it is operation- are best suited for CFI. Future research may focus on how
alized as NPD productivity. This suggests that respon- multicultural the individual NPD teams are and what
dents view a potential outcome of CFI as product effec- impacts this characteristic has on the success of new
tiveness rather than productivity. products. The result also shows that samples from high-

tech products appear to play a moderating role on the effect size, and that CFI performs different roles across
effect size. NPD’s time constraint and technology conver- stages of NPD. A future implication is to research not only
gence are among the reasons why NPD team in high-tech the impact of CFI in various stages of NPD, but also the
industry is more likely to seek integration with other factors and conditions that impact the CFI strategy in each
functional areas. The samples from western countries also NPD stage. Not surprisingly, goods vs. services as well as
show stronger negative impact on effect size than do industrial products vs. non-industrial products do not
samples from other countries. This result can be explicat- explain the variance of the effect size. Given that the
ed by the individualistic norm of people in western coun- overall combined mean is significant, we conclude that
tries. Workforce in these countries is less likely to commit CFI is important for NPD irrespective of the nature of the
to rigid group norm or CFI team. The result also implies types of products.
that CFI in development stage has negative impact on the

Tanawat Hirunyawipada
P.O. Box 311396
Denton, TX 76203
Phone: 940.565.3120
FAX: 940.565.3837
E-Mail: Hirunyat@unt.edu

THE CONTINGENT EFFECT OF PRODUCT QUALITY ON NEW
PRODUCT PERFORMANCE: A CONCEPTUAL MODEL
Kwaku Atuahene-Gima, City University of Hong Kong, Hong Kong
Gloria Barczak, Northeastern University, Boston
SUMMARY pleski et al. 1993; O’Neal and LaFief 1992). Thus, accep-
tance of marketing’s key role in establishing product
Many firms have instituted quality programs in an quality implies a contingency view of product quality
effort to upgrade and continually improve the quality of (Varadarajan and Jayachandran 1999). However, there
their products for the ultimate purpose of creating com- has been little, if any, research on if and how environmen-
petitive advantage (Clark and Fujimoto 1991; Menon tal factors moderate the relationship between product
et al. 1997). This emphasis on product quality is well- quality and performance.
deserved as there is strong evidence that product quality
is an important determinant of firm performance (Buzzell Since product quality is a deliberate firm strategy to
2004; Hildebrandt and Buzzell 1991; Jacobson and Aaker differentiate one’s offerings from its competitors’ (Jacob-
1987; Phillips, Chang, and Buzzell 1983; Sethi 2000). son and Aaker 1987; Phillips, Chang, and Buzzell 1983;
The positive impact of product quality on performance Porter 1980; Varadarajan and Jayachandran 1999), we
(Buzzell 2004; Hackman and Wageman 1995; Hilde- propose that the effectiveness of such a strategy will be
brandt and Buzzell 1991; Jacobson and Aaker 1987; influenced by the nature and dynamics of its environment
Phillips, Chang, and Buzzell 1983; Sethi 2000) is of (internal and external). This proposition resonates with
particular interest because it suggests a direct effect rela- Varadarajan and Jayachandran’s (1999) argument that
tionship. Research has investigated the direct effect of the “market performance outcome of a business’s deci-
product quality on various performance dimensions such sion to offer products of a high quality is contingent on . . .
as market share and profitability (Buzzell 2004; Hilde- consumers’ characteristics and competitors’ reactions”
brandt and Buzzell 1991; Phillips, Chang, and Buzzell (p. 130).
1983), price (Jacobson and Aaker 1987; Phillips, Chang,
and Buzzell 1983) and costs (Jacobson and Aaker 1987; We define product quality from a customer, rather
Phillips, Chang, and Buzzell 1983). Though the evidence than a supplier, perspective (Bounds et al. 1994; Kordu-
is substantial, uncritical acceptance of this result suggests pleski et al. 1993; Morgan and Piercy 1998). A customer-
that product quality defies the logic of contingency theo- focused approach views quality as the customer’s percep-
ry. Such a view is subject to theoretical skepticism given tion of how well a given product meets their needs and
that a variety of marketing strategies (e.g., market orien- expectations (Kordupleski et al. 1993; Morgan and Piercy
tation, innovation, strategic alliances) have been found to 1998; Parasuraman, Zeithaml, and Berry 1985). Put an-
follow contingency theory. other way, quality is “simply conformance to [the] re-
quirements” (O’Neal and LaFief 1992) of the customer.
Practically, uncritical acceptance of the direct effect Although product quality may influence new product
of quality on performance down plays the role of market- performance directly, it is possible that external market
ing in ensuring market performance. The importance of factors may help or harm that relationship (Varadarajan
the marketing function in building product quality is well- and Jayachandran 1999). Our theoretical arguments sug-
established (Cravens et al. 1988; Kordupleski et al. 1993; gest that when technological uncertainty is high, firms
O’Neal and LaFief 1992). In fact, it has been argued that may need to commit to certain aspects of the technology
marketing, as the most customer-focused function (Kor- before they know and understand the ramifications of
dupleski et al. 1993), should play the lead role in establish- their actions. The level of quality based on those actions
ing product quality (Kordupleski et al. 1993; O’Neal and is likely to be rendered inferior by new technological
LaFief 1992) as its job is to understand customer needs developments which may have a negative effect on prod-
and translate those needs into products with high custom- uct quality. Similarly, in situations of high market uncer-
er value (Cravens et al. 1988; Kordupleski et al. 1993; tainty, project teams need to monitor market shifts and
O’Neal and LaFief 1992). Creation of value through high adjust product specifications accordingly (Menon et al.
product quality depends on marketing skills in analyzing 1997). Since it takes time for firms to adjust the level of
and determining customer needs, competitor analysis, quality to meet both rapidly changing customer needs and
identifying appropriate target segments, product position- competitive actions, the positive effect of product quality
ing, communication, pricing, distribution, and speedy on new product performance is likely to diminish. Greater
implementation of launch (Cravens et al. 1988; Kordu- customer demandingness may indicate that customers are

not satisfied with existing products (Li and Calantone customers and perceptions of value will be determined
1998), suggesting a lower level of competition for a (Buzzell 2004; Hildebrandt and Buzzell 1991; Jacobson
specific product of high quality. Hence, product quality and Aaker 1987; Phillips, Chang, and Buzzell 1983).
will be more positively related to new product perfor-
mance under high level of customer demandingness. Our model makes practical and theoretical contribu-
Finally, as the hostility of a new product introduction tions. From a practical perspective, we focus on under-
increases, so does the strength of competitors’ reactions standing the effect, if any, of specific internal environ-
(Heil and Walters 1993; Heil and Robertson 1991; Moore mental variables with product quality. Because these
1992). Effective competitor responses suggest an erosion internal variables can be influenced and/or controlled by
of the competitive advantages of the focal firm’s new managers, the findings of this study should provide useful
product. recommendations for enhancing product quality and ulti-
mately, new product performance. It is possible that some
Regarding internal moderators we argue that a prod- of these variables may have an adverse reaction with
uct concept which is familiar to the organization will be product quality. Thus, our findings can provide insight as
easier, less time consuming and less expensive to design, to which variables enhance or weaken product quality and
develop, and market. Hence, a project team is more thus, product performance. In terms of theory, an impor-
capable of maintaining the level of quality than a product tant contribution of this study is its extension of the
concept high new to the firm. Further, in a new market, previous research on product quality to include potential
where customers know little about how to evaluate a moderators of the relationship between product quality
product’s attributes and benefits, product quality may and performance. Unlike existing research, this study
have greater positive effect on performance because the draws on contingency theory and argues that the impact of
nature of the product “may be able to influence how product quality on product performance depends on par-
attributes are valued [and] define the ideal attribute com- ticular internal product and strategy characteristics as well
bination” (Kerin, Varadarajan, and Peterson 1992, p. 35). as external market characteristics. Another major contri-
In other words, a quality new product in a new market may bution of this study is its focus on specific product
define the product category and the nature and level of outcomes as opposed to most prior research that examines
quality expectations (Kerin, Varadarajan, and Peterson firm-level aggregate outcomes such as the quality of a
1992). Finally, a quick implementation implies that prod- firm’s products relative to competitors (Clark and Fujim-
uct, price, distribution and promotion tactics and activities oto 1991; Menon, Jaworski, and Kohli 1997; Morgan and
are operational simultaneously, at full strength, over a Piercy 1998). By focusing on product quality at the
relative short period of time. The faster the implementa- individual product level, this study can provide insights
tion of a specific product’s marketing strategy, the more about factors that interact with quality and the differential
likely that the product will be known and accessible to effects they have on product performance.
For further information, contact:

Kwaku Atuahene-Gima
Department of Management
City University of Hong Kong
Hong Kong
E-Mail: Mgkwaku@cityu.edu.hk

CONSUMER EXPERIENCE OF SOCIAL POWER DURING SERVICE
CONSUMPTION: AN EXPLORATORY STUDY
Kalyani Menon, Wilfrid Laurier University, Waterloo
Harvir Bansal, Wilfrid Laurier University, Waterloo
ABSTRACT it arises during service consumption. The objective is to

initiate an in-depth understanding of the nature and com-
A survey derived scripts of consumer experiences of ponents of consumer experiences of powerfulness and
powerfulness and powerlessness during service consump- powerlessness during service consumption. It is hoped
tion. Most experiences occurred in high contact services, that such an understanding will enable service managers
underlining the social nature of consumer power. Power- to develop tools and strategies to enhance service experi-
fulness and powerlessness were linked with distinct ante- ences. The following section draws on the social psychol-
cedents and consequences that differ from non-consump- ogy literature to describe the concept of power, what is
tion situations, and powerfulness appears to be a more known of the experiential aspects of power and its impli-
complex experience than powerlessness. cations for consumer behaviour. We then present our
empirical study of the subject. A survey of consumers of
INTRODUCTION a range of services was conducted and qualitative and
quantitative data collected on their experience of power
As repeatedly documented by consumer researchers, and the lack of it. The data were coded and analyzed to
the social interaction between consumers and service develop scripts of consumer powerfulness and powerless-
providers is among the most significant determinants of ness. Such scripts map the mental representation of rele-
consumer experience in the service (Bitner, Booms, and vant experiences (Fitness and Fletcher 1993; Shaver et al.
Tetrault 1990; Keaveney 1995; Menon and Dubé 2000). 1987) and may be crucial to understanding the dynamics
Service satisfaction and dissatisfaction depends greatly of consumer – provider interaction.
on how the provider relates to consumers, and how they in
turn perceive the provider. SOCIAL POWER AND CONSUMER – SERVICE
PROVIDER INTERACTIONS
Research on service provider – consumer dynamics
has looked at it primarily from a relationship marketing Social power has been defined as the control pos-
perspective, to understand benefits to consumers and sessed by an individual over another and refers to attempts
firms due to long term relationships with service providers to enhance and protect differentiation of the individual
(e.g., Gwinner, Gremler, Bitner 1998). These studies have (Depret and Fiske 1993; Moskowitz et al. 1994). As an
focused on various examples of interpersonal behaviour extension, we argue it may be thought of as the ability to
such as empathy (e.g., Butcher, Sparks, and O’Callaghan control another’s outcome. High power and low power
2003), friendship (Price and Arnould 1999) and affective have been conceived as polar opposites with high power
commitment (Shemwell, Cronin, and Bullard 1994). So- being manifest in frequent behaviours of dominance and
cial psychologists stress that these behaviours are exem- assertiveness and low power expressed by submissive
plars of two key meta-constructs of power and commun- acts. Power dynamics are thought to permeate most social
ion that underlie most social interactions and map the relationships, whether explicitly or implicitly. This con-
various modes of relating to the world. (Moskowitz et al. cept of power and control may occur explicitly in services
1994). Understanding consumer-service provider inter- such as the control a doctor has over a patient, or a banker
action and the impact on the consumption experience has over a mortgage application. While the power equa-
requires we go beyond examples of behaviour to examine tion in such transactions is clearly weighted in favour of
how the two meta-constructs of power and communion the service provider, there may be several other service
occur in services where providers and consumers are situations where the power equation may not be so obvi-
required to interact during service delivery and consump- ous. For instance, a restaurant patron has power in so far
tion. To our knowledge, no research has thus far adopted as their patronage adds to the business of the restaurant
such a theoretically anchored approach to the study of the and of the individual waiter. But the restaurant and the
parameters of consumer-service provider interaction. waiter in turn can affect the outcome of the diner’s eating
experience, and thus may have some power as well. There
This paper presents an initial investigation of the may also be a range of services where the power equation
meta-construct of power from a consumer perspective as is mostly absent such as services at the post-office, routine

banking, and other such services where the consumer’s perceptions of their power in a service situation? Finally,
and/or the provider’s ability to control the outcome may with the objective of examining the impact of power on
be relatively irrelevant. consumer service experience, we also examine how per-
ceptions of two marketing outcome variables – service
While initial research examined individual level fac- quality and consumer satisfaction – vary due to the power
tors such as gender as sources of social power (Ridgeway experience. The following section presents a survey used
1992), more recent research stresses the interaction be- to capture the components of consumer experiences of
tween individual and situation level variables. Gender by power – antecedents, cognition, expectations, emotions,
itself may not impact social power quite as much as the and expression – as well as consumer evaluations of
gender composition of the social group with greater service quality and satisfaction when they are powerful or
stereotypical gender-power relationship in same gender powerless in a service context.
interactions than in opposite gender interactions (Macoby
1990; Moskowitz et al. 1994). Continuing with the idea METHOD
that power is a function of the interaction between a
person and the situation, social and professional roles Following the method used by previous research on
have been shown to impact power, For example, individ- developing scripts of interpersonal dynamics (Fitness and
uals express more power in the role of a supervisor but less Fletcher 1993; Shaver et al. 1987), we used a self-admin-
in the role of a supervisee (Moskowitz et al. 1994). The istered survey with a mix of qualitative and quantitative
first question of interest to this paper is what elicits service measures. Research has shown that repeated experiences
consumer experiences of powerfulness and powerless- render people highly knowledgeable about various com-
ness? Are they mirror images of each other e.g., does the ponents of their experiences and that it is possible to elicit
absence of what elicits powerfulness elicit powerlessness memory representations or scripts of these experiences
and vice versa? (Fitness and Fletcher 1993; Shaver et al. 1987). Typically,
researchers provide individuals with a series of prompts
This paper also examines the experiential conse- designed to elicit retrospective reports of the experience,
quences of power and lack of power on consumer cogni- and scripts are developed using the most frequently men-
tion and expectations, emotions, and emotion expression. tioned features of the experience.
The cognition and expectations of individuals experienc-
ing power or the lack of it tend to have specific biases since Sample and Procedure
powerful individuals are motivated to defend their power
and the powerless to compensate for their lack of it A convenience sample of adult consumers of services
(Richeson and Ambady 2002). Thus, in order to preserve (n = 138; 54% male, 46% female) were approached on a
their power, powerful consumers may have an exaggerat- university campus and randomly assigned to one of two
ed conception of their role and ability in the service, or versions of the survey questionnaire – either the version
may have low expectations of others in the setting. Pow- asking them to recall an experience of powerfulness (n =
erless individuals may focus their attention and cognitive 60) or one of powerlessness (n = 78) during consumption
resources on monitoring others in the social setting since of any type of service. The respondent pool was mainly
they perceive themselves as being dependent on others. composed of the members of the university community
such as administrative staff, employees of local business-
In addition to its impact on cognition, powerfulness es, and some faculty. We did not approach any undergrad-
and powerlessness may impact emotion experiences and uate students. Majority of the respondents were between
expression. According to the interpersonal view of emo- the ages of 26 and 50 (approx. 67%), 42 percent had
tions, emotions emerge and are defined during interaction completed an undergraduate degree, 25 percent were
between individuals (Parkinson 1995). There are well- graduates, and 14 percent held post-graduate degrees.
established norms about the appropriateness of emotions
relative to one’s power and emotions such as anger and In each version of the questionnaire, respondents
joy are typically experienced when one is in a powerful were instructed to recall and report the actual experience
situation while emotions such as anxiety and sadness are of powerfulness/powerlessness in as much detail as pos-
experienced in powerless situations. Further, the extent to sible. Powerfulness and powerlessness was defined for
which individuals express emotions may also be impacted the respondents as control/certainty and lack of control/
by their experience of power with powerful people being certainty respectively in a service context. The question-
more willing to express their emotions than the powerless naire began with demographic questions and was fol-
(Conway et al. 1999). Thus the second research question lowed by a series of prompts to aid recall of a relevant
of interest to this paper is: How does consumer percep- experience of powerfulness or powerlessness (e.g., what
tions of power or the lack of power in a service context happened, when, who was present, etc.). This was fol-
manifest itself? What kind of cognition, expectations, lowed by measures for the dependent variables.
emotions, and expressions occur as a result of consumer

Dependent Variables RESULTS
Open-ended questions asked respondents to provide Antecedents

details of the antecedents of powerfulness/powerlessness
(what happened to make them feel powerful/powerless), It is noteworthy that the majority of the experiences
specifically the type of service setting in which the event recounted occurred in high contact services (retail, pro-
occurred, gender of the individuals present at the time of fessional) rather than in low contact services (miscella-
the event, and details of the eliciting event. neous services such as online, cable) supporting the
contention that power is intrinsically social in nature and
Open-ended questions then captured the experiential arises in the context of interpersonal interaction. Results
consequences of consumer thoughts and expectations. indicate that while the majority of the powerless experi-
Respondents were asked to describe their thoughts, and ences (54%) occurred in the context of professional ser-
their expectations of the service provider as they experi- vices, the majority of powerful experiences occurred in
enced powerfulness/powerlessness during the event de- the context of either retail services such as computer
scribed. Multi-item 5 point scales (not at all – strongly) stores, restaurants (48%) or professional services such as
drawn from past research (Richins 1997) measured expe- financial services, legal services, and government servic-
riential consequences of some major consumer emotions. es (45%). In keeping with past research, while consumers
Anger was measured by 5 items (anger, frustration, irrita- were almost equally likely to feel powerful in mixed
tion, outrage, hostility; α = 0.94), anxiety was measured gender and same gender groups, powerlessness was expe-
by 3 items (nervousness, worry, anxiousness; α = 0.79), rienced primarily in mixed gender groups. The mixed
shame was measured by 3 items (shame, embarrassment, gender groups involved either a service provider or other
guilt, α = 0.71) and happiness was measured by 3 items consumers of the opposite gender in close proximity
(joy, excitement, happiness; α = 0.85). The experiential during the event.
consequences of emotion expression was measured by a
single item 7 point scale (not at all – very much so) asking Interesting differences arose in terms of the nature of
respondents to indicate the extent to which they expressed the eliciting events for powerful and powerlessness. Pow-
their emotions using hand and body gestures, facial or erfulness occurred primarily when consumers perceived
vocal expressions. they had high levels of knowledge in the service setting
(67%). This knowledge could relate to knowledge of the
The marketing outcome variables were all measured nature of the service (e.g., knowledge of financial instru-
by multi item scales. Service quality was measured by a 4 ments a consumer wishes to purchase, knowledge of
item (employee professionalism, quality of information computer hardware in a retail store) or to knowledge of the
provided, care of customers, overall service quality; α = script of the service encounter (e.g., consumers knew the
0.95) 5-point scales (poor – excellent). Consumer satis- process to file a complaint, knowledge of the various
faction was measured by a 2-item 7-point scale (very phases of consumption in a high-end restaurant). On the
dissatisfied – very satisfied; very disappointed – very other hand, while 32 percent of powerless experiences
delighted, α = 0.94) adapted from Oliver (1997). occurred despite consumer perceptions of knowledge, the
majority of powerlessness occurred due to service failures
Data Analysis such as service delays, pushy service provider, and being
taken advantage of (68%). Powerless consumers did not
The qualitative responses dealing with the anteced- report lack of knowledge as a cause for their experience,
ents, cognition, and expectations were scrutinized and thus indicating that powerlessness, rather than being the
coded by three coders working independently to reflect mirror image of powerfulness as documented by past
underlying dimensions and/or categories. After careful, research, may be a distinct experience.
repeated readings, the data was sorted in to groups reflect-
ing underlying similarities. For instance, when coding Experiential Consequences
responses to the question of the eliciting event, data about
delayed service and condescending service provider were Turning now to the experiential consequences of
assigned to the category of service failure. We ensured powerfulness and powerlessness, powerfulness was linked
that the data assigned to each category was more similar almost equally with provider oriented (49%) and action
to each other than to data assigned to other categories. oriented thoughts (43%). For example, an action oriented
Conflicts were resolved by majority vote. The full list of thought may concern a consumer thinking of, “how to get
categories is presented in Table 1. Aggregate scores were it [laptop under repair] back ASAP,” or being “better
created for the quantitative measures. Chi square tests and aware,” so as to “to make a better decision.” Provider
difference of means were conducted to develop scripts of oriented thoughts took the form of a consumer thinking of
powerfulness and powerlessness. “being ripped off as a consumer [by the service provider]

TABLE 1
Frequencies and Mean Values for Episodes of Powerfulness and Powerlessness.
Powerful Powerless
N = 60 N = 78
Antecedents
Service setting (% frequencies)

χ2 = 14.33 (p < 0.05)
Retail 0.48 0.21
Professional 0.45 0.54
Miscellaneous 0.07 0.25
Gender composition (% frequencies)
χ2 = 4.83 (p < 0.05)
Mixed 0.43 0.62
Same 0.57 0.38
Eliciting event (% frequencies)
χ2 = 18.49 (p < 0.05)
Consumer knowledge 0.67 0.32
Service failure 0.33 0.68
Experiential Consequences
Cognition (% frequencies)
χ2 = 21.49 (p < 0.05)
Action oriented 0.43 0.15
Provider oriented 0.49 0.55
Rumination 0.06 0.16
Avoidance 0.02 0.14
Expectations (% frequencies)
χ2 = 9.07 (p < 0.05)
Core service related 0.46 0.25
Interpersonal service related 0.36 0.59
No expectations 0.18 0.16
Emotions (mean values)
Anger
t = -8.11 (p < 0.05) 1.97 3.55
Happiness
t = 8.44 (p < 0.05) 2.30 1.48
Worry
t = -5.42 (p < 0.05) 1.78 2.69
Shame
t = -3.21 (p < 0.05) 1.30 1.69
Emotion expression (mean values)
t = 1.65 (p < 0.10) 4.33 3.90
Service quality perceptions (mean values)
t = 8.47 (p < 0.05) 3.69 2.15
Consumer satisfaction (mean values)
t = 9.11 (p < 0.05) 4.87 2.49
having already paid for this expensive chalet,” or focus- action-oriented thoughts. These results reflect past re-
sing on how the provider was delivering the service. search indicating the “other-focus” of powerless consum-
Powerless consumers tended to have provider-oriented ers while powerful consumers focus on how they can
thoughts (55%) and reported a very low frequency of any accentuate and protect their powerful position in the

situation. Interestingly, two categories of thoughts – rumi- powerfulness, respondents reporting powerless experi-
native (e.g., what can I do, how did I get in to this situation) ences did not cite lack of knowledge as a cause of their
and avoidance (I’ll never come back here again, get out powerlessness. Rather service failure (which can occur
immediately) occurred to some extent in powerlessness despite consumer knowledge) led to powerlessness. Pow-
but not in powerfulness. erfulness appears to be a more complex experience than
powerlessness as evidenced by the fact that for the ante-
Consumer expectations of the service provider dur- cedents such as service setting, and experiential conse-
ing their experience of powerfulness and powerlessness quences of cognition and expectations, there was no one
showed a pattern similar to that of consumer cognition. clearly dominant category for powerfulness. Powerful-
Powerfulness was accompanied by expectations that the ness was equally likely in retail and professional services,
provider would attend to both the core service (46%) as led to both action oriented and provider oriented thoughts
well as the interpersonal service (36%). Powerlessness, and to expectations regarding both the core service and the
however, was dominated by expectations that the provid- interpersonal service. Powerlessness, on the other hand,
er will attend to the interpersonal aspects of the service had more clearly dominant categories. It tended to occur
(59%). For instance, the core service expectations could most in professional services, was dominated by provider
take the form of, “yes, when I pay for a hotel room or a oriented thoughts and expectations regarding the inter-
chalet, I expect it to be ready when they tell me” or “. . . I personal aspects of the service. In this context it should be
expected answers to all questions regarding food and the mentioned that informal feedback from respondents after
restaurant.” On the other hand, examples of interpersonal having completed the surveys indicated that those recall-
service expectations were, “they [the service provider] ing experiences of powerfulness had greater difficulty in
would be helpful and knowledgeable,” “expected the articulating the experiential aspects of this as compared to
usual prompt attention,” “politeness,” “be friendly, pleas- those recalling experiences of powerlessness. This may
ant,” or “consumer should be utmost.” indicate that the concept of power becomes more salient
for consumers when they perceive the lack of it, and
Regarding the emotions experienced, in keeping with therefore it is more clearly represented in their minds.
the fact that powerlessness occurred primarily due to Since powerlessness was linked with a negative event
service failure, powerless respondents reported higher such as service failure, it is also possible that, since
intensity of all negative emotions (anger = 3.55, worry = individuals place more emphasis on the negative, these
2.69, shame = 1.69) than did powerful respondents (an- experiences were more sharply sketched in their minds.
ger = 1.98, worry = 1.84, shame = 1.21; all p < 0.05), while
the latter reported greater happiness than did the former Managing consumer perceptions of power appears to
(2.30 and 1.48 respectively; p < 0.05). Further, while past be an important managerial issue since powerful consum-
findings indicate that powerless individuals are highly ers had higher perceptions of service quality and satisfac-
conscious of their self-presentation and thus may be less tion. For instance, consumers tended to feel more power-
expressive about their emotions, results here indicate that ful in same gender groups than in mixed gender groups,
although powerless consumers were indeed less likely to and this could have implications for segmenting consum-
be expressive than powerful consumers, the difference ers and assigning service provider responsibilities accord-
was only marginally statistically significant (means 4.33 ing to the gender of the consumer. Further, while service
and 3.90 respectively, p < 0.10). failure is clearly to be avoided, merely preventing failure
may not enhance consumer power. Rather, providing
Marketing Outcomes consumers with knowledge of the structure and process of
the service may go a long way in doing so.
Results indicate that powerful consumers have high-
er perceptions of service quality (mean = 3.69) and satis- Surprisingly, powerless respondents reported high
faction (mean = 4.87) than powerless consumers (2.15 intensity anger, an emotion generally thought to occur
and 2.49 respectively; p < 0.05). when an individual feels in control of a situation or wishes
to gain control over a situation (Conway et al. 1999).
DISCUSSION Further, contradicting past findings indicating that pow-
erless individuals are highly conscious of their self-pre-
This paper studied consumer experiences of power, sentation and thus may be less expressive, results indicate
an important construct in our understanding of consum- that both powerful and powerless consumers are likely to
er – service provider interactions during service con- equally express their emotions. While the nature of our
sumption. The research provides details of consumers’ data does not allow an examination of the reasons for this,
mental representations of their experiences of powerful- we speculate that this may be a deliberate strategy by the
ness and powerlessness. Evidence indicates that power- powerless consumers to convey their experience to the
fulness and powerlessness may not be simple mirror service provider and perhaps recruit the provider’s help in
images of each other. While consumer knowledge elicited navigating the remainder of the service. Such a reasoning

is in keeping with past research indicating that consumers gy and it will be interesting for future research to examine
are highly aware of the communicative role of their consumer power in the context of consumer interactions
emotions and use it accordingly (Parkinson 1995). It is with technology as substitutes for service providers (Moon
also possible that since most people are aware of the 2000). Further, the concept of social power may not be
power standing connoted by discrete emotions (e.g., Con- restricted to service situations and consumer – service
way et al. 1999), experiencing and expressing high power provider interactions. Consumer perceptions of their power
emotions may be a strategy to try and gain more power in may exist when they interact with other consumers such as
the setting. during communal consumption activities generated by
brand communities, and in fact any consumption situation
While as noted previously, experiences of powerful- embedded in a social context.
ness and powerlessness occurred primarily in high con-
tact services thus underlining the social nature of power, While important insights into consumer experiences
it will be interesting to study the occurrence of social of power were gained through this exploratory study,
power in services that are migrating from high contact to future research is required to address the shortcoming of
self service. Research indicates that consumers attribute this study (e.g., the use of a relatively small convenience
human characteristics to their interactions with technolo- sample) as well as the queries that arise from its findings.
REFERENCES ing, 76 (3), 285–307.

Moon, Youngme (2000), “Intimate Exchanges: Using
Bitner, M.J., B. Booms, and M. Tetrault (1990), “The Computers to Elicit Self-Disclosures from Consum-
Service Encounter: Diagnosing Favorable and Unfa- ers,” Journal of Consumer Research, 26 (4), 323–40.
vorable Incidents,” Journal of Marketing, 54, 71–84. Moskowitz, D., E. Suh, and J. Desaulniers (1994), “Situ-
Butcher, K., B. Sparks, and F. O’Callaghan (2003). “Be- ational Influences on Gender Differences in Agency
yond Core Service,” Psychology and Marketing, 20 and Communion,” Journal of Personality and Social
(3), 187–208. Psychology, 66 (4), 753–61.
Conway, M., R. Difazio, and S. Mayman (1999), “Judg- Oliver, R. (1997), Satisfaction: A Behavioral Perspective
ing Others’ Emotions as a Function of the Others’ on the Consumer. McGraw-Hill.
Status,” Social Psychology Quarterly, 62, 291–305. Parkinson, B. (1995), Ideas and Realities of Emotions.
Depret, E. and S. Fiske (1993), “Social Cognition and New York: Routledge.
Power: Some Cognitive Consequences of Social Price, L. and E. Arnould (1999), “Commercial Friend-
Structure as a Source of Control Deprivation,” in ships: Service Provider – Client Relationships in
Control Motivation and Social Cognition, G. Weary, Context,” Journal of Marketing, 63, 38–56.
F. Gleicher, and K. Marsh, eds. New York: Springer- Richins, M. (1997), “Measuring Emotions in the Con-
Verlag, 176–202 sumption Experience,” Journal of Consumer Re-
Fitness, J. and G. Fletcher (1993), “Love, Hate, Anger, search, 24, 127–46.
and Jealousy in Close Relationships: A Prototype and Richeson, J. and N. Ambady (2002), “Effects of Situa-
Cognitive Attribution Analysis,” Journal of Person- tional Power on Automatic Racial Prejudice,” Journal
ality and Social Psychology, 65, 942–58. of Experimental Social Psychology, 39, 177–83.
Gwinner, K., D. Gremler, and M.J. Bitner (1998), “Rela- Ridgeway, C.J. (1992), “Gender, Interaction, and Ine-
tional Benefits in Services Industries: The Custom- quality,” Personality and Social Psychology, 76 (5),
er’s Perspective,” Journal of the Academy of Market- 805–19, New York: Springer.
ing Science, 26, 101–14. Shaver, P., J. Schwartz, D. Kirson, and C. O’ Connor
Keaveney, S. (1995), “Customer Switching Behavior in (1987), “Emotion Knowledge: Further Exploration
Service Industries: An Exploratory Study,” Journal of a Prototype Approach,” Journal of Personality
of Marketing, 59, 71–82. and Social Psychology, 52, 1061–86.
Macoby, E. (1990), “Gender and Relationships: A Devel- Shemwell, D., J. Cronin, and W. Bullard (1994), “Rela-
opmental Account,” American Psychologist, 45, 513– tional Exchange in Services: An Empirical Investiga-
20. tion of Ongoing Customer Service-Provider Rela-
Menon, Kalyani and Laurette Dubé (2000), “Ensuring tionships,” International Journal of Service Industry
Greater Satisfaction by Engineering Salesperson Management, 5 (3), 57–68.
Response to Customer Emotions,” Journal of Retail-

School of Business and Economics
Wilfrid Laurier University
Waterloo, ON, N2L 3C5
Canada
Phone: 519.884.0710, Ext. 2704
FAX: 519.884.0201
E-Mail: kmenon@wlu.ca

WHAT DRIVES CUSTOMER SUCCESS? CONSUMER PERCEPTIONS OF
ENABLING AND RESTRAINING FORCES ASSOCIATED WITH
PERFORMING CONSUMPTION TASKS
Pete C. Honebein, University of Nevada, Reno
SUMMARY elements of the consumption environment that must be in

place to enable consumer performance. This not only
The consumer’s role in unlocking the value that includes such things as policies, but also procedures,
exists in products and services continues to attract the tools, interfaces, and technologies. Incentive defines the
interest of the marketing field. Spearheaded by Vargo and intrinsic and extrinsic motivators that stimulate consum-
Lusch (2004), the new dominant logic for marketing ers to perform in a certain way. Often, these motivators
defines the consumer’s role of coproducer, where the take the form of rewards, punishments, or both. Expertise
consumer is continually involved in the production of reflects the knowledge and skills possessed by the con-
value as he consumes products and services. The consum- sumer, developed through training programs and direct
er’s success as a coproducer is dependent upon the con- experience.
sumer’s physical and mental skills, or if the consumer
lacks those qualities, the embedding of those attributes in Method and Procedure
the product or service itself. Vargo and Lusch (2004)
suggest that organizations that recognize this balance can Given the exploratory nature of our research the
more effectively develop offerings that enable consumers design of the study is based on critical incident technique
to realize greater value. (Bitner, Booms, and Tetreault 1990; Meuter et al. 2000).
This methodology involves collecting actual stories that
The development of a consumer’s physical and men- describe, in as much detail as possible, situations in which
tal skills through such strategies as consumer and custom- an organization enables or restrains customer perfor-
er education is well recognized, as is the embedding of mance. These stories are then classified to uncover pat-
such qualities in products and services through self- terns or themes, in this case using the coproduction
service technologies (Meuter, Ostrom, Roundtree, and experience model as the primary classification structure.
Bitner 2000). Yet, a consumer’s ability to perform activ- A convenience sample of 144 students in an introductory
ities that realize benefits associated with products and marketing class was invited to participate in the study.
services is dependent not only upon the acquisition of Participation involved posting up to two stories on a web-
skills or the transfer of skills to another entity, but upon the based bulletin board during a two week period in ex-
expectations, feedback, and motivation the consumer change for extra credit. Participants were asked to write
possesses and receives from external sources (Bateson stories of at least 200 words that reflected their personal
2002; Gilbert 1996). The combination of these factors, experience with performance success or failure in one of
which this paper defines as vision, access, incentive, and the following roles: (1) a consumer; (2) an employee
expertise, forms a coproduction experience model that working with consumers; or (3) an observation of another
offers organizations a framework of the causal factors that consumer. Two example stories prepared by the principal
make self-service possible in the context of coproduction investigator provided participants a model of the form,
and value creation. The aim of this research is to investi- structure, and length expected for the stories. One exam-
gate which consumer performance factor has the strongest ple story described a tax preparer’s checklist, and the
influence on consumers’ success and satisfaction with other described a retail clerk teaching a customer about
products and services. using a camera.
Coproduction Experience Model Results
A performance orientation for consumers focuses on Seventy percent of the stories reflected situations
four key constructs: vision, access, incentive, and exper- where consumer performance was enabled, while 30
tise. We define vision in terms of two key components, the percent of the stories reflected situations where consumer
consumer’s awareness of key performance goals or out- performance was restrained. In terms of the consumer
comes, and the feedback the consumer receives related to performance model, a vast majority of the stories (72%)
the achievement of the goals or outcomes. Access defines described situations where access had the strongest influ-

ence on consumer performance, followed by expertise er performance (54%), followed by human policies/pro-
(13%), incentive (9%), and vision (6%). Within access, cedures (29%), and then other non-human solutions (17%).
SSTs were the most often identified influence on custom-

Peter C. Honebein
Managerial Sciences Department
University of Nevada at Reno
5450 Wintergreen Lane
Reno, NV 89511
Phone: 775.849.0371
E-Mail: honebein@unr.nevada.edu

THE EFFECT OF EVENT VALENCE ON WAIT
MANAGEMENT STRATEGIES
Elizabeth G. Miller, Boston College, Boston
Barbara E. Kahn, University of Pennsylvania, Philadelphia
Mary Frances Luce, Duke University, Durham
SUMMARY In addition, various methods of coping suggest that

delay can be preferred. In particular, emotion-focused
Most service experiences involve waiting and thus coping attempts often take the form of denial or avoidance
the total experience can be described as being comprised (e.g., Lazarus 1991). To the degree that people practice
of multiple parts, e.g., arrival, wait, main event, departure, “wishful thinking” or feel that the event is psychological-
that extend over time. Each of the components of the ly further away during a wait, delays may be preferable.
experience may be associated with different emotions, Similarly, to the degree people use delay to cope, for
and consequently, may require different coping resources example, by taking deep-breaths and telling themselves
and strategies which may interact. We examine such that things will be okay (more of a problem-focused
interactions by investigating how the expected valence of strategy), delay may also be beneficial.
an event affects two different wait management strate-
gies – providing duration information and reducing wait Wait Duration and Duration Information
times – for managing the stress experienced during the
wait. Four experiments demonstrate that event valence Thus, several literatures suggest that under certain
can moderate, and even reverse, the effects of wait man- circumstances delay can be beneficial and desired. In
agement strategies. particular, such considerations are likely to become rele-
vant when people are waiting for events which they find
Waiting and Event Valence stressful, negative, or otherwise aversive. We suggest that
in such negative situations, there are two sources of stress:
While the literatures on services and operations man- that associated with having to wait and that associated
agement have examined many facets of waiting, such with the event itself. Further, considering our arguments
research has not considered the effect of the waited-for about avoidance, coping, and procrastination, we believe
event on the waiting experience (other than to suggest that that interventions designed to attenuate wait stress may
people are willing to wait longer for more important have unintended negative consequences on the stress due
events; Houston et al. 1998; Maister 1985). In addition, to the event. In particular, we believe that since duration
such research has typically been conducted in non-stress- information – one popular intervention for reducing wait
ful environments, such as banks or restaurants. In these stress – is thought to operate by reducing uncertainty (Hui
environments, waiting is assumed to be negative because and Tse 1996; Osuna 1985), it may be particularly harm-
the event being waited for is either positive or neutral, and ful. That is, we predict that when the waited-for event is
waiting is assumed to be the focal source of consumer aversive, duration information can actually increase stress,
stress. In contrast, we argue that waits are not always even though this information attenuates stress in more
negative and the valence of the event matters. In particu- neutral situations. Further, given that increased wait time
lar, we propose that delay may have positive effects when can be used for coping, we predict that longer waits may
the event being waited for is negative or unwanted. actually result in reduced stress (and thus, a better waiting
experience) for aversive events, while the reverse finding
The literatures on procrastination, approach-avoid- is typically found in more neutral situations.
ance, and coping provide support for this contention.
Specifically, the procrastination literature suggests pro- We examine these two hypotheses across four studies
crastination is more likely when people want to avoid a in several different waiting environments. In the first
task (Greenleaf and Lehmann 1995), while Miller’s (1959) study, in which participants were asked to wait prior to
approach-avoidance conflict theory suggests desires for participating in a discussion group where they would
avoidance increase and are stronger than desires for either taste soft drinks or be required to give an impromptu
approach as one nears a feared stimulus. These tendencies speech, we find evidence that duration information can
suggest that desire for delay may predominate over the exacerbate stress when people are waiting for an aversive
desire to have the experience completed. Consequently, event and that these adverse effects occur because dura-
people may view delay less negatively (and perhaps even tion information interferes with the coping process. These
positively) in such situations. results are replicated and further supported in study 2,

where we find that people who choose to cope with particular, managers seeking to apply wait management
avoidance-oriented strategies respond more negatively to strategies that have been used successfully in one context
duration information when waiting for an aversive event should be careful when trying to apply those strategies in
(watching an aversive film) than those who do not. Stud- other contexts, especially when those contexts differ in
ies 3 and 4 extend these findings to a second wait manage- their valences. For example, hospitals should be wary of
ment strategy – reducing wait times – while also lending mimicking strategies currently employed by banks (and
credence to the notion that delay can have beneficial other industries) to reduce the negativity of waits. In
effects. addition, this research suggests that for situations involv-
ing negative or stressful aspects for consumers, managers
Implications should consider wait management strategies other than
duration information to manage these waits. Strategies
Based on these findings, we argue that marketers that assist consumers in managing their stress (from all
should consider the characteristics of their service or sources) will likely yield the most positive experiences.
good, as well as the emotional state of their customers, References available upon request.
when determining the best way to manage delays. In

Elizabeth Miller
Boston College
140 Commonwealth Ave.
Chestnut Hill, MA 02467
Phone: 617.552.2988
FAX: 617.552.6677
E-Mail: milleliz@bc.edu

SEEMINGLY UNRELATED REGRESSION: AN ALTERNATIVE TO
TRADITIONAL BRIDGING IN CONJOINT ANALYSIS
Niels J. Blunch, Aarhus School of Business, Denmark
SUMMARY The name “seemingly unrelated” is due to the fact

that the one and only connection between the two equa-
Splitting a conjoint job involving many attributes tions is the (possible) correlation of the error terms.
into several jobs each containing a sub-population of the
original attributes is a well-known method to overcome In the present application, the equations model the
respondent overload. first and second bundle of conjoint judgements by a single
respondent, and the possible correlation between the
In traditional bridging the conjoint job is split into error-terms is due to the fact that the two equations refer
two or more jobs each with a subsection of the attributes to the same respondent – a fact that is ignored by tradition-
(taking care not to split correlating attributes into different al bridging. Also, we place the restrictions on the estima-
jobs) and then combining the partial utility functions into tion of the parameters that the coefficients referring to the
one using the technique of bridging. The name indicating bridging attributes are equal across the equations, and the
that a few of the attributes are included in both (all) same goes for the two constants.
conjoint jobs, and are – so to speak – used as bridges to
bring the evaluations on a common scale. Comparing the Two Techniques
This is done by estimating the “best” (in an OLS- In an experiment seventy-nine business students at
sense) scale factor, by which to multiply the utilities for master level were asked to express on a scale from 0 to 10
the bridging attributes in the second design to match the their willingness to by 2 x 27 different cars, each described
utilities in the first. The hope being that the coefficients of in an 18-attribute profile. A week later they were asked to
all factors in the second equation should also be made evaluate five car profiles (with all 18 attributes). This
comparable with those of the first equation by multiplica- second round was a holdout round, i.e., the data from this
tion with the same constant. Only rarely will the “regulat- round was not used to calculate the preference functions.
ed” utilities of the bridging attributes from the second Rather they were used to evaluate the functions by com-
design be the same as those estimated from the first paring the results of the holdout round with predictions
design, and consequently we end up with two estimates of based on the preference functions calculated on data from
the utilities of the bridging attributes, as well as two the first round. The main results were as follows:
estimates of the constant. This problem is solved by
averaging these utilities. 1. Compared to traditional bridging SUR reduced the
MAD (Mean Absolute Deviation) by about 30 per-
Although bridging has been used in several conjoint cent and the correlation between actual and predicted
studies as an intuitively reasonable – although somewhat preferences was increased by no less than about 70
rough – method for solving the problem of respondent percent. These figures are averages taken over the
overload, the method is not quite satisfactory, as it is best five holdout profiles.
described as a sequence of ad hoc’eries. Consequently the
statistical (distributional) properties of the model are 2. The variation of MAD and correlations across the five
rather obscure, making traditional testing meaningless – profiles was considerable smaller for SUR than for
or at least complicated. traditional bridging – i.e., SUR is more stable.
In order to overcome the weaknesses of traditional 3. Traditional bridging results in larger uncertainty on
bridging, I would suggest Zellner’s seemingly unrelated parameters estimated in the second evaluation as
regression model (SUR). In its basic form, the model is: compared to the first.
Σi α X + ε
y1 = α0 + i i 1
To sum up: The advantages of SUR as compared to
traditional bridging are:
y =β + Σ βZ +ε
2 0 j j 2 1. Estimation and “bridging” is done in one go instead
j of in tandem.
where (possibly) cov(ε1, ε2) =/ 0

2. The estimation is based on all variables, which makes on data from the same respondent – are taken into
the result less erratic than traditional bridging, where consideration, whereas it is ignored in traditional
the result depends solely on the bridged attributes. bridging.
3. The uncertainty is more evenly divided among the 5. Unlike bridging, SUR is a formal model with well-
parameters in the model, whereas traditional bridg- known statistical properties; which means that statis-
ing favours the first conjoint evaluation. tical evaluation and testing etc. can be performed on
the total model, and not only on the separate (un-
4. The possible correlation of the error terms – caused bridged) models. References available upon request.
by the fact that the two parts of the model are based

Niels J. Blunch
Department of Marketing, Informatics, and Statistics
Aarhus School of Business
Haslegaardsvej 10
DK–8210 Aarhus V
Denmark
Phone: +45.8948.6688
FAX: +45.8615.3988
E-Mail: nib@asb.dk

WITHIN-INFORMANT BIAS IN MARKETING RESEARCH
James R. Brown, West Virginia University, Morgantown
Anjala S. Krishen, Virginia Polytechnic Institute and State University, Blacksburg
Pushkin Kachroo, Virginia Polytechnic Institute and State University, Blacksburg
Chekitan S. Dev, Cornell University, Ithaca
SUMMARY We illustrate our two approaches for reducing with-

in-informant bias using data from two concurrent and
Empirical researchers who investigate managerial related surveys of hotel-brand headquarters relationships
issues in marketing (e.g., sales force performance, marin the U.S. and Canada. In the first survey, we gathered
keting channel relationships) often rely upon the key data from the general managers of a number of hotels
informant methodology to report on organizational struc- representing two well-known U.S. lodging chains. The
ture and behavior, both within the firm and among orga- second survey asked brand headquarters field representa-
nizations. For example, Brown and Peterson (1994) as tives of these two chains to report on the relationships
well as Weeks, Chonko, and Kahle (1989) have asked between the brand headquarters and the hotels that the
sales managers to provide their subjective assessments of reps supervised. The two surveys yielded 245 matched
the individual salespeople whom they supervise. pairs – general managers and brand headquarters reps –
with complete enough data, representing a 14.2 percent
The problem with such an approach is that, for a response rate, to illustrate our approach. Further, virtually
variety of reasons, supervisors’ subjective ratings of their all of the headquarters reps reported on their relationships
subordinates are not independent. This lack of indepen- with multiple hotels. Hence, the potential for within-
dence violates a key assumption of multivariate analysis informant bias in our data set appears strong.
(e.g., ordinary least squares regression, structural equa-
tion modeling) that the error terms are independently Reliable and valid measures based on these data were
distributed. Violating this critical assumption can result in used to estimate an identified system of two simultaneous
biased standard errors (Johnston 1984; Kennedy 1998) equations, with hotel opportunism and brand headquar-
and unstable parameter estimates (Johnston 1984; Kennedy ters opportunism as the two dependent variables.
1998). Thus, the objective of this paper is to present and
illustrate two methods for reducing the biases that occur Our results show that, as compared to the parameter
when key informants report on multiple organizational estimates derived from the unadjusted data, the mean-
units (e.g., customer service reps, salespeople, retail man- centered bias adjustment method strengthens parameter
agers, channel member firms). estimates in some cases and weakens parameter estimates
in others, while in still other instances, this method has no
Perceptual reports of informants lack independence strong effect on the parameter estimates. The matrix
because of several factors, including halo effects, lenien- inversion-based bias elimination method operates simi-
cy or harshness, and central tendency (Cocanougher and larly.
Ivancevich 1978). Retrospective accounts of performance
may also lack independence because of a desire to appear In comparing the two correction methods, the infer-
consistent and rational as a result of the consistency motif ences drawn from the 2SLS analysis are not different,
(Podsakoff, MacKenzie, Lee, and Podsakoff 2003; Pod- although certain parameters do vary somewhat. Thus, for
sakoff and Organ 1986). our sample, the particular method used to adjust for
within-informant bias appears to matter less than the
The problem of within-informant bias can be ad- simple act of correcting for that bias.
dressed in at least two ways. One is to use data based upon
completely independent observations (e.g., John and Reve When key informants (e.g., sales managers) report on
1982). Another is to correct for the within-informant bias more than one organizational entity (e.g., salespeople),
mathematically. We take this latter approach here. those reports cannot be considered independent. When
used in multivariate analyses, these dependent reports
One method that we propose to adjust for within- violate the important assumption that the multivariate
informant bias is the mean-centering bias adjustment error terms are independently distributed. As a result, the
method. Another is the matrix invest-based bias elimina- parameters estimated from these multivariate analyses
tion method, a refinement of the first method. can lead researchers to make inferential errors. Such

errors can be avoided, however, by correcting for within- informant subjective assessments). References available
informant bias (i.e., the lack of independence among key upon request.

James R. Brown
College of Business and Economics
West Virginia University
P.O. Box 6025
Morgantown, WV 26506–6025
Phone: 304.293.3053
FAX: 304.293.5652
E-Mail: jamesb@vt.edu

BUILDING FORMATIVE CONSTRUCT MEASURES: THE EXAMPLE OF
CORPORATE REPUTATION
Sabrina Helm, University of Duesseldorf, Germany
SUMMARY crease in one indicator is accompanied by increases of the

other indicators. If reputation were modeled as a reflective
The epistemic relationship between variable and in- construct, the indicators – understood as a stakeholder’s
dicators in latent variable structural equation modeling perceptions of (e.g., product quality, treatment of employ-
(SEM) is often not considered by researchers, leading to ees, management quality, care for the environment, etc.) –
measurement model misspecification (Burke et al. 2003). are interpreted as “effects of a construct“ (Bollen and
Latent variables may be associated with reflective or Lennox 1991, p. 305). Reputation leads to these effects
formative indicators. From a conceptual and methodolog- meaning that reputation determines the quality of prod-
ical standpoint, it is very important which kind of indica- ucts, the quality of management, the treatment of employ-
tor specification is used. Taking the example of the ees, and so forth, as outcomes of reputation.
construct of corporate reputation, the paper therefore
raises two research questions: Formative indicators “cause” the latent variable,
they represent different dimensions of it. The construct is
1. How can the correct mode of indicator specification a summation of the formative observed variables associ-
be determined for a complex construct? ated with it meaning that changes in the indicators change
the construct. The indicators need not be correlated or
2. How can a formative approach to measuring reputa- represent the same underlying dimension (Bollen and
tion be developed and tested? Lennox 1991). Conceptualizing corporate reputation as a
formative construct means that the indicators lead to the
Due to a resurgence of interest in corporate reputa- construct as inputs which seems to be the more suitable
tion, an abundance of different definitions of the construct relationship between indicators and construct. Reputation
can be found in the literature. Wartick (1992, p. 34) for is an aggregation of all its indicators such as product
example defines corporate reputation as “the aggregation quality or treatment of employees. This implies that be-
of a single stakeholder’s perceptions of how well organi- cause it delivers high quality products, a firm has a good
zational responses are meeting the demands and expecta- reputation; because it treats employees right, it has a good
tions of many organizational stakeholders.” In accor- reputation, and so forth. The process of conceptualizing a
dance to this and similar definitions, corporate reputation formative measurement model for reputation is different
can be understood as a construct based on a firm’s contri- to the process used for reflective modeling. In the paper,
butions to its stakeholders. This understanding is also the process is discussed in detail following a set of
relevant concerning the most discussed reputation mea- conceptual criteria to clarify the construct’s epistemic
sures such as Fortune’s “Most Admired Companies” or nature. An empirical study in the consumer goods sector
the “Reputation Quotient” developed by the Reputation serves to illustrate the tasks associated with formative
Institute. In empirical terms, this definition of reputation modeling. Diamantopoulos and Winklhofer (2001) sug-
results in the formative conceptualization of the measure- gest a four step approach for index construction, which is
ment model for reputation as rankings are used to measure used for guidance in our study of reputation. It includes
reputation and to compare companies based on their content specification, indicator specification, test for
reputation. Rankings and indices are classical examples multicollinearity, and external construct validity.
of formative construct conceptualization. But as the liter-
ature on the two measures implies, they are created using The results of the study support a formative measure-
reflective indicators. ment model containing 10 indicators. As some of the
indicators have low weights (which are the equivalent for
In the paper, the meaning of a formative and reflec- loadings in reflective measurement models), the discus-
tive structure of a measure is pointed out in detail. Using sions part of the paper is devoted to the implications of
reflective indicators, the researcher assumes that the ob- formative modeling and the handling of low weights.
servable indicators represent the construct, the direction Generally, ex-post removal of indicators is not considered
of causality runs from the construct to the items. These are an option if the formative construct is changed by that
interchangeable which means that the construct is uni- deletion procedure (Diamantopoulos and Winklhofer
dimensional and the items correlated; they are required to 2001). The findings imply that the individual differences
share the same antecedents and consequences. An in- in the importance that stakeholder groups attach to the

proposed reputation dimensions need to be investigated that different stakeholder groups “give different weight”
more closely. If the importance – meaning the weights of to the reputation dimensions. Eliminating low-weighted
the formative indicators – would prove to be low in one indicators would result in a shortened, stakeholder-spe-
stakeholder setting, but high in another, measurement cific list of indicators. Of course, this would reduce the
models for reputation could be adapted to the stakeholder capacity of the measure to compare different stakeholder
groups if this were acceptable for formative models. This groups’ perceptions of a certain company – thereby lim-
finding would corroborate some authors’ understanding iting a study’s generalizability.
SELECTED REFERENCES sumer Research, 30, 199–218.

Diamantopoulos, A. and H.M. Winklhofer (2001), “Index
Bollen, K. and R. Lennox (1991), “Conventional Wisdom Construction with Formative Indicators: An Alterna-
on Measurement: A Structural Equation Perspec- tive to Scale Development,” Journal of Marketing
tive,” Psychological Bulletin, 110, 305–14. Research, 38, 269–77.
Burke, Jarvis C., S.B. Mackenzie, and P.M. Podsakoff Wartick, S.L. (1992), “The Relationship Between Intense
(2003), “A Critical Review of Construct Indicators Media Exposure and Change in Corporate Reputa-
and Measurement Model Misspecification in Mar- tion,” Business & Society, 31, 33–49
keting and Consumer Research,” Journal of Con-

Sabrina Helm
University of Duesseldorf
Geb. 23.31, Universitaetsstrasse 1
40225 Duesseldorf
Germany
Phone: ++49.211.811.13 49
FAX: ++49.211.811.52 26
E-Mail: sabrina.helm@uni-duesseldorf.de

EFFECTS OF ONLINE STORE ATTRIBUTES ON CUSTOMER
SATISFACTION AND LOYALTY
Miao Zhao, Roger Williams University, Bristol
Ruby Roy Dholakia, The University of Rhode Island, Kingston
SUMMARY Online Store Attributes and Customer Satisfaction
With the emergence of computer-mediated commu- In this paper, the main focus is on the relationships
nication (CMC), online stores are experimenting with between online store attributes and customer satisfaction.
attributes that are unique to the new media. There is a While Baker (1986) and Bitner (1992) proposed catego-
choice of many attributes (such as search engine, ordering ries of attributes that impact consumer responses to retail-
system, order status tracking, customer survey, personal- er cues, Eroglu, Machleit, and Davis (2003) argue that
ization, and virtual reality display, etc.), each performing these typologies do not easily translate into the online
a specific function and distinct from other attributes world. Existing empirical research is limited.
within the website. In addition to deciding which at-
tributes to include and how to specifically operationalize We report on an empirical study designed to test
the selected attributes, online store managers are also whether retail store attributes affect users’ satisfaction
concerned about the impact of an attribute or a set of with and loyalty to the website. Secondary data, collected
attributes on customer satisfaction and loyalty. at the online store level, were compiled directly from
www.bizrate.com in August 2003. Customer ratings on
While there is an established body of literature and 1079 individual online stores were collected.
decades of experience regarding the design of physical
stores, the new world of online stores and website at- First, factor analysis of 12 attributes (excluding “over-
tributes are now beginning to receive attention. In addi- all look and design of site”) was performed and the results
tion to describing multiple attributes, several authors have indicate that the time at which the measures are taken
attempted to categorize them into “must have” and “should contribute to the factor loadings; hence the two factors are
have” attributes (e.g., Burke 2002). Given the large num- labeled “at check out” and “after delivery.” The two
ber of possible attributes as well as the changing nature of factors explain 68 percent of the variance. Second, several
technology that makes new attributes increasingly possi- multiple regression analyses are performed to examine
ble, it is not surprising that there is a lack of consensus the relationship between the attributes and the attributes
regarding “must have” and “optional” attributes. and customer satisfaction and loyalty (intention to revis-
it).
Website Attributes
From the analysis of the customer ratings, we draw
Using Ghose and Dou’s (1998) classification of web- some conclusions regarding the influence of online store
site attributes, we first attempt to identify potential at- attributes on site design, satisfaction and repeat purchase
tributes across four different types of websites – commu- intentions. First of all, “ease of finding what you are
nication, entertainment, information, and transaction (on- looking for” and “clarity of product information” are the
line store). With so many website attributes to choose two most important attributes for generating positive
from, our analysis suggests that transaction (or online ratings of overall look and design of the site. The analysis
store) sites are more likely to include certain types of also suggests how specific attributes are operationalized
attributes. It also suggests that online stores are “attribute are as important as whether or not a specific attribute is
rich” – potentially containing the maximum number (16) included.
of the 25 specific attributes. It is also very likely that online
stores will differ not only on the number of specific We also find that the time gap between interacting
attributes incorporated within specific websites, but also with a site and evaluating the experience or indicating
how a specific attribute such as “customer support” is revisit intentions affects the impact of attributes; some
operationalized. These variations create important chal- attributes persist in their impact. Others appear to lose
lenges for research on the effects of individual attributes some of their impact.
on customer satisfaction and loyalty.

Since the analysis relied on secondary data from met expectations” became the dominant attributes influ-
Bizrate, interpretation of results and conclusions must encing online store ratings once the customer received
consider several methodological limitations. While the actual delivery of the product. These fulfillment variables
customer ratings were acquired twice, the temporal asso- are biggest challenges to all non-store retailing, including
ciation of the second set of attribute ratings with the internet retailing. This suggests that the most creative,
dependent variables seems to be the primary reason for the interactive, vivid online site won’t compensate for weak
observed relationship. Also, it is not surprising that “ful- fulfillment and customer support capabilities. References
fillment” variables such as “on time delivery,” “product available upon request.

Miao Zhao
Roger Williams University
One Old Ferry Road
Bristol, RI 02809
Phone: 401.253.5351
E-Mail: mzhao@rwu.edu

ORGANIZATIONAL FACTORS RELATED TO EFFECTIVE CUSTOMER
INFORMATION SYSTEMS PRACTICES
Debra Zahay, Northern Illinois University, DeKalb
SUMMARY five firms identified specific organizational factors perti-

nent to the management of customer information in a
Organizations are continually reassessing and re- strategic context. One exemplary company was compared
aligning their capabilities and seeking to enhance their to four others to uncover organizational issues and pro-
business performance. There are many factors that con- cesses leading to effective management of customer in-
tribute to organizational success. In recent years, many formation.
firms have turned to a new area, the management of
Customer Information Systems (CIS) (Zahay and Griffin During the interviews a pattern emerged which indi-
2004) or the use of customer information in systems such cated how customer information is integrated throughout
as Customer Relationship Management (CRM) systems the organization in a manner similar to the process by
(Reinartz, Krafft, and Hoyer 2003), to contribute to firm which new products are successfully managed within the
profitability. In addition, failure rates of applications organization. Co-location, teamwork, and functional inte-
associated with customer information, like Sales Force gration were recurring themes. Although one exemplary
Automation (SFA) and Customer Relationship Manage- company seemed to do an outstanding job of collecting
ment (CRM), remain high, averaging 50–60 percent (Rig- and disseminating information, all firms struggled with
by et al. 2002) and empirical support for their contribution issues of inter-functional conflict, including the role of the
to management success is ongoing. sales force in contributing data to these systems. Howev-
er, there also appear to be organizational factors some-
Therefore, factors not included in these prior studies what unique to the management of customer information.
must explain the variance between firms in terms of the For example, customer-centric strategies are developed
use of customer information and its translation of the use interactively as a dialogue between middle and upper
of those difficult-to-imitate resources into ultimate firm management, using customer data and competitive trends.
performance. One possible explanation for these differ- The similarities and differences of the CIS process to the
ences is organizational implementation factors. For ex- NPD process is the organizing structure of the results
ample, in New Product Development, organizational fac- section of the paper, which is available upon request from
tors are found to be important in developing successful the author.
new products In-depth interviews with 17 managers in
REFERENCES Rigby, D.K., Frederick F. Reichheld, and Phil Schefter

(2002), “Avoid the Four Perils of CRM,” Harvard
Reinartz, Werner, Manfred Krafft, and Wayne D. Hoyer Business Review, 80 (2), 101–109.
(2003), “Measuring the Customer Relationship Man- Zahay, Debra and Abbie Griffin (2004), “Customer Learn-
agement Construct and Linking it to Performance ing Processes, Strategy Selection, and Performance
Outcomes,” Teradata Center for CRM at Duke Uni- in Business-to-Business Service Firms,” Decision
versity Working Paper Series (www.teradataduke. Sciences, 35 (2), 169–203.
org).

Debra Zahay
Northern Illinois University
DeKalb, IL 60115
Phone: 815.753.6215
FAX: 815.753.6014
E-Mail: zahay@niu.edu

MARKETING CONSIDERATIONS IN WEIGHT CONTROL:
PRELIMINARY FINDINGS
Angela Hausman, University of Texas – Pan American, Edinburg
SUMMARY This study attempts to create a more comprehensive

understanding of the factors affecting weight control
Americans are fat, and getting fatter (Martin, Robin- efforts by employing non-student samples and building
son, and Moore 2000). Despite this over-consumption, on Perugini and Bagozzi’s (2001) Model of Goal Directed
people are not eating healthier (Kim, Nayga, and Capps Behavior (GDB). The resulting model, while developed
2001). Consumers are not happy about this trend – that is in the context of weight control, might be equally valuable
why they spend billions of dollars trying to correct the in understanding similar purposive behaviors, including
defects caused by overeating (Thompson and Hirschman other health behaviors such as smoking cessation and
1995) and 120,000 each year will die prematurely due to exercise, dark-side consumption behaviors, such as gam-
nutrition related ailments (Frazao 2000). Governments bling and eating disorders (Hirschman 1992), and other
are not happy about it either – that is why they spend process behaviors (Bagozzi and Edwards 2000). Hypoth-
billions of dollars promoting healthier food consumption. eses tested were:
Taxpayers and insurance carriers are not happy – since
medical costs associated with obesity account for five H1: (a) Negative attitudes toward diet foods and (b)
percent of direct and ten percent of indirect costs (Martin, negative attitudes toward giving up non-diet foods
Robinson, and Moore, Robinson, and Moore 2000). The negatively affect the relationship between desires to
food industry is not happy as customers line up to file control weight and performance of weight control
lawsuits blaming restaurants for their excess weight (CBS behaviors.
Evening News 2003).
H2: The perceived usefulness of food information posi-
Nothing seems to help. Consumers take off weight tively affects the relationship between desires and
through expensive diet programs, pills, surgery, and exer- performance of weight control behaviors.
cise only to put it back on again. Promotional advertising
and educational programs run by the government and H3: The perceived ease of implementing food restrictions
insurance companies fall on deaf ears (Wansink 2002). positively affects the relationship between desires
Nutritional labeling, once thought to be the panacea for and performance of weight control behaviors.
enabling consumers’ desires to eat healthier, does not
appear to be having the desired effect (Hill et al. 2002). H4: Mood states affect the performance of weight control
behaviors.
Marketing studies have developed models attempt-
ing to explain why consumers engage in behaviors that H5: Satisfying social needs negatively affects perfor-
thwart weight control goals. For instance, Bagozzi and mance of weight control behaviors.
various colleagues have developed the theory of trying,
and more recently, the theory of goal directed behavior, A preliminary questionnaire was developed and dis-
which explain more of the variance in observed behaviors tributed to members of the community using a quota
by incorporating past efforts, control, and desires as sampling technique. Using a system similar to the one
antecedents of food consumption (cf., Bagozzi and War- employed by Keaveney (1995), data quality was assessed
shaw 1990; Bagozzi and Edwards 2000). Recently, Wan- by contacting randomly selected respondents from each
sink (2002) contributed to this understanding by review- sample. These procedures produced a usable sample size
ing research related to World War II efforts to modify of 199. Existing scales were used to measure constructs.
food consumption. His study underscored the importance Reliability assessment demonstrated acceptable Cron-
of food availability and familiarity in evaluations of food bach’s alpha for all scales.
acceptability. Unfortunately, most of these studies suffer
from narrow sampling frames – mainly employing stu- Analysis showed 95 respondents were not trying to
dents – and low explanatory ability. Developing reliable lose weight, while 101 were, using a median split on
and valid measures of modeled constructs has also con- intentions to lose weight next week as the criterion. As
tributed to the lack of theoretical progress in understand- expected from prior research, more women than men were
ing these behaviors (Bhaskaran 2002). trying to lose weight. Dieters also tended to be under 40,

single, and better educated. The effect of income and then on subsets of those who are attempting to lose weight
dieting appeared bimodal, with lower and higher income and those who are not. Variables were added in sequential
earners being more likely to be dieters than those in blocks starting with exogenous variables and ending with
between. Demographic factors affected these variables, the proposed moderators on performance. Resulting re-
showing that women are likely to diet more frequently gressions were all significant (at .05 level) and the ex-
than men, as are those who are married and working full plained variance was 41.6 percent for the entire dataset,
time. 51.1 percent for dieters, and 47.4 percent for non-dieters,
providing superior explanatory ability over those ob-
To understand factors effecting weight control, a tained using the GDB alone (r2 = .25), although the
series of hierarchical regressions were conducted using variables are best at explaining the behavior of dieters as
performance of weight control behaviors as the dependent would be expected. References available upon request.
variable. The regression was run first on the entire dataset,

Angela Hausman
Management, Marketing, and International Business
University of Texas – Pan American
1201 W. University Drive
Edinburg, TX 78504
Phone: 956.381.2826
FAX: 956.384.5065
E-Mail: hausman@panam.edu

PRODUCT INVOLVEMENT AND PLACE ATTACHMENT: INSIGHTS
FROM THE ENVIRONMENTAL PSYCHOLOGY LITERATURE
Merlyn A. Griffiths, University of California, Irvine
ABSTRACT function for attaining security, comfort, connectedness,

and survival (Ainsworth and Bell 1970). An apparent
Discussions of person-product bonding relationships congruity exists between the bonding relationships con-
in the consumer behavior literature have focused on sumers develop with products and that which they devel-
consumer involvement with one category of products; op with places. Although consumer product involvement
namely physical goods. Yet, evidence exists that people is extensively studied (e.g., Andrews et al. 1990; Beatty
have similar relationships with other types of products et al. 1988; Bloch 1982, 1984; Higie and Feick 1989;
like place. Environmental Psychology recognizes person- Laaksonen 1994; Lastovicka 1979; Lastovicka and Gard-
place bonding relationships as “place attachment.” Draw- ner 1978; Laurent and Kapferer 1985; Richins and Bloch
ing on the literatures of product involvement in consumer 1986; Zaichkowsky 1985, 1986, 1994), relatively little is
behavior and place attachment from environmental psy- known about consumer involvement with places. Unlike
chology, this paper provides a conceptual framework the products studied in the consumer behavior literature
through which to understand consumer bonding relation- regarding involvement (e.g., cosmetics, automobiles, cloth-
ships with places. ing, etc.), places offer a wide variety of uses, activities,
settings, cultures, and landscapes that marketers can use to
INTRODUCTION build an ongoing relationship with consumers, emphasiz-
ing various enticing and attraction factors. These at-
Although products have been defined in the typical tributes of places, maintain stability as the geographic
marketing textbook to include person, physical good, location and spatial nature of the place never changes
organization, service, events, properties, places, experi- (e.g., Brazil never ceases to be Brazil), however, the
ence, information, and ideas; anything that can be offered adventures, excitement, and experiences one can engage
to a market to satisfy consumers want or need (for exam- in is limited only by the individual’s imagination.
ple, Kotler 2003, p. 407), a preponderance of the consum-
er behavior literature on product involvement emphasizes Person-place bonding relationships have been recog-
consumers’ involvement with physical goods. However, nized in environmental psychology as “place attach-
evidence exists that people have similar relationships with ment.” This literature offers potentially significant theo-
other types of products like places. Bonding with places, retical contributions to the product involvement literature
similar to involvement with products, objects, personal by expanding the range and type of products studied to
possessions, and other people, is recognized as a universal include “place” as a product. First, it broadens our under-
human phenomenon. standing of person-place relationships that contribute to
the individual’s identity. The strength of individuals’
Person-place bonding relationships although not clas- attachment to place is determined by the degree to which
sified as such, exist minimally in the consumer behavior they perceive the place as being an innate part of their
literature as part of consumer consumption of places. For existence, self image, and self concept (Sirgy 1982),
example, Penaloza (2000, 2001) demonstrates consum- similar to the strength of one’s attachment to personal
ers’ place attachment to the American West through the possessions. As Belk (1988) describes, possessions be-
recreation and commoditization of the Wild West era in come a part of the extended self. The extended self, he
American history, through Rodeo and Stock Shows. This summarizes, include “body, internal processes, ideas and
reenactment of America’s rich historical legacy has inter- experiences, and those persons, places, and things to
national appeal, and is relived in the United States, Europe which one feels attached” (p. 141). Thus, one’s identity is
(e.g., Old Texas Town in Berlin), and the United King- tied to the categories of things equated to be a part of the
dom. It is the cultural meanings ascribed to the West, the self, including places. Consequently, if place can be a part
values of freedom, family, and naturalism which facilitate of the extended self like material possessions, then place
the formation of person-place bonds with this historic must be ultimately tied to one’s identity. Place as it relates
place of an era long ago. to one’s identity is identified in the environmental psy-
chology literature as place identity which is one of the
From a psychological standpoint, person-place bond- determinants of place attachment representing the emo-
ing, similar to person-person attachment (like the bond tive and affective aspects including preferences, feelings,
between infant and caregiver), is a biological human and values (Proshansky 1978).

Second, place attachment adds to our understanding of-mouth. These institutions strive to endear alums to the
of the mobility of meanings (McCracken 1986) ascribed school independent of social ties to other alums. For many
to place by individuals. McCracken’s (1986) theoretical alumni the physical place is endowed with relevance and
account of meaning movement emphasizes the transfer of meanings often ascribed to churches, synagogues, or
meaning from the culturally constituted world to the religious institutions. For example, some universities like
consumer good, to the individual consumer. The place University of Richmond and University of Virginia are
attachment literature presents a theoretical framework offering columbariums or burial sites on campus, the
through which to explore the breadth of meaning move- ultimate extreme in person-place bonding with the phys-
ment by extending our understanding of how and why ical place even after death.
person-place attachments are formed, the symbolic mean-
ings ascribed to place and the relevance and importance Place as an Entity
that characterizes these meanings for the individual.
Place refers to space that has been given meaning
Third, understanding how different types of attach- through personal, group, or cultural processes (Low and
ments are classified will help researchers clarify bound- Altman 1992). As a spatial entity to which people develop
aries of not only place attachment, but also attachments to attachments, place can vary in scope, size, and scale. It can
brands, products, and material possessions. Kleine and be tangible versus symbolic, known and experienced
Baker (2004) recognized the lack of overlap between the versus unknown or not experienced (Low and Altman
place attachment literature and possession attachment in 1992). The tangible, known and experienced places are
consumer behavior, and make a pressing call for research- for example, homes, neighborhoods, communities, and
ers to integrate both literatures to broaden the scope of cities. Bonding with these everyday familiar places occurs
attachment research, further clarifying boundaries, mean- as a result of long term residential exposure. Similarly,
ing, and value of person-object and person-place bonding recreational places visited (e.g., Appalachian Trail, Ha-
relationships. waii, Jamaica, Disneyland) are also tangible, known and
experienced places however, the experience is short term
Understanding consumers’ involvement and attach- and non-residential in nature. Yet, special feelings of
ment to places has practical implications for both relation- connectedness develop toward these places that are expe-
ship and place marketers (i.e., tourism and destination rienced for a short period of time. Symbolic, unknown and
marketers) as well. As many segments of the tourism unexperienced places (e.g., heaven, Italy, if one has never
industry reach the maturation stage, these providers of visited) are places that have not been physically experi-
experiential consumption scramble to enhance custom- enced by the individual. The activities, rituals, or land-
er’s psychological attachment to place by seeking to build scape are elaborated on only through the individual’s
long term programs that communicate investments of imagination, influenced by stories, readings, or pictures
love, status, and reciprocal loyalty (Morais et al. 2004). (i.e., TV, magazines, and internet). Yet for many individ-
Leisure and recreation providers have sought to not only uals and groups, although there is no actual exposure in
understand the formation of and processes involved in the physical sense with these places, an existential bond
person-place bonding relationships, but also the impact develops.
on consumer commitment, satisfaction and loyalty to
place (Iwasaki and Havitz 2004). For relationship market- An example of an individual’s involvement and sub-
ers whose primary focus is building strategies of person- sequent attachment to a place is the 1995 film “While You
alization (i.e., one-on-one interaction) individualization Were Sleeping,” where the main character (played by
(i.e., specific needs and preferences) and continuity (i.e., Sandra Bullock) has a fantasy world of romance that
repeat patronage) (Gordon et al. 1998) comprehension of includes an attachment to Florence, Italy, a place she has
the process and degree to which consumers’ attachment to strong desires and affinity for, but has never visited. She
places evolves, presents invaluable information in influ- carries a passport daily in hopes of one day going there.
encing consumer preference, expectation, satisfaction, Her endearment to Florence is further evident in the way
and loyalty (Lau and McKercher 2004). Although the she speaks of the place as a significant part of her life. The
social aspects of place are often emphasized by these level of her involvement and attachment to this place is
marketers, some promote the importance of bonding with based on the feelings and romantic meanings she ascribed
the physical place. Summer camps, for example want to Florence, implying a revered place that if she were to
individuals to be attached to the place independent of the physically experience it, would represent a phenomenal
people that may be a part of the place. In this respect, the accomplishment. Thus, a place with strong self relevance
physical space, rather than merely social ties, drives in its meanings to the individual will be salient in thoughts,
campers to return each year, even if their friends do not. have a high degree of emotional significance and expec-
Institutions like colleges and universities yearn to in- tation of future experiences imagined or anticipated to be
crease place bonding relationships with their alumni to possible in that place.
gain benefits like increased donations and positive word-

Much of the environmental psychology literature knowledge, and belief) and practice (action and behavior)
emphasizes longevity in residence as a requirement for (Low and Altman 1992). Within the intellectual milieu,
place attachment to develop. However, Stueve, Gerson, place attachment research builds on the premise that
and Fischer (1975) found that the effect of length of people develop long-standing and meaningful relation-
residence on feelings about a locale is largely explained ships with places. Place attachment can develop through
by the mediation of local ties. This finding weakens four distinct ways; namely; biological, the evolutionary
studies that argued that attachment develops as a direct and physiological adaptations and ecological fit of people
result of length of residence. The narrow geographic view to places (Riley 1992); environmental, the interactional
and the bounded condition relegated by the length of experiences of cultural ecology (Hufford 1992); psycho-
residence have limited our understanding of place attach- logical, the experiences across life stages from childhood
ment, a complex and multi-faceted phenomenon (Manzo to adult life (Chawla 1992; Cooper-Marcus 1992; Rubin-
2003). This limitation was addressed when Low and stein and Parmelee 1992); and sociocultural, the social
Altman (1992) extended the spatial range of previous norm and ideologies, ritual performance, and culturally
studies, and asserted that people can be attached to places shared meanings (Ahrentzen 1992; Low 1992).
of varying scale, specificity, and tangibility, from the very
small, like objects, to the nation, planet Earth, or the Like product involvement, place attachment has wide
Universe. In a recent study, Hidalgo and Hernandez variability in its definition and operationalization across
(2002) found that attachment can also develop after brief studies. According to Low and Altman (1992) place
contact with place. The subjects in the study demonstrated attachment is the bonding of people to places. Hummon
attachment type behaviors after only short term exposure (1992) identifies it as an “emotional involvement with
to place, in particular where the opportunity for social places” (p. 256). Low (1992) considers it as “an individ-
interaction was absent. This finding demonstrates that ual’s cognitive or emotional connection to a particular
people can develop attachment to places after short term setting or milieu” (p. 165), and Shumaker and Taylor
exposure, and in the absence of social relationships that (1983) views it as a “multilevel person-place bond that
are typically formed through long term exposure. evolves from specifiable conditions of place and charac-
teristics of people, and that has implications for the
Based on the arguments above, it is my position that attitudes and behaviors of individuals toward their socio-
place attachment can also be formed through short term physical environments” (p. 223). Consensus however,
non-residential exposure (i.e., recreational or tourist plac- leads to the definition adopted in this paper: place attach-
es) or even non-exposure (i.e., wanting to visit Italy) to a ment is “a positive affective bond between an individual
place. This paper introduces place attachment to the and a specific place, the main characteristic of which is the
consumer behavior literature in an effort to extend our tendency of the individual to maintain closeness to such a
understanding of person-place bonding relationships. place” (Hidalgo and Hernandez 2001). In this respect,
Drawing on well-established theories and empirical stud- maintaining closeness includes nearness in physical prox-
ies of place attachment from environmental psychology, imity, cognitive closeness through the imagination, and/
this paper aims to provide a conceptual framework through or emotional closeness through the feelings evoked when
which to understand the characteristics of people-place the place is considered by an individual.
bonding relationships. The goal is to examine the proper-
ties of product involvement in its applicability to place as Studies of place attachment in environmental psy-
a product. In the sections to follow, an abbreviated review chology first establish a geographic or conceptual terrain
of the place attachment and product involvement litera- of interest as the boundary place to which attachment
tures is presented, followed by a summary of the key develops. With few exceptions studies have emphasized
concepts of person-place bonding relationships and po- limited spatial range of place to routinely familiar places:
tential implications for the product involvement litera- homes (Ahrentzen 1992; Brown et al. 2003; Harris et al.
ture. 1996; Hidalgo and Hernandez 2001), near-home territo-
ries (Fuhrer et al. 1993), home town (McAndrew 1998);
LITERATURE REVIEW (Abbreviated) neighborhood (Bonaiuto et al. 1999; Low 1992) commu-
nity (Hummon 1992; Pretty et al. 2003; Riger and Lavra-
Place Attachment kas 1981) and city (Brown and Perkins 1992). These
limited spatial ranges represent the tangible, known and
Place attachment is a complex phenomenon that experienced aspects of place, which carries a necessary
integrates many aspects of person-place bonding. In the requirement of longevity in residence as a determinant of
environmental psychology literature, theories of place place attachment. Each type of place offers different
attachment involve the affective aspects of bonds people experiences and exposure that facilitates the forming of
form with places (Low and Altman 1992). The emotional bonds.
qualities are often accompanied by cognition (thought,

Spatial Ranges of Place City: The city, larger in spatial range than home,
neighborhood and community, represents collective or
Home: Home as a place we inhabit has many mean- group ownership of public space and place. As a posses-
ings. It is the stage of much of our everyday performances, sion to which individuals and groups become attached,
and cultural artifacts eliciting important meanings to Belk (1992) identified city as “a collection of monuments,
people (Ahrentzen 1992). As a primary territory, home including buildings, cemeteries, and museums that extend
affords residents a sense of control that is rarely experi- our personalities” (p. 43). The city encompasses the
enced in other locations. Home is a safe haven, a place of community, neighborhood, and homes. In this respect, the
connection with family, a setting for enjoyed activities, city allows for further exploration by the residents (insid-
and a medium for identity displayed (Harris et al. 1996). ers), and non-residents (outsiders) alike. As the city is
Thoughts of home can elicit powerful memories and public domain, it is likely that transient non-resident
affective responses, as among other meanings, home can individuals more than residents would utilize the collec-
mean the house one grew up in, the setting where love was tive sites (museums, monuments) and other public estab-
first felt, the dwelling where one raised children (Cooper- lishments.
Marcus 1992). In his focus on attachment to possessions,
Belk (1992) posits that we spend much of our lives in our Determinants of Place Attachment
homes; our desire for what he calls “homeyness” results in
feelings of attachment to home. Several concepts have emerged in the environmental
psychology literature focusing on person-place bonding,
Neighborhood: Neighborhood has been considered that have been used inconsistently and in some cases
the range most important in the formation of attachment interchangeably in the literature. In particular, place iden-
bonds, thus the most often studied (Hidalgo and Hernan- tity (Proshansky et al. 1983); place dependence (Shumak-
dez 2001). A neighborhood has as its main composition er and Taylor 1983), and sense of place (Hay 1998) are
people and the interrelated social ties among them. Ac- most commonly used. The concepts are behaviorally
cording to Rubinstein and Parmelee (1992) “life stage and related in that they are concerned with the bonds between
patterns of interdependence are consistent influences on people and places. However, in some studies, the concepts
the nature and objective manifestations of emotional have been operationalized as consequences of place at-
bonds with neighborhoods” (p. 150). Neighborhood so- tachment, while in others one term has been used to
cial ties are held together through group collective behav- encapsulate the others. Twigger-Ross and Uzzell (1996)
iors that also serve to maintain order and cohesiveness for example, purport that place attachment functions to
among the residents. The physical boundaries of neigh- support and develop aspects of place identity. Jorgensen
borhoods vary in size and configuration, and are a result and Stedman (2001) posits that place attachment, place
of the residents’ collective demarcation and sentiments identity, and place dependence are three dimensions of
(Lawrence 1992). Neighborhoods are an extension of sense of place. Kyle et al. (2004) claims place identity and
individuals collectively coexisting. Brown et al. (2003, place dependence are dimensions of place attachment,
2004) examined attachment to neighborhood by focusing and Pretty et al. (2003) emphasizes place attachment and
on the neighborhood blocks as the level analysis. The place dependence are two of three indicators of place
complexity in a neighborhood has at its source the wide identity. The significant overlap in terminology and def-
heterogeneity of residents, demographically, socioeco- initions of these place related constructs and the lack of
nomically, and ethnically, making it a more dynamic uniqueness in operationalization should signal research-
concept for place attachment studies. ers to consider whether these concepts are the same or
whether they are theoretically different ways of looking at
Community: Communities are a composition of mul- the same phenomenon: person-place bonding. In this
tiple neighborhoods, allowing for a wider range of inter- regard, it is logically necessary to examine these con-
action among residents of the neighborhoods that make up structs to ascertain each of their relation to place attach-
the community. According to Riger and Lavrakas (1981) ment and applicability in understanding individual and
“attachment to communities refer to the neighborhood as group bonding relationships with place.
a “network of necessities, that go beyond the minimal
level of functional necessity and become conscious com- Place Identity: Place identity has been defined as
munities, in which attachment persists because of adher- “those dimensions of the self that define the individual’s
ence to a clear set of values, despite the absence of personal identity in relation to the physical environment
traditional functions which formerly bound people to by means of a complex pattern of conscious and uncon-
neighborhoods” (p. 57). A key indicator of community scious ideas, beliefs, preferences, feelings, values, goals
attachment is greater freedom of behavior, exploration, and behavioral tendencies, and skills relevant to this
confidence, and affective responses within the local com- environment” (Proshansky 1978, p. 155). Place identity
munity (Fried 2000). relates to “the variety and complexity of physical settings

that define the day-to-day existence of every human sense of place is not imbued in the physical setting, but
being” (Proshansky et al. 1983, p. 59). Thus, place iden- resides in human interpretations of the setting (Jorgensen
tity is one of many facets of an individual’s self-identity and Stedman 2001). As viewed by Hay (1998), sense of
that helps in structuring experiences with physical places place can be broader in context than place attachment, as
(Shumaker and Taylor 1983). Place identity is most con- a result of the “subjective qualities, and the sensing of
cerned with how places form identity (Moore 2000), and place to create personal meaning” (p. 7). Shamai (1991)
how a place shapes or builds aspects of an individual’s self describes sense of place as having levels of intensity of
identity. For example, a person who identifies with the feeling and behavior from belonging (affiliation) and
neighborhood of the South End of Boston might identify attachment (special affinity) to commitment (ready to do
himself or herself as a “Southie,” expressing the dimen- something for the place). He further notes that “place is
sion of self that is defined by and related to the physical never merely an object, but a part of a larger whole that is
environment. The strength of the individual’s emotional being felt through the ‘actual’ experience of meaningful
attachment to place is based on self definitions of who and events . . . the experience is felt through all the senses
what he or she is relative to place (Proshansky et al. 1983). (sight, hearing, smell, taste, and touch)” (p. 348). From
Thus, for an individual to become attached to a place they this perspective, it can be surmised that sense of place is
must first identify with the place. It is the meanings they a necessary requirement for place attachment, as it is the
ascribe to the place that enforces the feeling of similarity, meanings individuals and groups ascribed to the place that
belonging, and relatedness. Place attachment is integral to feeds attachment.
self-definitions (Brown and Perkins 1992), making place
identity the reinforced symbolic representation of the self Measurement Approaches of Place Attachment
that identifies with the particular place. In this respect,
place identity is representative of the emotive and affec- Across studies, the most common measure used to
tive aspects of place attachment. It is the intimate self- identify the existence of place attachment has been “length
interpretation of what an individual may take to be a sign of residence” (Bonaiuto et al. 1999; Brown et al. 2003;
or locus of one’s identity. Riger and Lavrakas 1981). As an indicator of place
attachment, length of residence hinges on residents’ dem-
Place Dependence: Place dependence refers to “an onstration of attachment through physical manifestations
occupant’s perceived strength of association between him like upkeep of the appearance of their homes and property
or herself and specific places” (Stokols and Shumaker (Jorgensen and Stedman 2001), protecting the home and
1981, p. 457). The strength of the association between neighborhood from crime and other incivilities (Brown
person and place is based on the individual’s comparative et al. 2003; Brown et al. 2004), and interacting with other
judgment of the current place as a mechanism to satisfy residents (i.e., knowing names of neighbors and neigh-
needs and goals, and expectations of having goals and borhood children) (Mesch and Manor 1998). Length of
needs met by alternative comparable places (Shumaker residence, by the nature of its mandatory requirement of
and Taylor 1983). The outcome from each option may be long term participation, neglects attachment bonds that
negative; meaning that the current or alternative place can be formed with temporary residences, places to which
may not be ideal in satisfying the individual’s goals. exposure is short term or limited, places that are not yet
However, in comparison, the judgment is made to choose physically experienced, symbolic and other intangible
the better of the two less than ideal options. Place depen- places. Harris et al. (1996) demonstrated that attachment
dence is therefore setting specific, as its main concern is to temporary residence is possible, with their study of
the adequacy of fit between the individual’s goals and the attachment to a student housing facility. Kyle et al.’s
particular place, and the achievement of those goals. It is (2004) study of attachment to the natural setting Appala-
the functionality of the place that drives the level of chian Trail demonstrated that attachment does form through
dependence an individual develops to the place. Place non-residential and short term exposure to place. The
dependence differs from place attachment in two ways. length of residence measure is acknowledged as a power-
First, place dependence can be negative depending on the ful correlate of attachment; however, it is theoretically
limitations of the place in achieving the individual’s problematic as it excludes attachment to non-residential
desired goals; and second, the strength of the connection places. A brief review of product involvement is present-
between the person and place is based on specific behav- ed here to explore the factors characteristic of product
ioral goals rather than general affect (Jorgensen and involvement that may be applicable to person-place bond-
Stedman 2001). The functional aspect of place depen- ing.
dence suggests it is representative of the “commitment to
place” aspect of place attachment. PRODUCT INVOLVEMENT
Sense of Place: Sense of place is the meaning at- Consumer behavior researchers have examined the
tached to a spatial setting by a person or group. Further, person-product bonding relationships by focusing on

consumer involvement with different product categories. Brand Commitment: Brand commitment and prod-
A significant number of empirical studies of product uct involvement have been identified as related but dis-
involvement begin by first setting specific parameters tinct constructs (Traylor 1981). These constructs differ in
identifying a product or object of interest, like automo- that the object of involvement is a product, while the
biles (Bloch and Richins 1983) cosmetics (Coulter et al. object of commitment is a brand. In fact, according to
2003) and fashion clothing (O’Cass 2001). As a construct, Fournier (1998), “brand has no objective existence at all:
product involvement has been operationalized in numer- it is simply a collection of perceptions held in the mind of
ous ways, resulting in a great deal of variability in the the consumer. The brand cannot act or think or feel –
interpretation of empirical evidence across studies. Coulter except through the activities of the manager that adminis-
et al. (2003) defines product involvement as the personal ters it” (p. 345). Brand commitment has been defined as an
relevance or importance of a product category; O’Cass emotional or psychological attachment to a brand or
(2001) sees it as a person specific characteristic that exists product class (Beatty et al. 1988; Robertson 1976; Traylor
among consumers in varying degrees; Bloch (1986) 1981; Warrington and Shim 2000). Strong brand commit-
recognizes it as an unobservable state reflecting the amount ment has been associated with high levels of involvement
of interest, arousal or emotional attachment a consumer (Warrington and Shim 2000), implying the greater the
has with a product; and Zaichkowsky (1985), Celsi and brand commitment, the more firmly fixed is the brand as
Olson (1988) and Warrington and Shim (2000) believe the only choice for the consumer (Traylor 1981). The
product involvement is a person’s perceived relevance of strength of the consumer commitment to a brand lies in the
the object or product class based on inherent needs, distinguishable attributes of the brand and the salience of
values, and interests. these attributes relative to the consumer’s belief system in
regard to the product (Robertson 1976). However, in-
It is well documented that levels of product involve- volvement is not a necessary requirement for commit-
ment exists on a continuum from high to low (Antil 1984; ment, as a consumer can become involved with a product
Lastovicka and Gardner 1978; Leavitt et al. 1981). Deter- and not be committed to it. Involvement occurs when
mination of high versus low is in large part a function of values important to the individual’s self-image are made
the level of interest in the product and the level of salient. Commitment results when these values, self-
cognitive processing and behavioral activities that the image or important attitudes are cognitively connected to
consumer engages in (Antil 1984). Consumers with high a specific situation. In this respect, product involvement
product involvement exhibit strong interest in a product leads to brand commitment (Beatty et al. 1988; Coulter
that conceivably occupies their thoughts (Richins and et al. 2003; Fournier 1998).
Bloch 1986) and demonstrates strong commitment (War-
rington and Shim 2000) and enthusiasm (Bloch 1986). Brand Loyalty: Brand loyalty has also been linked to
Consumers with low involvement product rely on short both brand commitment and product involvement. As
term memory and routine behaviors or minimal cognitive closely related concepts, distinction in operationalization
processing (Leavitt et al. 1981) as they are not closely tied is often blurred. For example, Quester and Lim’s (2003)
to the product or brand outside of their focal attention study examining the relationship between brand loyalty
(Robertson 1975). and product involvement, states that “brand loyalty devel-
ops when the brand fits the personality or self-image of the
As an individually defined phenomenon, involve- consumer or when the brand offers gratifying and unique
ment with products requires ongoing commitment with benefits that the consumer seeks . . . in both instances,
regard to thoughts, feelings, and behavioral response personal attachment develops toward the brand” (p. 26).
(Quester and Lim 2003), emotional connection (Bloch However, in an earlier study Beatty et al. (1988) claimed
1986) as a function of the meanings ascribed to the that not brand loyalty, but “commitment results when
product, and strong personal relevance (Bloch 1986; these values, self-images or important attitudes become
Bloch and Richins 1983; Higie and Feick 1989). Empir- cognitively linked to a particular stand or choice alterna-
ically studied, product involvement has been said to have tive” (p. 152). Consumer loyalty, as described by Oliver
robust influencing effects on consumer cognitive and (1999) is “a deeply held commitment to rebuy or repa-
behavioral responses, including memory, attention, pro- tronize a preferred product/service consistently in the
cessing, early adoption, search, brand commitment, satis- future, thereby causing repetitive same-brand or same
faction, and opinion leadership (see Laaksonen 1994). brand-set purchasing” (p. 34). Thus, loyalty has two
distinctive concepts; namely, behavioral which relates to
Determinants of Product Involvement the purchase and repeat aspect, and attitudinal which
emphasizes the commitment depth relating to the pre-
Several concepts have emerged in the consumer ferred product or service. This distinction is the basis for
behavior literature relating to consumer involvement with the difference between product involvement, brand loyal-
products. Most often found intermingled with product ty, and brand commitment.
involvement are brand commitment and brand loyalty.

KEY CONCEPTS: PLACE ATTACHMENT AND ence their interest to the extent to which they become a
PRODUCT INVOLVEMENT wine enthusiast. Product involved enthusiasts have been
characterized by Bloch (1986) as having high levels of
From the literature review on place attachment and information seeking, opinion leadership, innovativeness,
product involvement, it can be determined that both are product nurturance, or care and represent a significant
psychological constructs that involve the interplay of marketplace force.
individually determined self relevant meanings ascribed
to the product or place of interest. However, similarities CONCLUSION
and differences exist between these constructs. Place, like
product can become irreplaceable. When an individual The position taken in this paper is that bonding with
identifies with place and endows it with meanings the place is a cognitive, affective, and behavioral commit-
extent to which place becomes an interpretation of the self ment. Further, short term exposure as well as non-expo-
which draws on place as a locus of identity (Hummon sure can also result in the forming of bonds with place.
1992), place becomes unsubstitutable. Like products, Experience through fantasy, desire, and imagination sup-
these places are those that for the individual there are no ply equally strong interest and involvement to which
alternatives or substitutes that have the same meaning as attachment develops. Developing a bond with place does
the original. not require physical experience. Similarly, developing a
bond with products does not require ownership of the
The factors that drive the development of place product. It is generally accepted that a product can be a
attachment (i.e., biological, psychological, environmen- place. Place, like tangible products, can become an inte-
tal, and socio-cultural) contributes to our understanding gral part of one’s self image and identity. However, the
of product involvement. The aspects of an individual’s properties of place that influence an individual in the
upbringing, the environmental and sociocultural facets forming of bonds are not fully explored. In this respect,
indigenous to where one grew up and the cultural norms researchers are encouraged to explore the dynamic nature
and rituals experienced may predispose one to become of place to deepen our understanding of person-place
involved with specific products. For example, an individ- bonds, and extend the product involvement literature
ual growing up in a region where wine consumption is a using the insights of place attachment from the environ-
ritualistic part of meals, (e.g., Italy or France) may influ- mental psychology literature.
SELECTED REFERENCES ment: A Conceptual Inquiry,” in Place Attachment,

Irwin Altman and Setha M. Low, eds. Vol. 12, New
Ainsworth, Mary D. Salter and Silvia M. Bell (1970), York: Plenum Pres.
“Attachment, Exploration, and Separation: Illustrat- Penaloza, Lisa (2000), “The Commodification of the
ed by the Behavior of One-Year-Olds in a Strange American West: Marketers’ Production of Cultural
Situation,” Child Development, 41, 49–67. Meanings at the Trade Show,” Journal of Marketing,
Bloch, Peter H. and Marsha L. Richins (1983), “A Theo- 64 (October), 82–109.
retical Model for the Study of Product Importance Proshansky, Harold M., Abbe K. Fabian, and Robert
Perceptions,” Journal of Marketing, 47 (3), 69–81. Kaminoff (1983), “Place-Identity: Physical World
Hidalgo, M. Carmen and Bernardo Hernandez (2001), Socialization of the Self,” Journal of Environmental
“Place Attachment: Conceptual and Empirical Ques- Psychology, 3, 57–83.
tions,” Journal of Environmental Psychology, 21, Richins, Marsha L. and Peter H. Bloch (1986), “After the
273–81. New Wears Off: The Temporal Context of Product
Laaksonen, Pirjo (1994), Consumer Involvement: Con- Involvement,” Journal of Consumer Research, 13,
cepts and Research. New York: Routledge. 280–85.
Low, Setha M. and Irwin Altman (1992), “Place Attach- Complete set of references will be provided upon request

Merlyn A. Griffiths
Graduate School of Management
University of California, Irvine
Irvine, CA 92697
Phone: 714.366.6998
E-Mail: griffitm@uci.edu

EMERGENCY CONTRACEPTION: EXPECTATIONS OF
PRODUCT NEED AND USE
Andrew M. Parker, Virginia Polytechnic Institute and State University, Blacksburg
Melanie A. Gold, University of Pittsburgh, Pittsburgh
SUMMARY primary contraceptive, it is possible that experience with

EC could lead individuals to think of it as such. On the
Emergency contraception (EC) is a form of contra- other hand, EC use can cause a number of temporary but
ception that reduces the risk of pregnancy after unprotect- uncomfortable side effects, which if seen as significant
ed sex or contraceptive failure. Though available in some may actually decrease future willingness to engage in
form since the late 1960’s, only recently has it been the unprotected sex. Lastly, it is expected that the perceived
focus of behavioral research. Recent FDA deliberations risk of getting pregnant from an episode of unprotected
on switching one EC product (Plan B) from prescription- sex will moderate the positive relationship between actual
only to over-the-counter status have brought EC even unprotected sex and actual EC use, with this relationship
more public attention. Focused on a specialized market being stronger for those with higher risk perception.
(particularly young women) and being potentially stigma-
and affect-laden (because it is a medication used solely In a longitudinal study, 301 sexually active female
after unprotected intercourse to prevent pregnancy), EC adolescents participants completed an initial, one-on-one
represents a rather unusual product class. From a market- interview conducted in a research office, followed by six
ing perspective, in trying to deliver the product to the monthly 10-minute telephone interviews (see Gold et al.
consumer in a way that is most useful, it is important to 2004). At each interview, the perceived likelihood of their
understand how consumers think about, use, and are next sexual intercourse being unprotected and of using EC
affected by EC, particularly adolescents and young wom- in the next month were gathered, as well as whether they
en. From a public-health perspective, this same informa- had engaged in unprotected sex and/or used EC in the
tion could be invaluable in protecting individuals’ safety preceding month. Participants also indicated the per-
and well-being. ceived risk of pregnancy from one episode of unprotected
sex.
Because use and misuse of EC reside at the center of
this important public health debate, understanding what Across the six months, only 5 percent to 9 percent
motivates individuals to seek out and use EC (i.e., market predicted that their next sexual intercourse would be
demand) is crucial. The focus here will be on one piece of unprotected, while 24 percent to 33 percent predicted EC
the EC product-consumer relationship that has not been use during the next month. The overall pattern, across
considered – expectations of the future. Expectations, in visits, is that expectations of having unprotected sex
this case of having unprotected sex and possibly needing correlated positively with reports of subsequent unpro-
EC, are central to most theories of choice. To the extent tected sex (omnibus p < .0001; Strube 1985), while
that the perceived likelihood of needing EC is high (i.e., expectations of EC use correlated positively with subse-
unprotected sex seems likely), and to the extent that quent reports of actual EC use (p = .03). Significant
having EC on hand will diminish the chances of an correlations were not found between expectation of un-
unwanted outcome (i.e., pregnancy), seeking out EC will protected sex and EC use or between expectation of use
be a more attractive option. and unprotected sex. Finally, reports of actual unprotect-
ed sex correlated positively with EC use, although this
It is hypothesized that expectations of having unpro- relationship was not moderated by risk perception. A very
tected sex will correlate positively with future reports of similar pattern arose when looking at the relationships
actual unprotected sex. Similarly, expectations of EC use between past behavior and expectations of the future.
will correlate with future reports of actual EC use. How-
ever, each expectation is hypothesized to correlate poorly In summary, expectations do correlate with behavior,
with the other behavior. A crucial concern is whether the both past and future. However, a key distinction is be-
expectation of using EC will in fact lead to more irrespon- tween the expectation of having sex (and potentially
sible behavior (i.e., unprotected sex). To the extent that needing EC) and the expectation of actually using EC.
EC expectations are not related to subsequent unprotected Much of the public policy debate regarding EC has
sex, however, such a concern holds less weight. A similar centered on how EC could encourage irresponsible sexual
issue involves the expectation of unprotected sex and and contraceptive behavior, particularly among adoles-
actual past EC use. While EC is not recommended as a cents. Two of the results here argue against that concern.

First, expectation of EC use did not predict future reports For healthcare providers and marketers, these results
of unprotected sexual behavior. Second, past EC experi- suggest the validity of interventions that recognize the
ence did not predict expectation of future unprotected sex. usefulness of expectations. If a young woman expects that
From the perspective of advance provision and EC’s over- she will use EC in the future, then this expectation should
the-counter status, while unprotected sex definitely influ- be identified during a health care visit and counseling
ences EC use (EC wouldn’t be necessary without it), we should include a discussion about how to facilitate timely
have no evidence that EC use is influencing unprotected use of EC and when the medicine is most effective.
sex.
REFERENCES haviors,” Journal of Pediatric and Adolescent Gyne-

cology, 17, 87–96.
Gold, Melanie A., Jennifer E. Wolford, Kym A. Smith, Strube, Michael J. (1985), “Combining and Comparing
and Andrew M. Parker (2004), “The Effects of Ad- Significance Levels from Nonindependent Hypothe-
vance Provision of Emergency Contraception on sis Tests,” Psychological Bulletin, 97, 334–41.
Adolescent Women’s Sexual and Contraceptive Be-

Andrew M. Parker
Department of Marketing (0236)
Virginia Tech
Blacksburg, VA 24061
Phone: 540.231.3096
FAX: 540.231.3076
E-Mail: anparket@vt.edu

INVESTIGATING DRIVERS OF CUSTOMER DEFECTION: A RELATIVE
WEIGHT APPROACH
Thomas Hollmann, Arizona State University, Tempe
Cheryl Burke Jarvis, Arizona State University, Tempe
SUMMARY Second, much of the extant research in satisfaction

and loyalty uses behavioral intention probabilities (e.g.,
Customer satisfaction, loyalty and retention are cen- “likelihood to renew the contract” on a 7-point Likert
tral tenets in the practice of customer relationship man- scale) rather than actual defection data (e.g., Anderson
agement (CRM) for both goods and services, and these and Sullivan 1993; Parasuraman, Zeithaml, and Berry
concepts have been the focus of heavy study. Yet custom- 1988), despite the fact that Garland (2002) has shown that
er defection rates remain quite high in many service-based self-reported switching probabilities overestimate actual
industries, as shown by Griffin and Lowenstein (2001). defection rates. Thus, both measuring satisfaction as a
Therefore it may be surprising that defection itself has predictor of defection and measuring self-reports of be-
received far less attention in both research and practice havioral intentions fall short in providing researchers a
than its counterparts of loyalty and retention. Only recent- valid tool for identifying and analyzing drivers of custom-
ly, a variety of studies have started to look at specific er defection. Different methodological approaches are
drivers of customer defection (e.g., Ahmad 2002; Capra- needed to more effectively measure and evaluate the
ro, Borniaczyk, and Srivastava 2003; Keaveney 1995) or system of factors that influence customer switching be-
more generally at influences on repurchase decisions haviors.
(e.g., Bolton and Lemon 1999; Jones, Mothersbaugh, and
Beatty 2003). Therefore, the goal of this paper is to extend the
methodological arsenal of customer loyalty research by
Interestingly, Keaveney’s (1995) qualitative research proposing and illustrating the use of a defection weight
found that the majority of defected customers identified a method (DWM) of data collection and analysis to assess
complex combination of multiple reasons for a decision to the relative influence of key drivers of customer defec-
switch service providers, rather than a single driver of tion. This method allow researchers to investigate the
defection. This finding highlights the need for both aca- relative influence of a constellation of factors that contrib-
demics and practitioners to understand drivers of defec- ute to switching decisions, rather than evaluating influ-
tion in the context of a comprehensive model, rather than ences in isolation from each other. DWM employs graded
through individual investigations of single factors in paired comparisons to understand the preference structure
isolation as is typical of the literature to date. of defection drivers.
In addition, measurement has complicated the study The paper reports results from a survey of recently
of defection drivers in two ways. First, researchers often defected consumers on a variety of services using the five
have relied on measures of customer satisfaction to iden- elements of the Servqual scale, price, and the physical
tify problem areas in service delivery that might influence product as defection driver categories. The study shows
future defection. Satisfaction has been found to have a that a vast majority of defection decisions are multi-
significant impact on repurchase intentions, but that im- dimensional. Less than 10 percent of the respondents have
pact is typically small (Bolton 1998), and it has been a uni-dimensional defection weight profile in which one
shown that in some cases as much as 65 percent to 85 reason dominates, i.e., represents 75 percent or more of
percent of defected customers report being “satisfied” or the decision to defect. This finding supports Keaveney’s
“highly satisfied” (Reichheld 1996). Even dissatisfaction (1995) qualitative findings of the complexity of the defec-
is not necessarily a predictor of defection (Hennig-Thurau tion decision.
and Klee 1997). For a loyal customer, it may take repeated
incidences of dissatisfaction with various elements of a The paper also compares the relative influence of the
product purchase or service encounter before the custom- tested defection drivers using the DWM technique to the
er chooses to switch. Or, it may require that a customer relative influence of the same drivers as reported by
reach a threshold level of dissatisfaction on a particular traditional customer satisfaction measures using regres-
factor or combination of factors to choose to leave. Even sion analysis. Significant differences in the relative weights
if a customer is not particularly loyal, defection may not of the tested factors are demonstrated between the two
occur despite unsatisfactory experiences, because of con- methods, supporting past evidence that traditional mea-
venience or competitive issues. sures of satisfaction are not valid proxies for drivers of
defection.

Finally, we use the DWM results to perform a cluster enced to defect by the elements of service quality as
analysis of customers based on defection drivers, illustrat- expressed by the Servqual, while about 30 percent of
ing a method that practitioners and academics alike could respondents were more strongly influenced by the non-
use to segment customers and differentiate services. Close Servqual elements. References are available upon request.
to 70 percent of respondents were more heavily influ-

Thomas Hollmann
W.P. Carey School of Business
Arizona State University
P.O. Box 874106
Tempe, AZ 85287–4106
Phone: 480.965.3621
FAX: 480.965.8000
E-Mail: Thomas.Hollmann@asu.edu

INVESTIGATING THE MODERATORS OF THE CUSTOMER
SATISFACTION-LOYALTY LINK: EVIDENCE
FROM RETAILING
Heiner Evanschitzky, University of Muenster (MCM), Germany
Gianfranco Walsh, University of Strathclyde, Scotland
SUMMARY restricted model. With four degrees of freedom more, the

restricted model exhibits a significant Chi-square differ-
Research on the relationship between customer satis- ence (at p < .05) for three effects: “critical incident,”
faction and customer loyalty has advanced to a stage in “expertise,” and “income.” The Chi-square difference for
which moderator variables need to be examined more the “loyalty-card membership” is slightly less than 9.49,
thoroughly. The relevance of customer satisfaction for indicating a difference significant at a level slightly below
maintaining successful relationships with customers has .05. Because this effect just fails the threshold, we decided
been discussed widely (e.g., Oliva, Oliver, and MacMill- to test for moderation here as well. Only “age” and
an 1992; Reichheld 1993; Hennig-Thurau and Klee 1997) “gender” showed no general moderating effect (Chi-
and early work on that relationship demonstrates a posi- square difference with four degrees of freedom were
tive and direct relationship. However, more recent studies 1.913 for age and 1.282 for gender). Therefore, we did not
argue that satisfaction can have a direct and indirect test for specific moderator effects. In summary, it appears
impact on loyalty (e.g., Hennig-Thurau 1997). that “loyalty-card membership,” “expertise,” “critical in-
cident,” and “income” are relevant general moderators of
Consistent with the notion of a non-linear relation- the link between satisfaction and loyalty.
ship (e.g., Mittal, Ross, and Baldasare 1998; Audrian
2002), some research suggests that the satisfaction-loyal- Next, we analyzed specific moderator effects of the
ty relationship is influenced by certain moderator vari- two satisfaction dimensions on the two loyalty dimen-
ables (Jones and Sasser 1995). The objective of this study sions for the moderators. Loyalty-card membership im-
is to provide additional insight into the relationship be- pacts only one out of the four effects. It moderates the link
tween customer satisfaction and customer loyalty by between “satisfaction with employees” and “positive word-
empirically examining the effects of selected moderators of-mouth.” Those customers who do not hold a loyalty
(loyalty-card membership, critical incidents, expertise, card seem more likely to relate to their positive experienc-
gender, age, income) on this relationship. es with the retailer’s employees to others. Apparently, the
loyalty card holders seem to expect employees to be
The sample consisted of 776 customers of a large friendly and competent and by doing that, expectancy
DIY chain. Testing for moderation, we first looked at a level of that factor is high. This result questions the
non-restricted model and then restricted the four paths appropriateness of investing in loyalty-card programs.
from satisfaction to loyalty to be equal across subgroups. Our finding is in line with findings from Reinartz and
Chi-square differences (with four degrees of freedom) Kumar (2002) who found little support for the general
were assessed. Next, we compared two models that only positive link between loyalty and profitability.
differed in one effect of one satisfaction dimension on one
loyalty dimension. One model restricts the parameter to be Three effects are moderated by “critical incidence.”
equal across groups while the second model allows vari- It is worth mentioning that we only looked at those critical
ation in one of that parameter across groups. The restricted incidents that were resolved to the satisfaction of the
model has one degree of freedom more than the general customer, meaning we are looking at “recovered custom-
model. A moderating effect would be present when the ers.” With that in mind, it can be noted that recovered
improvement in Chi-square moving from the restricted to customers – if satisfied with the employees – are far more
the non-restricted model is significant, meaning the Chi- loyal than customers who did not experience a critical
square difference between the two models (and one de- incident. In addition, the link between “satisfaction with
gree of freedom) is larger than 3.84 (p = .05). the assortment/tangibles” and “positive word-of-mouth”
is also moderated by a positively resolved critical inci-
After confirming the influence of the four postulated dent. Interestingly, “expertise” moderated only the link
main effects, we tested for moderator effects. A Chi- between “satisfaction with assortment/tangibles” and re-
square difference test was conducted for all six possible purchase intention. It is economically highly relevant that
moderator effects, comparing a restricted and a non- lucrative professional customers (e.g., craftsmen such as

carpenters or plumbers) buying their products at the DIY mensions are equally important to men and women and all
retailer are far more likely to repurchase if satisfied with age-groups in their effect on customer loyalty. A reason
the assortment. for that could be that DIY stores usually offer low-
involvement goods that are normally purchased less im-
The analysis of income as a moderator shows mixed pulsively. Particularly for certain high-involvement prod-
results. The link between satisfaction with the assortment ucts, we would expect a moderating role of gender and age
and positive word-of-mouth is stronger for wealthier on the satisfaction-loyalty link. Despite the fact that we
customers. Moreover, the link between satisfaction with found no support for the moderating effect of gender and
employees and both loyalty dimensions is stronger for age, there is sufficient evidence for a non-linear relation-
customers with lower income. It seems that high-income ship between the two constructs “satisfaction” and “loy-
consumers are more concerned with the availability of a alty.” Our research suggests that the two satisfaction
wide variety of products and less concerned with the dimensions, “satisfaction with assortment/tangibles” and
friendliness and competence of the employees. “satisfaction with employees,” positively influence the
two loyalty outcomes “repurchase intention” and “posi-
The link between satisfaction and loyalty is less tive word-of-mouth.” In particular, we found that “loyal-
straightforward and more complex than previous studies ty-card membership,” “expertise,” “critical incident,” and
have suggested. Gender and age do not moderate the link “income” are relevant general moderators of the link
from satisfaction to loyalty. Contrary to previous findings between satisfaction and loyalty. References and tables
(e.g., Homburg and Giering 1999), the satisfaction di- with results available upon request.

Heiner Evanschitzky
Marketing Center Muenster
University of Muenster
Am Stadtgraben 13–15
D-48143 Muenster
Germany
Phone: +49.251.83.22036
FAX: +49.251.83.22032
E-Mail: evanschitzky@ifhm.de

THE INTERPLAY OF COGNITION AND AFFECT IN THE FORMATION
OF CUSTOMER SATISFACTION: A DYNAMIC PERSPECTIVE
Nicole Koschate, University of Mannheim, Germany
Wayne D. Hoyer, The University of Texas at Austin, Austin
SUMMARY and that the ability of both cognition and affect to predict
customer satisfaction should increases.
Previous research has recognized that both cognition
and affect significantly predict satisfaction judgments. The results of an experimental study based on a real
However, only a few studies have investigated cognitive consumption experience indicate that the impact of cog-
and affective antecedents of customer satisfaction simul- nition on the satisfaction evaluation increases over time
taneously. Moreover, these studies have been static (i.e., and that the influence of affect decreases. Moreover, these
cross-sectional) in nature. This represents a significant effects are more pronounced in the case of consistent
shortcoming since it is well established that customer performance experiences. Finally, the study shows that
satisfaction is a dynamic phenomenon. Despite the strong the variance in customer satisfaction jointly explained by
recognition that customer satisfaction should be viewed cognition and affect increases as experience accumulates.
from a dynamic perspective, the role of cognitive and The findings have several important implications for
affective influences has not been systematically studied in marketing managers. For example, the study helps man-
this manner. The few studies which have investigated the agers to understand customer satisfaction in a more thor-
antecedents of customers satisfaction from a dynamic ough way. It sheds light on the formation process of
perspective have focused on the cognitive component of customer satisfaction and shows that customers satisfac-
customer satisfaction. tion has a more stochastic character in early stages. Thus,
it is more easily for managers to influence the satisfaction
Against this background, this study provides a dy- judgment in the early stages where the satisfaction evalu-
namic analysis of the simultaneous influence of cognition ation has not yet been crystallized. Further, it is common
and affect in the satisfaction formation process. The in practice for managers to think of customer satisfaction
fundamental proposition is that the role of cognition and in a logical, rational manner (i.e., if the product or service
affect may change over time. We argue that affect plays its performs well, satisfaction will be higher). The results of
strongest role at the early stages of satisfaction develop- this study point out that affect can play a critical role as
ment, whereas the impact of cognition should increase well, particularly in the early stages of the satisfaction
over time. In addition, we build a case that this phenom- formation process. Thus, for new relationships or new
enon also depends on the level of consistency of the products and services managers must pay close attention
consumption experience (i.e., is it consistently positive or to affective aspects and be careful to manage them effec-
consistently negative). Furthermore, we argue that satis- tively.
faction judgments should become more stable over time

Christian Homburg
Institute for Market-Oriented Management (IMM)
68161 Mannheim
Germany
Phone: +49.621.181.1555
FAX: +49.621.181.1556
E-Mail: homburg@bwl.uni-mannheim.de

MARKETING SIX SIGMA: ZERO DEFECTS IN INTERCULTURAL
SERVICE QUALITY
Martin C. Reimann, Freiberg University, Germany
Ulrich F. Luenemann, California State University, Sacramento
ABSTRACT were the two factors integration of the external factor and
service quality variation. It was assumed that, as a result
The Six Sigma methodology, traditionally referring of cultural value differences between the service provider
to defect reduction and quality improvement in manufac- and the customer, the integration of the external factor
turing, can also be applied successfully to marketing. could be defective. This might be based on the perception
Some service quality defects are related to intercultural of the customer, who measures the service’s quality
differences, especially when taking the direct integration according to his/her own specific cultural values while the
of external factors – international customers – into ac- service provider might base the service on another culture
count. In a study involving German, Spanish, and Swed- value. Thus, the integration of external factor – the inter-
ish customers, the authors found that people from cultures national customer – is defective. Since the perceived
with a high degree of uncertainty avoidance were less service quality varies based on cultural value differences
satisfied when their service expectations were not met. and there is a dependency between service quality and the
This suggests that Six Sigma can reduce service quality integration of the external factor (Hutchens 1989; Stamatis
defects related to intercultural differences if preceded by 1996), another assumption was that if the integration of
sound intercultural operative planning and training of the external factor was defective, the level of perceived
service personnel. service quality would be low.
INTRODUCTION In this paper, different cultural values and their im-

pact on international service marketing will be explained
Six Sigma, originally developed in Japan and adapted first. Subsequently, the two relevant service characteris-
at Motorola in the late 1980s to eliminate waste by tics – the integration of the external factor as well as
achieving near-perfect results in production processes, is quality variation – will be examined with regard to cus-
credited with saving billions of dollars also at other well- tomers responding differently to a delivered service and
known U.S. multinational companies, like General Elec- its quality. At this stage, cultural value differences and
tric and Allied Signals, just to name a few (Biolos 2002). their influence on customer and service provider interac-
Traditionally, Six Sigma refers to reduction of defects and tions will be taken into account. In this intercultural
improvement of quality in manufacturing; however, it can context, the Six Sigma methodology will be introduced.
be applied to marketing as well (Ehrlich 2002). Especially The study will show how Six Sigma can help to reduce the
services with high involvement of the customer, the Six defects in the integration of the external factor by focusing
Sigma methodology can help to improve service quality on quality variation. Thus, this paper suggests that Six
(Pande, Neuman, and Cavangh 2000; Biolos 2002; Ehrlich Sigma can be used as a workable tool for enhancing
2000). In this regard, Six Sigma can be defined as a data- international service marketing provided that it is built
based methodology or process to prevent service defects into the company’s operative programs and the service
and improve customer satisfaction. The statistical concept personnel has undergone adequate cultural awareness and
behind it represents the amount of process variation in intercultural preparation training.
relation to customer requirements as the source of all
service quality considerations. If a process functions on CULTURAL VALUES
the Six Sigma level, variation is almost non-existent.
Thus, the process results are defect-free in 99.9997 per- The growth and spread of multinational business on
cent of all cases, which means only 3.4 defects per million a global scale puts strong emphasis on the importance of
process steps. integrating cultural elements in international service mar-
keting. According to Hofstede (1997, p. 9), culture is “the
According to Kotler (2003), a service features the collective programming of the mind which distinguishes
following three characteristics: (a) the intangibility of the the members of one group or category of people from
service; (b) the integration of an external factor (an object another.” Thus, “culture is not a characteristic of individu-
or a subject on which the service is applied and which is als; it encompasses a number of people who were condi-
integrated into the service process); and (c) the quality tioned by the same education and life experience” (Hofstede
variation of the service. Of special interest for this paper 1997, p. 5). Values, the most basic manifestation of

culture, are defined as “broad tendencies to prefer a • the degree of long-term orientation (indicating the
certain state of affairs over others” (DeMooji 1997, p. 46). extent to which a society exhibits a pragmatic future-
Values are among the first things children learn, not oriented perspective rather than a conventional, his-
consciously but implicitly. Developmental psychologists toric short-time point of view.
consider that by the age of 10, most children have their
basic value system definitely in place and that changes These five cultural value dimensions (Hofstede 1997;
beyond this age are difficult to obtain (Rokeach 1973). Hofstede and Bond 1988) can be used to make important
Since people are not consciously aware of the values they predictions of intercultural differences in work-related
hold, it is difficult to discuss or observe them (DeMooji values including customer satisfaction.
1997, p. 46). Based on the receptivity to the idea of cultural
values as an important factor for organizational success, INTEGRATION OF THE EXTERNAL FACTOR
however, the need for intensified cultural value research – AND CULTURAL VALUE DIFFERENCES
especially for multinational companies – became widely
acknowledged. During the last two decades, many re- Cultural values influence how service providers in-
searchers have tried to make specific predictions of inter- teract with customers. This paper assumes that service
cultural differences and the related behaviors (Hall 1984; defects in the integration of the external customer will
Hall and Hall 1990b; Trompenaars and Hampden-Turner happen if different cultural values are not understood. The
1998; Schwartz 1999). The most comprehensive study to analytical ability to determine the uncertainty avoidance
date on cultural differences in work-related values was orientation seems to be the most important cultural value
done by Hofstede (1980, 2001), who obtained data from dimension that refers to defects in intercultural service
over 116,000 questionnaires answered by employees at quality. Uncertainty avoidance posits that humans reduce
all levels of a large U.S. multinational company. His their inherent uncertainty by dint of technology, law, and
research took place in more than 60 countries around the general rituals (Hofstede 2001, p. 147). As shown in
world over a period of six years. In his fundamental Figure 1, the degree to which uncertainty is generally
approach, Hofstede (1980) concentrated on four basic acceptable within a given culture can differ greatly from
dimensions of cultural values to which the selected coun- another (Hofstede 2001, p. 151). For this study, the
tries have found different answers in explaining. These German, Spanish, and Swedish cultures were selected due
dimensions were: to availability and easy access of data.
• the degree of power distance (indicating the extent to In his research, Hofstede (1980, 2001) compares low
which a society accepts the fact that power in institu- and high uncertainty avoidance in societies and uses the
tions and organizations is distributed unequally); degree of uncertainty avoidance to distinguish between
societal norms. With regard to peoples’ beliefs, attitudes,
• the degree of uncertainty avoidance (indicating the and behaviors, low uncertainty avoidance refers to: low
extent to which a society tries to avoid uncertain levels of stress and anxiety; weaker superegos and less
situations by, for example, establishing more formal showing of emotions; aggressive behavior is frowned
rules and believing in, and/or striving for expertise); upon; greater tolerance and acceptance of diversity and
uncertain situations; strong belief in general approaches
• the degree of individualism (indicating the extent to and common sense to problem solving, where people
which relationships are based on loose social frame- should be rewarded for innovative approaches; commit-
works rather than on collectivism, where people are ments are less binding and relationships are built quickly
tightly integrated in primary groups, such as families but can also be dissolved as quickly; focus on short-term
and organizations); planning (up to five years); rules and laws should be
adaptive and changed if they don’t work; more acceptance
• the degree of masculinity (indicating the extent to of dissent; and willingness to take unknown risk.
which dominant values or roles in society are viewed
“masculine,” for example achievement, assertive- On the other hand, high uncertainty avoidance refers
ness and performance, when measured against its to: higher stress levels and an inner urge to be busy; robust
opposite pole, femininity, defined as quality of life, superegos and more showing of emotions; aggressive
caring for other people as well as social and gender behavior of self and others is accepted; less tolerance and
equality (Hofstede 1980, 1997, 2001). acceptance of unclear situations; less acceptance of dis-
sent and strong need for consensus, clarity and structure;
Based on their research in Asia, Hofstede and Bond strong belief in expertise and knowledge for problem
(1988) found a new dimension, which was later added to solving, where accuracy should be rewarded; commit-
Hofstede’s (1997, 2001) research as a fifth dimension and ments are long-lasting and relationships, which are built
labeled: slowly, are expected to last for a long time; focus on long-

FIGURE 1
Country UAI*
Greece 112 Colombia 80 Norway 50
Portugal 104 Brazil 76 New Zealand 49
Guatemala 101 Venezuela 76 South Africa 49
Uruguay 100 Italy 75 Canada 48
Belgium 94 Czech Republic 74 Indonesia 48
El Salvador 94 Austria 70 United States 46
Poland 93 Pakistan 70 Philippines 44
Japan 92 Taiwan 69 China 40
Peru 87 Arab World 68 India 40
Argentia 86 Ecuador 67 Malaysia 36
Chile 86 Germany 65 Ireland 35
Costa Rica 86 Thailand 64 United Kindom 35
France 86 Finland 59 Hong Kong 29
Panama 86 Iran 59 Sweden 29
Spain 86 Switzerland 58 Denmark 23
South Korea 85 West Africa 54 Jamaica 13
Turkey 85 Netherlands 53 Singapore 8
Hungary 82 East Africa 52
Mexico 82 Australia 51
Israel 81
* UAI = Uncertainty Avoidance Index
225
term planning (up to 20 years); strong need and adherence process variables, which will then be measured, ana-
to rules and regulations to make behavior predictable; lyzed, improved, and controlled. The Measurement
concern with security in life; and knowing about risks phase comprises the setup of a capable measurement
(Adler 1997, p. 53; Hofstede 2001, p. 161). Based on these system to measure the dependent variables: in this
general comparisons, service providers can and should case customer satisfaction with regard to a specific
use Hofstede’s (1980, 2001) uncertainty avoidance index service. In the Analysis phase, the independent vari-
to integrate the external service factor in their operations ables are assessed, which was the service variable
by defining service quality for international customers in “delivery time” in this study. Then, in the Improve-
terms of cultural awareness and intercultural prepared- ment phase, the value of the independent variables
ness of service employees in addition to their obvious will be increased. Finally, the Control phase is neces-
business and organizational skills. A specially tailored sary to review the measurement system and the cor-
intercultural training – including service factors such as rectness of its outputs (Pande, Neuman, and Ca-
reliability, responsiveness, competence, courtesy, credi- vanagh 2000, p. 39).
bility, security, access, communication, and understand-
ing the customer – can provide appropriate and useful • Customer Satisfaction Measurement: Six Sigma aims
approaches for adaptation to international customers’ at the achievement of fullest customer satisfaction by
different value systems and behaviors. This is especially providing a defect-free process or service. As already
true when the training focuses on the aforementioned mentioned above, customer satisfaction can be se-
uncertainty avoidance differences as the result of different lected as dependent target variable, which is influ-
cultural backgrounds. Therefore, to achieve a high level enced by one or more process drivers. Its measure-
of external factor integration, service providers should ment should be done before as well as after the
strive continuously to improve the level of customer process improvement to compare and view progress.
satisfaction, which should include intercultural under-
standing and display of correct and appropriate behavior • Cultural Awareness and Preparation Training: Six
towards international customers. Based on the cultural Sigma also integrates change management. Improv-
differences in Hofstede’s uncertainty avoidance dimen- ing a process also means changing human behavior in
sion, it can be hypothesized: H1: Customers from cultures organizations to minimize defects. In international
with a high degree of uncertainty avoidance will be less service marketing, operational planning for special
satisfied when their service expectations are not met. cultural awareness as well as preparation training on
Consequently, in order to achieve customer satisfaction intercultural differences is necessary and especially
with people from high uncertainty avoidance cultures, the essential for the integration of the external factor
service provider needs to meet all relevant service expec- when service providers and customers have different
tations. cultural backgrounds and opposing views on how to
deal with uncertainty avoidance. As explained be-
QUALITY VARIATION AND SIX SIGMA fore, the differences between low and high degrees of
uncertainty avoidance can be so severe between
Quality variation characterizes service as well. The people from diverse cultural backgrounds that under-
central source of quality variation is the defective integra- standing their behavior and being aware of their
tion of the external factor (Hutchens 1989; Stamatis different perceptions are very important to prevent
1996). Among the reasons for defective customer integra- irreparable service defects. Therefore, proactive in-
tion are some of the following service personnel short- tercultural communication training programs based
comings: lack of responsiveness and timeliness, missing on Hofstede’s (1980, 2001) uncertainty avoidance
competence and courtesy, miscommunication and faulty dimension have to emerge as critical events in the
understanding of the customer (Heineke and Davis 1994). development of the international service provider’s
Kotler (2003) identifies three actions to overcome quality management strategies to improve service quality.
variation: process improvement, customer satisfaction Such training can effectively enhance future service
measurement, and service personnel training. All three quality both for service providers and customers in
actions can be achieved with the Six Sigma methodology: overseas assignments, as well as in multicultural and
ethnically diverse domestic settings. The information
• Process Improvement: Six Sigma not only implies gained and behavioral skills learned will not only
statistics; it also uses managerial tools for process help to prevent service defects, but also enable the
improvement. A Six Sigma project usually follows service provider to perform on a much higher quality
the so called D-M-A-I-C approach (Define-Measure- level. Thus, customers will experience the service
Analyze-Improve-Control). In the Definition phase, function on a much higher satisfaction level.
customer requirements are surveyed, potential sav-
ings are evaluated and the process is mapped. The There are basically two ways to achieve a Six Sigma
customer requirements directly lead to the relevant level. One way is through reduction of scattering, the

other is through expansion of tolerance. As shown in ened as much when encountering deviations from their
Figure 2, the defects in the upper normal distribution refer normative expectations. To use the same hotel example, if
to a three sigma level with a possible yield of 93.32 customers with a low degree of uncertainty avoidance do
percent. not get the expected and/or reserved room at check-in,
they are more likely to accept a wider range of alterna-
The two lower distributions refer to a six sigma level tives. Thus, a possible conflict can be contained on the
with a 99.9997 percent yield. While the reduction of level of fairness, flexibility, and common sense without
scattering focuses on minimizing variance, the expansion leading to a service defect. Even if all service expectations
of tolerance focuses on the customer-related requirements are not met, such customers are still able to come across
(the acceptance of a longer-than-promised delivery time). with a positive and satisfying service experience. If con-
fronted with such a situation, the service provider can
Achieving Six Sigma Through Reduction of Scat- safely assume that the customers’ tolerance level is high
tering: As the description of differences between low and enough to allow for certain deviations from the expected
high uncertainty cultures has shown, customers coming service and still perceive a high quality service. As a
from cultures with a relatively high degree of uncertainty result, it can be hypothesized for this study: H3: If custom-
avoidance have a much lower tolerance for ambiguity. ers posses a low degree of uncertainty avoidance, then a
They do not accept unclear situations and deviation from wider tolerance with regard to the promised delivery time
the normal variation as easily as customers coming from will be accepted. Consequently, the expansion of toler-
cultures with a relatively low degree of uncertainty avoidance could be an adequate instrument and part of the
ance. A high uncertainty avoidance index generally also service provider’s operational planning to meet the re-
indicates higher anxiety and stress levels, a greater pro- quirements of customers with a low degree of uncertainty
pensity to display emotions, and a tendency towards avoidance.
aggressive behavior when challenged. If a service provid-
er interacts with such an external factor, there is only very STUDY AND HYPOTHESES TESTING
little chance to prevent a service defect if the customer
encounters a situation or behavior that does not conform To test all three hypotheses, a study was carried out
to the customer’s cultural and normative expectations. within a Six Sigma project of a global company from the
For example, if hotel customers with a high degree of chemical industry. In a questionnaire, which was filled out
uncertainty avoidance do not get the expected and/or immediately after a specific service was delivered, 500
reserved room at check-in, they might not accept any customers of three different national cultures were asked
alternate arrangements without aggressive, emotional, about a certain service quality as well as their satisfaction
and stressful behavior which, in turn, may lead to a full- level with regard to only this service. Of the 303 received
size service defect. In such a situation, the service provid- responses (equaling a response rate of 60.6%), 34 percent
er has to consider the customers’ level or degree of were from Germany, another 34 percent from Spain, and
tolerance to be so narrow that any deviation or scattering 32 percent from Sweden. Using the Six Sigma methodol-
from the promised or expected service would automati- ogy in this study, customer satisfaction was treated as the
cally lead to the customers’ perception of low service dependent variable, whereas a culturally varying degree
quality or even total dissatisfaction. Therefore, and within of uncertainty avoidance, which influences customer ser-
the scope of the specific study, it can be hypothesized: H2: vice quality satisfaction, was the independent variable.
If customers posses a high degree of uncertainty avoid- The study had the following parameters: it used a precise
ance, then a wider tolerance with regard to the promised delivery time of 240 hours to measure the degree of
delivery time is not accepted. Consequently, the reduction customers’ service expectation; both data streams were
of scattering (or upgrading of service) should be an discrete: the degree of customer satisfaction (from 1 =
adequate instrument and part of the service provider’s “very satisfied” to 5 = “not satisfied”) as well as the degree
operational planning to meet the requirements of custom- of uncertainty avoidance; and it featured three cultures on
ers with a high degree of uncertainty avoidance. three different degrees of uncertainty avoidance (1 =
“high,” 2 = “medium,” and 3 = “low”). The question asked
Achieving Six Sigma Through Expansion of Toler- regarding customer satisfaction was: “In compliance with
ance: Contrary to the previous example, customers com- the achieved delivery time for this delivery, were you
ing from cultures with a relatively low degree of uncer- satisfied with our service?” According to Hofstede (2001,
tainty avoidance have a much higher tolerance for ambi- p. 151), and as shown in Figure 1, Spain ranks fairly high
guity. They see uncertainty as an inherent part of life and on uncertainty avoidance, Germany takes a rather medi-
accept more easily each situation as it comes. A low um position, and Sweden has a fairly low uncertainty
uncertainty avoidance index generally also indicates that avoidance index. Based on these different ranking posi-
people are more at ease, show less emotions and frown tions, the study was able to compare the relevant degrees
upon aggressive behavior. Based on their higher level of of uncertainty avoidance in the target cultures. Given the
tolerance, they are more flexible and do not feel threat- same performance level, hypothesis H1 predicted that

FIGURE 2
customers from cultures with a high degree of uncertainty 71.9%; the p-value is always below 0.05) between all
avoidance, in this case Spain, will be less satisfied then three variables.
customers from Sweden when their service expectations
were not met. To test this hypothesis, a Chi-Square-test When inserting the values for the degree of uncertain-
was applied. As shown in Figure 3, the test of hypothesis ty avoidance (“1” for Spain with a high degree, “2” for
H1 displayed a high chi-square (χ2 = 141.947) as well as Germany with a medium degree, and “3” for Sweden with
a p-value below 0.05 (p = 0.000). The high Chi-Square a low degree) as well as for delivery time (for example 238
resulted from combining the degree of uncertainty avoid- hours versus 245 hours) into the regression equation, it
ance with the level of customer satisfaction based on the showed that Spanish customers were less satisfied – and
outcome of the study. Therefore, the linkage between a much earlier – than their German and Swedish counter-
high degree of uncertainty avoidance and a certain satis- parts although the same level of service quality (slower
faction level was established. While 37.8 percent of the delivery time) was provided. Therefore, the central find-
Spanish customers were satisfied and 62.1 percent were ing of the study was that the degree of uncertainty avoid-
dissatisfied, there was almost the opposite picture for ance as a cultural variable has significant influence on
German customers (62.1% satisfied versus 14.6% dissat- customer satisfaction. As stated in hypothesis H2, it was
isfied). In Sweden, however, most customers were satis- found that customers from cultures with a high degree of
fied (85.5% satisfied versus 0.0% dissatisfied). In com- uncertainty avoidance do not accept a wider tolerance of
parison, this means that people in Spain were overall less quality variation as measured by the length of the time for
satisfied than those in Germany – and especially Swe- delivery. Customers from cultures with a low degree of
den – given the same service performance level. Thus, it uncertainty avoidance, however, do accept a wider toler-
was found that customer satisfaction is influenced by a ance as was hypothesized in hypothesis H3. Consequent-
culturally varying degree of uncertainty avoidance, which ly, H2 and H3 were also strongly supported.
strongly supports hypothesis H1.
LIMITATIONS
After that, hypotheses H2 and H3 were tested. As
already stated above, it was assumed that the customer Although this study provides a unique insight into the
with a lower degree of uncertainty avoidance would relationship between service quality, cultural value differ-
accept a wider tolerance in service quality, whereby a ences, and customer satisfaction, some limitations have to
fulfilled service would be referred to as a high service be highlighted. First, the study was conducted only in the
quality. In the study, a fulfilled service was the accom- chemical industry with a rather small sample of custom-
plishment of a specific product delivery in a certain ers. Second, although the observed cultures represented a
amount of time. The average delivery time was 240 hours. high, medium, and low degree of uncertainty avoidance,
Thus, it was defined that a faster or on-time delivery members of only three different cultures were inter-
(below or within 240 hours) would be perceived as high viewed. Thus, conducting more research in other cultures
service quality while a slower delivery (above 240 hours) is recommended. Third, the use of additional cultural
would be perceived as low service quality. This relatively variables, for examples Hall’s (1984) distinction of
small variation in delivery time referred to the “just-in- chronemics, which relates to the culturally different per-
time” production at the customer side. Taking longer than ception of time as a form of non-verbal communication, or
240 hours meant loss of production or sales since the economic differences in supply and demand situations of
customer was running empty on the chemical product the target cultures, may have changed or influenced the
delivered by the service provider. Thus, and according to study’s outcome. Fourth, the application of Hofstede’s
the Six Sigma methodology, a delivery taking longer than cultural value dimensions to further research has to be
240 hours was considered a defect in the service process. considered carefully. Critics argue that Hofstede’s model
The study found that customer satisfaction varied across lacks transferability and, therefore, is not representative
cultures even if the same variance within the service for a specific nation or a culture since only data from one
quality (slower delivery time) had been provided. The single U.S. company (IBM) were used as samples (Triandis
reason for this finding was the culturally varying degree 1982; Yoo and Donthu 1998). On the other hand, this also
of uncertainty avoidance. Respondents from cultures with ensures some consistency in the research since IBM
a higher degree of uncertainty avoidance were less satis- employees are somewhat similar in regarding organiza-
fied with the provided service quality than their counter- tional culture, job description, or educational level (Lowe
parts from cultures with a lower degree of uncertainty 1996). Overall, the above listed limitations show that
avoidance. As illustrated in Figure 4, it was found that further efforts must be made to understand the behavioral
service quality as well as cultural value differences (in this impact of cultural differences and, if possible, incorporate
case the degree of uncertainty avoidance) will influence other cultural values in future research models for more
customer satisfaction. The regression equation shows a complete explanations.
significant relationship (R2 = 72.1% and R2 adjusted =

FIGURE 3
χ2 = Σ (Observed Count – Expected Count)2

Expected Count
1 2 3 4 5
very satisfied not satisfied TOTAL
Spain Observed Count 19 20 0 23 41 103

(high degree Percentage 18.4% 19.4% 0% 22.3% 39.8% 100%
of uncertainty Expected Count 32.97 30.25 12.92 12.92 13.94
avoidance)
Germany Observed Count 32 32 24 15 0 103

(medium Percentage 31.1% 31.1% 23.3% 14.6% 0% 100%
degree of Expected Count 32.97 30.25 12.92 12.92 13.94
uncertainty
avoidance)
Sweden Observed Count 46 37 14 0 0 97

(low degree Percentage 47.4% 38.1% 14.4% 0% 0% 100%
of uncertainty Expected Count 31.05 28.49 12.17 12.17 13.13
avoidance)
TOTAL Observed Count 97 89 38 38 41 303
χ2 = 5.922 + 3.475 + 12.917 + 7.870 + 52.549 +

0.029 + 0.101 + 9.508 + 0.336 + 13.937 +
7.195 + 2.541 + 0.277 + 12.165 + 13.125
= 141.947
p = 0.000
230
FIGURE 4
DISCUSSION customer satisfaction measurement tool based on uncer-
tainty avoidance that there are major differences in the
For quite a while, cultural issues had been closely customers’ perceptions of service quality across cultures.
observed in terms of overall life and job satisfaction Subsequently, service process improvement should be
(Hofstede 2001). Previous research, however, has not applied according to the customers’ requirements in each
looked into the relationship between certain levels of country or culture (Fitzsimmons and Fitzsimmons 1994).
service quality satisfaction and cultural values. While The study’s findings clearly show that only a narrow
some researchers have recognized that international ser- service quality tolerance delivered in the service process
vice is people-centered and, therefore, culture must some- will be accepted by customers from cultures with a high
how play a role (Clark, Rajaratnam, and Smith 1996; De degree of uncertainty avoidance. This means that service
Ruyter, Wetzels, and Lemmink 1996; Dahringer 1991), managers have to especially plan and aim for a defect-free
their focus has not been on the impact of cultural values on process in high uncertainty avoidance countries like Spain
customer satisfaction in terms of a delivered service. in this case. One could also deduct from the findings,
Based on the study presented in this paper, it has been however, that cultural differences can play a role as a
shown that uncertainty avoidance – a very important moderator between lower service performance and cus-
cultural value – has a significant influence on global tomer satisfaction in low uncertainty avoidance countries
business and how customers from different cultures per- like Sweden.
ceive a certain service quality level differently. Given the
variance of uncertainty avoidance across cultures, the Therefore, and in addition to a stringent service
study also shows how global service marketing can im- process improvement, new operational service planning
prove efficiency and reduce customer dissatisfaction by for increased quality should include intense cultural aware-
adopting cultural education and training programs for ness and intercultural preparation training for all involved
service personnel. With regard to this study, the research service personnel. More specifically, training with the
implications demand a closer investigation of the cohe- main focus on Hofstede’s (1980, 2001) cultural dimen-
sion between cultural values and customer satisfaction in sion of uncertainty avoidance must be seen as critical part
terms of delivery time. For example, not only Hofstede’s of the Six Sigma methodology to achieve zero defects in
degree of uncertainty avoidance but also that of long-term intercultural service quality. The increased intercultural
orientation could have implications on the service of competence derived from such training not only gives the
delivery time. Furthermore, cultural value items of other multinational service provider the opportunity to adjust
researchers, such as Hall and Hall’s (1990a) degree of behavioral patterns accordingly, but can also help to
timing, which differentiates between cultures working relieve much of the normal anxieties experienced by
parallel on many tasks versus cultures working on one trying to integrate the external customer in a novel but
task at a time, might be taken into account as well. As satisfying cultural setting. It will also help to strive for a
suggested by Kotler (2003), one key characteristic in defect-free service process and enable the international
service marketing is quality variation. The managerial service provider to perform on the highest achievable
implications in service quality variations are straightfor- service quality level, which will result in an even higher
ward and can be achieved by applying the Six Sigma satisfaction level on the customers’ side.
methodology. The presented study illustrates through a
REFERENCES DeMooij, Marieke (1997), Global Marketing and Adver-

tising: Understanding Cultural Paradoxes. Thou-
Adler, Nancy (1997), International Dimensions of Orga- sand Oaks, CA: Sage.
nizational Behavior. Cincinnati, OH: South-West- De Ruyter, Ko, Martin Wetzels, and Jos Lemmink (1996),
ern. “The Power of Perceived Service Quality in Interna-
Biolos, Jim (2002), “Six Sigma Meets the Service Econ- tional Marketing Channels,” European Journal of
omy,” Harvard Management Update, 7 (11), 10. Marketing, 30 (12), 22–38.
Clark, Terry, Daniel Rajaratnam, and Timothy Smith Ehrlich, Betsie H. (2002), Transactional Six Sigma and
(1996), “Towards a Theory of International Services: Lean Servicing. Boca Raton, FL: St. Lucie Press.
Marketing Intangibles in a World of Nations,” Jour- Fitzsimmons, James A. and Mona J. Fitzsimmons (1994),
nal of International Marketing, 4 (2), 9–28. Service Management for Competitive Advantage.
Dahringer, Lee D. (1991), “Marketing Services Interna- New York: McGraw-Hill.
tionally: Barriers and Management Strategies,” Jour- Hall, Edward T. (1984), The Dance of Life. Garden City,
nal of Service Marketing, 5 (3), 5–17. NY: Doubleday.

____________ and Mildred Hall (1990a), Hidden Differ- ment in Hong Kong,” Asia Pacific Business Review,
ences: Doing Business with the Japanese. New York: 2 (1), 120–33.
Anchor-Doubleday. Pande, Peter S., Robert P. Neuman, and Roland R. Ca-
____________ and Mildred Hall (1990b), Understand- vangh (2000), The Six Sigma Way: How GE, Motor-
ing Cultural Differences: Germans, French, and ola, and Other Top Companies Are Honing Their
Americans. Yarmouth, ME: Intercultural Press. Performance. New York: McGraw Hill.
Hofstede, Geert (1980), Culture’s Consequences: Inter- Rokeach, Milton (1973), The Nature of Human Values.
national Differences in Work-Related Values. Bever- New York: The Free Press.
ly Hills, CA: Sage. Schwartz, Shalom H. (1999), “Cultural Value Differenc-
____________ and Michael Bond (1988), “The Con- es: Some Implications for Work,” Applied Psycholo-
fucius Connection: From Cultural Roots to Econom- gy International Review, 23.
ic Growth,” Organizational Dynamics, 16 (4), 4–21. Stamatis, Dean H. (1996), Total Quality Service. Delray
____________ (1997), Cultures and Organizations: Soft- Beach, FL: St. Lucie Press.
ware of the Mind. New York: McGraw-Hill. Triandis, Harry C. (1982), “Culture’s Consequences,”
____________ (2001), Culture’s Consequence: Com- Human Organization, 41 (1), 86–90.
paring Values, Behavior, Institutions, and Organiza- Trompenaars, Alfons and Charles Hampden-Turner
tions Across Nations. Thousand Oaks, CA: Sage. (1998), Riding the Waves of Culture. New York:
Hutchens, Spencer (1989), “What Customers Want: Re- McGraw-Hill.
sults of ASQC/Gallup Survey,” Quality Process, Yoo, Boonghee and Naveen Donthu (1998), “Validating
(February), 29–35. Hofstede’s Five-Dimensional Measure of Culture at
Kotler, Philip (2003), Marketing Management. Engle- the Individual Level,” Proceedings of the 1998 Sum-
wood Cliffs, NJ: Prentice-Hall. mer Marketing Educators’ Conference of the Amer-
Lowe, Sid (1996), “Culture’s Consequences for Manage- ican Marketing Association, Boston, MA.

Martin C. Reimann
Marketing and International Trade
Freiberg University
Lessingstrasse 45
Freiberg, Germany 09596
Phone: +49.3731.392004
FAX: +49.3731.394006
E-Mail: martin.reimann@bwl.tu-freiberg.de
Ulrich F. Luenemann
Department of Communication Studies
California State University, Sacramento
6000 J Street
Sacramento, CA 95819
Phone: 916.278.6688
FAX: 916.929.1638
E-Mail: calulli@saclink.csus.edu

THE IMPACT OF EXPERIENTIAL KNOWLEDGE AND CREATIVITY ON
PERFORMANCE OF INTERNATIONAL PROJECT
Taewon Suh, Texas State University, San Marcos
Hongxin Zhao, Saint Louis University, St. Louis
Seung H. Kim, Saint Louis University, St. Louis
Mark J. Arnold, Saint Louis University, St. Louis
Mueun Bae, Inha University, Republic of Korea
SUMMARY edge as intellectual market-based assets was assessed at

the multilevel. Both team-level and firm-level experien-
This study focuses on testing the relationships be- tial knowledge were identified as unique, independent
tween the constructs of knowledge and creativity of an constructs, exerting important roles in the proposed mod-
international project team and their consequential perfor- el. Particularly, team-level experiential knowledge was
mance. More specifically, this study attempts to fill in the significant on knowledge implication and transfer, and
research void by (1) establishing a research model at the firm-level experiential knowledge played importantly on
project-team level; (2) centering on the crucial factors the outcomes. Third, this study differentiated the two
concerning knowledge and knowledge-creation for the creativity constructs and their unlike influences in the
success of the international project (i.e., experiential knowl- modeled relationships. With the dual conceptualization, it
edge and creativity); and (3) constructing a structural is quite effectively questioning the seemingly uncon-
model comprising the multiple relationships between the scious assumption that creativity is a unitary construct.
explaining factors and performance. In the research mod- Fourth, this study has confirmed the fact that project
el, the constructs of experiential knowledge measured at performance is nested in the organization recognizing the
both the firm- and the team-level make the exogenous need for a multi-level study using the broad framework of
variables in the model. Experiential knowledge as pre- the resource-based views.
sented is associated with team creativity, project creativ-
ity, and project performance. Team creativity, subse- For better performance of a project, we can find two
quently, influences the other two endogenous variables, important factors according to the current research set-
project creativity and project performance. And, project ting: firm-level experiential knowledge and team creativ-
creativity is associated with project performance. ity. MNCs should accumulate the domain-specific knowl-
edge at the firm level and encourage creative behavior by
Hypotheses are tested through structural equation nurturing supportive environments and organizational
modeling using Korean MNCs’ sample. Hypothesis 1, culture. First, in terms of the experiential knowledge,
predicting that Team-Level Experiential Knowledge (EKT) knowledge accumulation would better be focused on
positively affects Team Creativity (TC), was supported developing routines and structures to manage operations
(EKT Î TC: t = 5.62, p < .001). The influence of firm- since such routines and processes are not sensitive in
level experiential knowledge on creativity was presented terms of geographic application. Second, summarizing
in Hypothesis 2. Firm-Level Experiential Knowledge the literature, the following actions will prove beneficial
(EKF) is not significantly associated with Project Cre- for better team creativity. Encourage employees to ex-
ativity (PC) (EKF Î PC: t = .93, p > .05). As predicted, press their ideas openly; provide help in developing ideas;
TC was highly associated with PC (t = 3.89, p < .001), provide time for individual efforts; encourage risk taking
which supports Hypothesis 3. Hypothesis 4 and 5 were and initiative; provide freedom for employees to enable
also held up. Project Performance (PP) was significantly them to do things differently; provide a non-punitive
influenced by EKF (t = 5.27, p < .001) and TC (t = 2.72, environment using a low level of supervision; encourage
p < .01). However, as the impact PC on PP failed to team members to interact and participate with other groups
achieve significance (t = -.80, p > .05), Hypothesis 6 was besides their own; maintain an optimal amount of work
not supported. pressure; provide realistic work goals; encourage the
delegation of responsibilities; demonstrate confidence in
Theoretical contributions of this study are summa- the workforce in a climate of mutual respect; allow indi-
rized to few points. First, the study first used creativity viduals to be part of the decision-making process; encour-
construct in the international business setting, investigat- age management to provide immediate and timely feed-
ing the structured relationships with experiential knowl- back to their team members.
edge and performance. Second, related, experiential knowl-

Taewon Suh
McCoy College of Business Administration
Texas State University
601 University Drive
San Marcos, TX 78666
Phone: 512.245.3239
E-Mail: ts21@txstate.edu

CHANGE AND THE MARKETING ORGANIZATION
Kelly D. Martin, Washington State University, Pullman
Jean L. Johnson, Washington State University, Pullman
SUMMARY the alpha type occurs within a fixed and stable system and
can be accurately measured with reliable dimensions.
More than ten years ago, Achrol’s (1991) work on the Beta change also involves periphery change within a fixed
evolution of the marketing organization envisaged the system, but is complicated by the fact that the dimensions
dynamic environment in which marketers exist today. He of measurement or calibration have changed as well. Beta
portends, “The future will be characterized most notably change may be likened to changes occurring in a stable
by unprecedented levels of diversity, knowledge rich- system, however fluctuating “rubber yardsticks” provide
ness, and turbulence” (Achrol 1991, p. 77). Our discipline unreliable measurement. We extend the discussion of this
has witnessed Achrol’s prophecies come to fruition, as the type of change by proposing two distinct forms of beta
marketing environment is characterized by unprecedent- change. The typology has relevant application to a broad
ed dynamism and turbulence. Broad social and cultural spectrum of change affecting marketing strategy, and we
change in the form of technological intensity, globaliza- highlight some examples for illustration.
tion and global outsourcing, and regulatory and political
unrest is illustrative of our marketing landscape. There- In response to the unique impact of change, a few key
fore, to extend the predictions of Achrol into the present, organizational phenomena will combine to determine
and in the spirit of “understanding diverse and emerging marketing’s ability to overcome change. Three relevant
markets, technologies, and strategies,” our aim is to ad- factors instrumental to the marketing function’s change
vance appreciation of change and its role in the marketing response include opacity, asperity, and intricacy. Re-
organization. search has demonstrated these variables’ impact on an
organization’s ability to manage and ultimately survive
Little is known about the effects of broad social, significant change (Hannan, Pólos, and Carroll 2003).
cultural, and strategic change on the marketing organiza- Opacity describes phenomena at the individual level,
tion in particular. Further, no commonly accepted frame- specifically concerning marketing managers’ perceptions,
work exists by which to categorize and thus better under- actions, or possible oversights. In particular, we consider
stand specific types of change in marketing. As a founda- opacity as marketing mangers’ inability to foresee, effec-
tion, we draw from existing literature in organizational tively comprehend, and thus plan accordingly in the face
theory on change. Most organizational theorists agree that of change. Asperity describes normative, cultural phe-
change may occur either incrementally and peripherally nomena affecting large groups or subgroups within the
which causes a small scale impact, or radically by altering organization. Within marketing, interrelatedness between
the core functioning of the organization. Because this normative structures and organizational culture makes the
first-order or periphery change is the most predominant marketing function more or less susceptible to failure in
type of change to be experienced by organizations (Fox- light of change. Finally intricacy describes the overall
Wolfgramm, Boal, and Hunt 1998), it is surprising that organizational design, specifically the degree of intercon-
contemporary research has made little attempt to refine nectedness between relevant organizational units. Al-
this conceptualization further. One important exception is though highly interconnected organizations may promote
the typology introduced by Golembiewski, Billingsley, positive communication structures and decentralized de-
and Yeager (1976), which delineates three distinct cate- cision making for example, organizations that are too
gories of change. This typology is considered a relevant highly interconnected may lose flexibility. Achrol (1991)
lens through which to view marketing specific change, emphasizes that the impact of change will be intensified in
and has been usefully employed in other marketing re- densely interconnected and interdependent settings. We
search (e.g., Cooper 2000). apply these organizational phenomena to the marketing
discipline with specific descriptions and examples of each
Similar to the radical change classifications of core opacity, asperity, and intricacy.
and second-order change, the typology considers gamma
change. The authors define gamma change as a “quantum Our research marries these marketing organizational
shift in ways of conceptualizing salient dimensions of phenomena to the various conceptualizations of change.
reality,” (Golembiewski et al. 1976, p. 138). Alpha change Just as alpha, beta, and gamma changes are defined so
involves smaller scale change such as that described by differently, it follows that their impact on opacity, asper-
the terms periphery, non-core, or first-order. Change of ity, and intricacy will fluctuate and vary. In worst-case

situations, these variables combine in such a way to divert ring as a result of missed opportunities. The ultimate
marketing managers’ time and attention away from op- consequence of missed opportunities and subsequent re-
portunities and thus potential resource generating activi- source depletion is marketing failure or even death. It
ties (Hannan et al. 2003). Thus, missed opportunities follows that understanding how specific combinations of
become the key, detrimental outcome of change misman- opacity, asperity, and intricacy impact the marketing
agement in the marketing organization. Depletion of organization’s ability to capitalize on, rather than neglect,
resources through missed opportunities can ultimately important resource-generating opportunities is impera-
drain the life from a marketing organization. Only signif- tive. We consider the consequences for other strategic
icant caches of resources accumulated prior to the change marketing outcomes, as well as overall firm performance.
can possibly offset substantial resource depletion occur- References are available from the authors upon request.

Kelly D. Martin
College of Business and Economics
Washington State University
Pullman, WA 99164–4730
Phone: 509.335.5848
FAX: 509.335.3865
E-Mail: kdmartin@mail.wsu.edu

PRIVACY CONCERNS AND CUSTOMERS’ WILLINGNESS TO PROVIDE
INFORMATION: A REVIEW WITH IMPLICATIONS FOR
FUTURE RESEARCH
Mona Srivastava, Texas A&M University, College Station
Robert Harmon, Portland State University, Portland
SUMMARY even the customer’s awareness or permission, and has

also helped make the provider, delivery process, and
Organizations and customers are engaged in a battle process controls invisible especially in an online environ-
over information. Too much information being collected ment, which has had a negative effect on trust. We posit
has raised privacy concerns amongst customers whereas an inverse relationship between privacy concerns and
too little information hampers organizations in their ef- trust which may be accentuated or attenuated depending
forts to create long-term customer relationships. Although the organizations’ need-to-know that information. Fur-
society may force privacy regulations on marketers through ther trust is the mediating factor in the relationship be-
governmental action or legal initiatives, we believe that a tween privacy concerns and the customer’s motivation for
market-based solution focused on educating organiza- a relationship with the organization.
tions on the benefits of finding the right balance between
too much and too little information would alleviate cus- Organization’s Need-to-Know: When approached
tomers’ privacy concerns as well as translate into cost- by an organization for information customers are con-
savings for organizations. This paper reviews extant liter- cerned about how this information will be used, can the
ature on the factors that affect the customer’s willingness organization can be trusted with this information and most
to provide information to organizations and explores importantly, whether the organization needs to have this
implications for future research to draw managerial atten- information for better satisfying its customers’ needs?
tion towards the ethics of collecting customer informa-
tion. The potential relationships between the factors that P1: The relationship between the customer’s privacy
impact the customer’s willingness to provide information concerns and the customer’s trust in an organization
to organizations are detailed below. is moderated by the organization’s need-to-know
that information.
Customer’s Privacy Concerns: Sieber (1998, p.
136) has defined privacy as a person’s “degree of control Customer’s Motivation for a Relationship: Bendapu-
of the access that others have to them and to information di and Berry (1997) argue that individuals participate in
about them.” Customers are increasingly concerned about relationships either because they want to or because they
retaining control over their personal data, and want orga- have to. Intrinsic desires like dedication drives high mo-
nizations to reward them for giving information and tivation for a customer relationship and such relationships
allowing for its use (Hof 2001). The degree of privacy are based on trust as well as dependence. However, when
concern differs as per the type of information (Nowak and customers are constrained to maintain a relationship,
Phelps 1992, 1995), type of industry, culture, age, and perhaps due to lack of alternatives, then dependence,
gender (Petrison and Wang 1995; Milne and Boza 1999; without trust, leads to a low motivation for a customer
Sheehan 1999). We posit that privacy concerns directly relationship. We believe that customers with a high moti-
impact trust. Since the role of trust in the relational context vation for a relationship would be more likely to acquiesce
has been acknowledged as crucial, it is important to study to an organization’s request for information, providing
the impact of privacy concerns on trust (Regan 2003; better quality and/or quantity of information, to achieve
Fonseca and McCarthy 2003; Cavoukian et al. 2002; Miln mutual benefits.
and Boza 1999).
P2: The customer’s trust in an organization increases the
Customer’s Trust in an Organization: Some of the motivation of the customer to have a relationship
factors that may determine the degree to which customers with an organization.
are willing to trust an organization include who their
personal information is shared with, how it is used, what P3: Customers with a higher motivation for a relationship
is the cost-benefit of sharing their information, and the with an organization would be more willing to pro-
reliability of the organization. Technology has enabled vide better quality and quantity of information to the
the collection and use of customer information without organization, than those with a lower motivation.

Customer’s Technology-Readiness: Parasuraman readiness index would be more willing to provide
and Colby (2001, p. 27) define technology-readiness as information to an organization.
“people’s propensity to use, embrace and employ new
technologies for accomplishing goals in home life and at P4b: Customer with high scores on the technology-readi-
work.” On one hand it may be argued that people who are ness index would be less willing to provide infor-
more technology-ready would be better able to interact mation to an organization.
with the technologies that organizations use to collect
customer information. However a counter argument may P5: The customer’s motivation for a relationship mod-
be that highly technology-ready customers would be more erates the relationship between the customers’ tech-
aware of how their information can be surreptitiously nology-readiness and the customer’s willingness to
collected and used, and so would be less willing to provide provide information.
information to organizations. Further we believe that the
impact of the customer’s technology-readiness on the P6: The customer’s technology-readiness moderates
customers’ willingness to provide information occurs the relationship between customers’ privacy con-
only in a relational context. So despite the customers’ cerns and the customers’ trust in an organization.
technology-readiness, they may not be willing to provide
information to an organization unless they trust the orga- The answer to this privacy question is not simple and
nization and have a relationship with that organization. opens up new avenues for research geared towards resolv-
Additionally we argue that the technology-readiness of ing the tug-of-war between organizations and their cus-
customers also impacts the relationship between privacy tomers over customer information, and thus making mu-
concerns and trust. For different levels of technology- tually beneficial customer relationships more fact than
readiness, the relationship between privacy concerns and fiction. By protecting customer privacy, the contribution
trust may be accentuated or attenuated, again depending made to society would be tremendous and the onus of
on the industry context i.e., the organizations’ need-to- protecting the rights of customers may shift somewhat
know. from policymakers to organizations, which then would
become the enforcers of privacy protection rather than
P4a: Customers with high scores on the technology- perpetuators of what has become a societal dilemma.
References available upon request.
Mona Srivastava
Texas A&M University
College Station, TX 77843
Phone: 979.845.4525
FAX: 979.862.2811
E-Mail: msrivastava@cgsb.tamu.edu

SHIFTS IN WORKPLACE ETHICS: OPPORTUNITIES FOR CONFLICT?
Paul L. Sauer, Canisius College, Buffalo
Paul Chao, Eastern Michigan University, Ann Arbor
SUMMARY For each workplace behavior we use two scales, one

to measure a self-report of strength of personal belief as to
The increased availability and use or abuse of office how wrong that behavior is and a second to measure self-
technology in the workplace is the concern of recent ethics report of perceived commonality of that behavior in the
research (Oz 2001; Stone and Henry 2003). Victor and respondent’s workplace. All items use a 7-point bi-polar
Cullen (1987, 1988) develop a classification of nine types scale with end-points anchored by strongly agree and
of ethical climate that vary along two dimensions: types of strongly disagree. This is equivalent to the wording of
criteria (or ethical criteria) and level of analysis (or locus items and scales used to assess ethical beliefs and confirm
of analysis). Peterson (2002) shows that giving gifts and dimensions of climate (Cullen and Victor 1987; Peterson
favors such as described in items that are contained in our 2002). We operationalize perceived incongruence by
gifts and entertainment section, is related to both the subtracting each respondent’s perception of ethical cli-
personal morality and corporate rules dimensions of the mate rating score from that respondent’s ethical belief
Victor and Cullen (1988) theoretical framework. Peterson rating score for each behavior. A score of zero would
(2002) also provides evidence that calling in sick and indicate a consonant state. The greater the positive differ-
lying such as described in items that are in our truth and ence, the greater is the potential for compelled incongru-
lies section, are related to Victor and Cullen’s (1988) ence. The greater the negative difference, the greater is the
ethical climate dimensions of personal morality and cor- potential for restraint incongruence.
porate rules. The purpose of our research to examine the
extent of temporal shifts in personal beliefs, in percep- The overall sample consisted of 14.8 percent top-
tions of the ethical climate in the workplace, and in the level managers or directors, 31.7 percent middle level
potential for conflict caused by changes in incongruence managers, and 35.5 percent lower level staff such as
between self-reported beliefs and climate perceptions technical employees or assistant managers. Regarding
office technology, independent sample t-tests reveal two
Building on the concept of value congruence be- significant shifts in the direction of weaker ethical beliefs
tween and individual and the office environment (Liedtka for Midwest respondents and three significant shifts occur
1989), an approach is developed to assess the shifts in the in the direction of stronger ethical beliefs for Northeast
congruence between individuals’ beliefs and their work- respondents. Midwest respondents indicate that use of
place climate. Subjective perception is the determinant of company e-mail for personal reasons is significantly more
whether conflict exists (Moser 1988) and depends on the common while Northeast respondents perceive use of an
ethical congruence between the individual and his or her office computer for Internet shopping to be more common
workplace. Restraint incongruence occurs when the indi- in post-2001. For Northeast respondents perceived com-
vidual does not perceive a behavior as wrong, but per- pelled incongruence increased for use of company e-mail,
ceives that such behavior is not common in the workplace. use of the office computer for Internet shopping
Compelled incongruence occurs when an individual per-
ceives a behavior as wrong, but the perceives the same Midwest respondents exhibited a shift toward weak-
behavior as common in the workplace. er beliefs regarding gifts and entertainment. For gifts and
entertainment, Midwest respondents perceive receiving a
Because there is evidence of regional geographic $50 gift from the boss as being significantly more com-
differences in individuals’ consumption habits, lifestyles mon while Northeast respondents’ perception of receiv-
and values (Kahle 1986; Mittal, Kamakura, and Govind ing $200 football tickets from a supplier is significantly
2004), we administer the survey to working professionals less common after 2001. Perceived restraint incongru-
enrolled in part-time management programs operated by ence increased for a $25 gift certificate from a supplier.
a private university in the northeastern region of the For Midwest respondents an increase in perceived re-
United States and a public university in the Midwest. The straint incongruence for a $75 raffle prize at a suppliers’
time frames for administration include pre-2001 and post- conference occurred. There were no shifts in ethical
2001 survey administration. We chose these time frames beliefs about truth and lies. No significant shifts in climate
to span the time before and after ethical scandals revealed occurred with respect to the two items measuring truth and
in late 2000 and in 2001. We did pre-2001 sampling in the lies. Overall only four (4) of the 108 tests for pre-2001 and
first half of 2000 and post-2001 in spring and fall of 2003. post-2001 differences for the gifts and entertainment and

truth and lies items were significant which is less than uals simply do not associate the personal behaviors used
chance. in this study with the higher corporate level of illegal and
unethical behavior. Another problem is that the pre-2001
Results with respect to shifts in beliefs indicate that and post-2001 data came from two different sets of
there is some change with respect to office technology, but respondents, thus there is no way to know if individuals
essentially none with respect to gifts and entertainment or changed, only if the aggregate sample changed. This also
truth and lies. In spite of the fact that no significant shifts may explain that not only were there a large number of
occur with respect to truth and lies behaviors, these forms non-significant differences, but also that there were dif-
of ethical abuse appear to provide the most fertile ground ferences that went in the opposite direction of what was
for perceived incongruence and conflict. Personal beliefs hypothesized. Improvements would include using match-
are very strong that these two behaviors are wrong, yet the ing samples over time and confining the study to employ-
perception is that they are generally common in the ees in one or two companies. References available upon
workplace, especially after 2001. One reason for the lack request.
of more significance effects may be the fact that individ-

Paul L. Sauer
Canisius College
2001 Main St.
Buffalo, NY 14208–1098
Phone: 716.888.2631
FAX: 716.888.3215
E-Mail: sauer@canisius.edu

INTERACTION BETWEEN THE SALESPERSON AND CUSTOMER:
A FRAMEWORK FOR IMPROVING THE SALES OUTCOME
Elizabeth Hemphill, University of South Australia, Australia
Chris Dubelaar, Monash University, Australia
Steven Goodman, University of South Australia, Australia
Gus Geursen, University of South Australia, Australia
ABSTRACT “sales” on behalf of the principal? Industry view has long

held the importance of rapport building to lead to sales.
This paper examines the role of disclosure on agency This paper examines this notion from an academic sense.
establishment. Structural Equation Modeling reveals the The objective of this paper is to explain what a salesperson
need for salespeople to reach a level of “rapport” for a should do to maximize their chances of establishing
phase transition prior to any eventual sale. Reaching this agency agreements by assembling and empirically testing
level is a key driver of successful outcomes, rather than a model of agent-principal relationship establishment.
the negotiating skills.
LITERATURE AND BACKGROUND
INTRODUCTION
The roots of agency relationships come from legal
Marketing centres on the process of exchange be- literature that adopts a perspective of enforceability of all
tween buyers and sellers (Kotler, Adam, Brown, and terms, covenants, and conditions of the agency agree-
Armstrong 2001). As sellers enter into relationships with ment. In legal literature (Reuschlein and Gregory 1990)
suppliers they act as agents in pursuit of consumers/ and contract law (e.g., U.S.A. District Court Case: Paul T.
buyers to the extent that “most of the world’s work is done Freund Corp. v. Commonwealth Packaging Co.) ele-
by agents” (Reuschlein and Gregory 1990). To date ments of agency include: consent, fiduciary agreement,
marketing literature has adopted a focus on relationship absence of gain or risk to the agent, and control by the
maintenance (Weitz and Jap 1995; Dahlstrom and Ingram principal (Reuschlein and Gregory 1990). Legally this
2003), essentially skirting around formation of this rela- fiduciary consensual relationship between agent and prin-
tionship (Dahlstrom and Ingram 2003) so that how sales cipal exists when “one person manifests an intention that
relationships are developed is not well covered, even in another shall act in his behalf and the other person con-
personal sales (Weitz and Jap 1995). This paper examines sents to represent him” (Reuschlein and Gregory 1990).
what it is about a specific seller that maximizes their Marketing literature has studied the application of agency
chance of establishing the right to pursue buyers in a way theory and the role of information in relationship forma-
overlooked until now in the literature. Specifically, the tion (Dahlstrom and Ingram 2003); the purchaser’s per-
approach taken in this paper is to understand the role of spective of influences that determine a decision to commit
rapport building in the establishment of agency relation- (Bagozzi and Dholakia 1999; Bagozzi 2000); relationship
ships. maintenance (Sitkin and Roth 1993; Sheth and Parvatuyar
2000; Singh 2000); agency relationship definitions and
When a seller (agent) is engaged by a supplier (as outcomes (Bergen, Dutta, and Walker 1992); drivers of
principal) to undertake some action (such as pursuing agent behavior (Richins, Black, and Sirmons 1987; Moore,
buyers) on behalf of that principal, an agency relationship Smolen, and Conway 1992; Marsh and Zumpano 1998);
is established. Thus, formation of the agency relationship and significant behavioral predispositions of a salesper-
is essential prior to any longer term association or market- son that influence sales performance (Weitz 1981).
ing exchange. In forming the agency exchange, an agent
“sells” to a buyer (principal) their offer to serve in a Relationship marketing literature adopts a focus on
master-servant relationship governed at least by the “rules the relationship rather than the transaction resulting from
of agency” (established consensually between agent and relationship formation (Sheth and Parvatuyar 2000), in
principal) and both agency and partnership law (Reus- which a process model clearly differentiates formation of
chlein and Gregory 1990). On the surface, establishing the the relationship from other aspects of relationship main-
agency agreement therefore looks like a normal sales tenance. However, even this literature fails to adequately
transaction. As a purpose driven process (Sheth and define the process of formation, slipping quickly into an
Parvatuyar 2000), the question remains: What can a sales emphasis demonstrated by Levitt (1983) in which the real
person really do to increase their chance of forming an value of a relationship between customer and seller occurs
agency relationship so they can subsequently pursue after a sale. Such a relationship is dependent on establish-

ing trust and cooperation (Morgan and Hunt 1994) and social self-disclosure in sales exchanges between life
avoiding cognitive dissonance (Festinger 1957) in order insurance sales agents and potential customers/investors
for it to continue. If we accept Sheth and Parvatuyar’s (Jacobs, Hyman, and McQuitty 2000). A seller’s commit-
(2000) observation that forming a relationship is more ment to a sale therefore depends on information disclo-
important than customer acquisition, agency formation sure between the principal and seller (as agent) on both
emerges as a critical part yet to be examined empirically transactional and personal levels (Jacobs et al. 2000).
for marketing. This paper is therefore an essential first
step in understanding how a supplier of agent services is Negotiation: Weitz and Jap (1995) further find that
selected from a pool of potential agents. This paper firstly the sales presentation comes before, and leads to, negoti-
adopts the agent’s perspective. Due to imperfections in ation. A clear link from salesperson’s behavior to a
self-assessment noted by authors such as (Jaramillo, Car- principal’s decision to commit exists in personal sales
rillat, and Locander 2003), we secondly offer results from literature based on appraisal of interactions with the sales
a pilot study of principals as a basis for future research to person (Lee and Dubinsky 2003), despite the specific
further compare the agent’s perspective with that of the decision influencing criteria of the principal being un-
principal. available to the salesperson as a matter of competition
(Spekman 1988). Lee and Dubinsky (2003) establish that
Theoretical Model courteousness, expertise, and friendliness are assessed to
determine satisfaction levels which then influence a pur-
Based on a number of propositions framed by Weitz chase decision. Throughout this process, known as “prim-
and Jap (1995) and Jacobs et al. (2000) we examine how ing” (Whittler 1994), purchase decisions can be influ-
the salesperson influences formation of the agency rela- enced by at least the salesman depending on their capacity
tionship. Weitz and Jap (1995) argue that relationships are to establish a sense of trust (Morgan and Hunt 1994), by
built around trust, communication and negotiation in what achieving a level of acceptance by the principal. This
is described by other authors as the first stage of relation- research paper tests the conceptual model below (Fig-
ship building, known also as pre-sale distinct from sale ure 1) using adaptations of propositions offered by Weitz
(Sheth and Parvatuyar 2000, p. 332). and Jap (1995) and Jacobs et al. (2000).
Communication: Models of customer-salesperson METHOD

interaction traditionally include an element of informa-
tion exchange. For example: Information exchange drives This study empirically tests propositions detailed in
a negotiation process (Olekalns 2002), the final stage of Table 1 above that have only previously been tested in
which is a level of “give and take” that result in a mutually other environments by Weitz and Jap (1995) and Jacobs
agreeable outcome; And a noted distinction between the et al. (2000). These are based firmly on avoiding cognitive
contributions of exchange specific self-disclosure and dissonance (Festinger 1957; Spekman 1988; Hunt 1991)
FIGURE 1
Conceptual Model
Social
Disclosure
P2, P5
People
Skills Negotiation Relationship
P1, P3 P6 Establishment
Exchange
Disclosure
P4
Empirical examination of the propositions presented in Table 1 tests this graphical model.

TABLE 1
Propositions Tested In This Research
Propos-
Research Propositions Instrument Item Construct tions
Interest is communicated and partner The agent came across as People P1

worthiness assessed (Weitz and Jap 1995) being competent Skills
Norms are used to provide a context for Information the agent acquired Social P2
messages (Weitz and Jap 1995) included which other agents Disclosure
were being considered
Communication increases over time spent The agent got to know the People P3
on the relationship (Weitz and Jap 1995) principal Skill
Salespersons must be skilled active listeners The agent got to know the People
that have the ability to communicate (Weitz principal Skill
and Jap 1995)
Exchange specific self-disclosure influences The agent got the principal to Exchange P4
the outcome of a sales attempt (Jacobs et al. share good things about the Disclosure
2000) property
Social self-disclosure influences the The agent got the principal to Social P5
outcome of a sales attempt(Jacobs et al. share why they were selling and Disclosure
2000) if they had already purchased
Negotiation of terms leads to a sale (Weitz The amount of changes to the Negotiation P6
and Jap 1995) agent’s proposed service
offering (sales commission, list
and reserve price, advertise-
ments)
and aim to determine which agent characteristics influ- cipal relationship is established (a sale of agent services
ence whether an agent-principal relationship was formed occurs). If unsuccessful, no legally binding agreement is
between agent and principal. established between that agent and that principal (no sale
occurs). Whether an agency agreement is established with
In this paper real estate agents, or realtors, are used as another agent, or does nothing at all is not measured.
sellers to examine agent-principal relationship establish-
ment because they engage in commonly occurring agency Construct Operationalisation
experiences with property sales (and purchases) that re-
quire establishment of an agent-principal relationship for Based on Churchill’s (1979) paradigm, discussions
technical or economic reasons. They operate in agent- provided a breadth and depth of qualitative data required
principal relationships in which people generally have for empirical instrument development to test Figure 1.
multi-agent experiences as they sell houses and property
at different points in their lives. In general, people who For this research we define negotiation as: the bar-
have engaged in such agent-principal relationships as gaining, trading-off, “haggling,” that results in change to
either agent and/or as principal (property vendors that the terms of the agent-principal agreement. This excludes
wish to sell their properties with the assistance of an agent) the design, plan, and assembling of the agent proposal,
readily share information about their experience. The anything prior to the completion of initial principal eval-
dependent variable for this research is a relationship uation of the agent proposal, anything prior to both parties
establishment attempt. This can be either successful or establishing their goals, measurement of the principal’s
unsuccessful. If successful, a legally binding agent-prin- previous experiences, and anything to do directly with

potential buyers. We did this because the literature clearly ing in total 274 usable sale attempts. Each response
indicates this as an important stage of negotiation (Ole- included information on two sale attempts – one success-
kalns 2002). We measure negotiation from the perspec- ful and one unsuccessful.
tive of its contribution to agent-principal relationship
establishment by examining the amount of change to a Differences between the two groups were examined
proposed agency agreement. closely to ensure that data were not distorted by self-
validity bias. Based on McQuitty (2003), sufficient level
Information disclosure underpins negotiation (Ole- of power was achieved to enable the model to verified
kalns 2002). We define this as the acquisition of exchange using SEM. Missing answers were imputed using expec-
information (about the property) and social information tation maximisation (Figueredo, McKnight, McKnight,
(about the property owner that is the potential purchaser and Sidani 2000; Flury and Zoppe 2000) to best preserve
of the agent services). Exchange disclosure is important in the integrity of this data set. The final response rate was
this sales process [Agent Focus Group]. “If a house has around eight percent. Although this could be considered
white ants we need to know it is fixed . . . and if buyers ask a major limitation to this research, there is only a six
we tell them to get an inspection done . . . we tread percent (.060412) chance that the sample does not repre-
carefully” [Agent W]. Due to legal requirements of full sent the population through calculation of the margin of
disclosure in the final sale to a property purchaser, we error (Kline 1998).
suggest that agents filter information by probing for
resolutions of potentially problematic information but Verification of sample representativeness involved
otherwise do not probe for information, thus the amount examining respondents’ personal previous month’s medi-
and nature of exchange information is important. an residential property price sold and comparing this to
published property prices for their city of operations.
Social information is important in the process of From this, two different types of agent sales performance
establishing an agent-principal relationship. Real estate estimations were evident. Firstly, a large number of agents
agents continually ask questions like “What were you estimated their sales performance close to published pric-
thinking of? Is that all right with you? Do you agree with es. These agents, on average, underestimated their sales
that?” [Agent Z] and “We try to find out who else the performance compared to published sales figures by only
vendor is considering and shy they are selling” [Agent approximately 12 percent. This is a relatively small dis-
W]. Agents in fact selectively appropriate social informa- crepancy, suggesting that respondents were representa-
tion in relationship establishment to simplify agent-prin- tive of their locations. In these cities, property prices were
cipal relationship maintenance and increase their chance relatively stable at the time of this research.
of being accepted by the principal. Our definition ex-
cludes how the agent uses the information, how the agent Secondly, a number of agents underestimated their
acquires the information, whether the information was sales performance compared with published property
requested, offered directly, observed indirectly, or ac- prices by up to 50 percent. Notably these agents operated
quired by other means, and the quantity of information in cities that experienced large house price increases over
about the vendor. We did this because “In any negotiation, the past 18 months revealing a telescoping effect resulting
information is power” (Latz 2001). Measurement of in- from agents being asked to consider back over time whilst
formation disclosure is focussed on the opportunity for we take a single snapshot. This means that should this
selective acquisition of property and property seller spe- research be replicated at a later date, with more stable
cific information. As agents’ selectively appropriate in- property prices, it would be reasonable to expect fewer
formation that will improve their chances of selling their discrepancies between estimations and published prices.
services to the principal, measurement of information is Responding agents were found to be reasonably represen-
focussed on the selectivity of exchange specific informa- tative of the regions in which they operate.
tion and more personal social information.
Anderson and Gerbing’s (1988) approach to Struc-
Data Collection and Analysis Method tural Equation Modeling (SEM) was used for data analy-
sis to test the theoretical model (Figure 1) so that indirect
A self-completion mail questionnaire was developed effects of model constructs could be estimated and inter-
and sent to 1600 real estate sales agents (as sellers) that actions between constructs could be examined. Direct
had experienced both successful and unsuccessful at- influences of model constructs on the dependent variable
tempts to establish supply relationships with property were examined with Logistic Regression (McFadden
vendors. These were randomly selected from the Real 1976). Variance and correlations were examined to affirm
Estate Institute of Victoria’s member database but follow- measure reliability, validity, and purity based on Camp-
up data collection involved convenience sampling from bell and Fiske (1959), Churchill (1979), Cronbach (1951),
the original database. Respondents were screened to en- Finn and Kayande (1997), Menezes and Elbert (1979) and
sure they were representative of agents in general provid- Peter (1981) prior to structural model evaluation in order

to justify implementation of a model evaluation technique Examination of Bollen and Stine p-values reveals
(Bagozzi 1982; Bhargava, Dubelaar, and Ramaswami that similar results could be expected in 39.9 percent of
1994; Churchill 1979; Finn and Kayande 1997; Menezes cases (95.6% for sales compared to 16.3% for no-sales).
and Elbert 1979; Nunnally and Bernstein 1994; Peter Such poor model indicators for the unsuccessful relation-
1981). ship establishment attempt (no-sale model) suggest that it
is not just the interaction with the salesperson that ac-
RESULTS counts for no-sale being made. Table 3 shows differences
between a sale and no sale estimates. Incorporating the
Diagrammatic representation of the resulting struc- work of Pearl (2000) links tested with SEM are significant
tural equation model is shown in Figure 2 below with when Critical Ratio estimates are greater than 1.96 (Ar-
Table 2 detailing results of model statistics. buckle and Wothke 1999, p. 74); or Differences between
relationship establishment attempt outcomes for a single
Structural equation model indicators in the Table degree of freedom are significant (if Discrepancy esti-
show the structural model estimated in the Figure below mates are less than 3.84146 then estimates are considered
delivers a reasonable fit within the measurement model significantly different from each other at a significance
(with Fit indicators close to 1), confirming a significant level of 0.05).
difference between successful (a sale) and unsuccessful
(no sale) relationship attempt outcomes. The likelihood This Table shows that people skills are significantly
that test statistics are correct exists because there is nearly related to exchange (CR > 1.96) and social disclosure
an 80 percent chance that acceptable structural models (CR > 1.96) and that the importance of social disclosure
were rejected, meaning that if not rejected the statistics are differs significantly between successful and unsuccessful
more likely understated than overstated to a level of 0.8 relationship establishment attempts (Significant differ-
for the full data set (McQuitty 2003, p. 7). ence between estimate of 0.40 and -0.47). Logistic regres-
FIGURE 2
Structural Model Estimates
1.07 1.68 2.01 2.23

eq 14 a eq 14 d eq 11 e eq 11 k
1 1 1 1
q 14 a q 14 d q 11 e q 11 k
1.23 1.15 1.00 .73
eppl 1.03 1.00 e exch
1 1
People .68 Exchange
Skills Disclosure
.33 -.05
6.16
.63
eneg
esoc 1
1 Social .02
Negotiation
Disclosure
2.14 1.59 1.00 1.00 .90 1.03
q 11 d q 11 b q 11 a q 13 a q 13 b q 13 c
1 1 1 1 1 1
1.08 4.60 4.08 3.58 3.38
1.85
eq 11 d eq 13 a eq 13 b eq 13 c
eq 11 b eq 11 a

TABLE 2
Model Statistics
Boot Bollen
Incre- Strap &
RMSEA Normed Relative Mental Comparative Mean SE of Stine
Chi-Sq Df P max Fit Fit Fit Fit Disc MEAN P-value
M’ment 18.97 29 .922 .02 .95 .92 1.03 1.00 38.38 .30 .97
Model
Structural 40.26 31 .123 .06 .94 .92 .99 .99 38.53 .26 .40
Model
Sale 21.20 31 .906 .03 .94 .91 1.03 1.00 41.47 .32 .96
No sale 47.72 31 .028 .10 .87 .81 .95 .95 36.78 .26 .16
TABLE 3
Differences Between a Sale and No-Sale (SEM Results)
A Sale (Successful No Sale (Unsuccessful

Relationship Relationship Significantly
Establishment Attempt) Establishment Attempt) Different
Est. S.E. C.R. P Est. S.E. C.R. P
People Skills to
Exchange Disclosure 0.63 0.16 3.98 *** 0.68 0.25 2.73 0.01 No
People Skills to
Social Disclosure 0.32 0.10 3.19 0.00 0.38 0.15 2.53 0.01 No
Social Disclosure to
Negotiation 0.40 0.28 1.43 0.15 -0.47 0.36 -1.28 0.20 Yes
Exchange Disclosure to
Negotiation -0.19 0.25 -0.75 0.46 0.16 0.31 0.53 0.60 No
sion results in the table below show significant relation- On a theoretical level, negotiation should depend on
ships between modelled constructs and the outcome of an information disclosure for setting up terms of a business
agent-principal relationship establishment attempt when relationship. Personal relationships set up the clients’
Sig. < 0.05. satisfaction prior to any sharing of information that may
lead to commencement of negotiations, or eventuate in a
These results show that getting to know the customer sale. We propose therefore that this can be viewed as two
(sig. = .009 in Table 4) is the best way to (a) achieve social distinct phases of the sales process. The people skills
disclosure and (b) establish the relationship that deter- construct can be viewed as what is commonly discussed in
mines the likelihood that a marketing exchange will industry sales literature and training courses as “rapport
occur; Knowing the competition is important aspect for a building,” or “empathy selling” (Golis 1991). These re-
sale (sig. = .037 in Table 4); That negotiation and ex- sults give empirical support to the notion that the selling
change specific disclosure will not assist in establishing a process can be observed as a “phase transition” where
sales exchange without the support of the people skills people skills (the ability to build rapport) are most impor-
component. tant and come first. Without establishing rapport, and

completing the phase transition to commitment to pur- strates the need for research in which data are gathered
chase, a sale (and an agency agreement) is not likely to be from the client side of the agency exchange. Admission of
achieved. personal inadequacies such as this would be unlikely, as
individual agents seek to preserve their self-esteem. Con-
The notion that self-evaluation in the area of sales and sequently, steps were taken to overcome such a bias in this
agency relationships is unreliable is supported in our research by conducting an exploratory pilot study of 200
results on the basis that people skills are clearly attribut- principals.
able to success but not necessarily attributed to failure to
achieve a sale. Model differences suggest a distinct inabil- This Table shows that the most important “sale”
ity to acknowledge fault when the sale relationship is not influences mirror closely the aspects that significantly
established. Given the nature of sales and personal selling, differ between the model for a “sale” and “no sale.” That
this difference is to be intuitively expected and demon- is, personal relationships between agent and vendor are
TABLE 4
Logistic Regressions Results
Construct Measured
Instrument Item Wald Sig. by Instrument Item
Why the vendor was selling 0.046 0.830 Social Disclosure

Which other agents they were considering 4.351 0.037 Social Disclosure
If the vendors had already purchased 1.127 0.288 Social Disclosure
Good things about the property 0.568 0.451 Exchange Disclosure
All there was to know about the property 0.157 0.692 Exchange Disclosure
Change in sales commission 0.409 0.523 Negotiation
Change in property list and reserve price 0.005 0.942 Negotiation
Changes in the advertising campaign 0.088 0.766 Negotiation
Agent persuasively demonstrating competence 0.428 0.513 People Skills
Agent getting to know the vendor 6.901 0.009 People Skills
TABLE 5
Purchaser Pilot Study Results
Influence in Purchase commitment Count % Responses
Relationship(acquaintance) with agent 101 27.2

Reputation 74 19.9
Recommendation 55 14.8
Results 31 8.4
Market Share 26 7.0
Verbal/written submission 16 4.3
Corporate identity 16 4.3
Media presence 12 3.2
Auction/For Sale boards 12 3.2
Website 5 1.3
Other 23 6.2
Total responses 371 100

established as the agent gets to know the vendor and We argue that it is not just trust and satisfaction that lead
perceives a position of competitive advantage in the to a sale but the degree to which a salesperson can “Get to
agent’s ability to perform contracted services after the know” the customer, thus entering their “space.” Direct
purchase. These results further support our view that the interaction and information disclosure between agent and
personal relationship with the agent is most influential in purchaser results in a level of commitment to a purchase
moving the purchaser through the initial phase of their decision prior to the actual sale transaction. However,
purchase commitment and accept the salesperson and only if rapport is right then transition can be direct to
their competence to perform tasks of the agency agree- relationship establishment without incurring further risk
ment. to the outcome via negotiation.
DISCUSSION/IMPLICATIONS This paper strongly argues therefore that agency

establishment (as a normal sales exchange) is a multi-
This paper examines the point of a sales relationship phase process. In this process the first phase requires
establishment from the perspective of agent-principal establishment of trust between the seller and purchaser
relationship establishment. We do this because (i) an (Morgan and Hunt 1994) and establishment of a personal
agent must establish these relationships with a principal relationship. This relationship must in fact include a level
prior to any further money making transactions on behalf of satisfaction (Weitz 1981) and avoidance of cognitive
of that principal (ii) as part of a sales transaction marketing dissonance (Festinger 1957). Subsequent to moving
is dependent on establishment of these relationships and through this first phase other elements of buyer behavior
(iii) more and more, business structures can be seen as models may account for the eventual outcome of a sales
agency structures as often business is conducted through attempt but in the first instance of phase transition, estab-
networks of independent agents. Establishment of the lishing rapport with a potential customer is critical.
agency agreement is clearly similar to normal sales trans-
actions in which the agent’s role of establishing rapport This paper confirms that agency establishment is a
with the principal is critical as purchasers are led to a sales normal sales exchange on the basis that (i) legal require-
commitment by their personal rapport with a salesperson. ments prescribe steps prior to agent action on behalf on the
principal (ii) vendor (principal) selection of an agent is a
The key principle this research addresses is the way competitive process and (iii) the transaction is basically a
in which a salesperson (agent) contributes to a position of services marketing transaction in which personal skills of
competitive edge in the mind of the purchaser (principal) the agent are the driving force. This paper also establishes
in what is essentially a normal sales process (establish- that the process of negotiation is incorrectly positioned in
ment of the agent-principal relationship). Sales literature existing literature because it: (i) cannot occur until a
strongly suggests that a movement towards commitment personal relationship has been established; (ii) does not
to a sale is based on the customers’ perception of trust and necessarily lead to a final sales transaction; and (iii) does
satisfaction with the sales person (Morgan and Hunt 1994; not necessarily rely on information disclosure from the
Weitz and Jap 1995; Jacobs et al. 2000; Olekalns 2002). principal (vendor).
FIGURE 3
Revised Model
Social
Disclosure
P2, P5
People Relationship
Negotiation Establishment
Skills
P6
P1, P3
Exchange
Disclosure
P4

Significant implications for practitioners (general Gopinath, and Nyer 1999; Bagozzi 2000). Adding to
sales people, agents, and agencies) and theory exist from Bagozzi and Dholakia’s (1999) definition of the consum-
these findings because: er’s pursuit of specific goals, this paper contributes to
sales presentation literature by empirically demonstrating
“Selling is a process of influence . . . the ability to that, as the interface with the client, the sales presentation
present information, without being abrasive or apol- is the highest level predictor variable. People skills are the
ogetic, is a learnable and important skill . . . [involv- agent’s mechanism (as a salesperson) with which to enter
ing a] . . . willingness to be shaped, influenced, and a client’s space so these skills represent the highest order
impacted upon by the customer” (Shaw 1981). of the variables in relationship establishment.
The important practical implication from our re- Negotiation emerges as a demand side benefit (post
search is in the area of sales training known currently to be relationship establishment, post-“sale”) for pursuit of a
a process of (i) developing a taxonomy of behaviors (ii) final transaction and is not needed for relationship estab-
practicing skills and (iii) providing support and reinforce- lishment. Currently, agent-principal relationships (sales)
ment (Shaw 1981). This research takes the first step in begin in the case of real estate by presenting a proposed
defining a taxonomy of behavior that leads to sales rela- marketing campaign to a client in the context of selective
tionship establishment. successes and failures that is often tweaked and altered.
Our research demonstrates that negotiation is not even
By modeling and empirically testing behavior that desirable when tendering a relationship proposal to a
influences the likelihood that a sale is made, the outcome potential client. As a result agency owners should hire
of a sales attempt is found not to be determined, as agents that have excellent people skills and can get along
described by Bergen et al. (1992), by processes that follow with a potential client because of the importance of these
initial contact between an agent and a principal in which to a relationship establishment attempt outcome directly;
terms of an agency agreement are negotiated. It is instead have adequate processes in place so that agents are not put
determined at the point at which an initial context, or in the position of having to negotiate terms of an agent-
reference point, is established. This point is translated by principal relationship because this will not assist in rela-
the principal into their measure, or perception, of equity tionship establishment; and ensure adequate training pro-
and fairness that actually determines the outcome of the cesses are available for personal skill acquisition, prac-
agent-principal relationship establishment attempt rather tice, and support.
than any negotiation.
Agents need to be most concerned about gaining
For the literature, this paper contributes the finding acceptance by a client at the initial point of contact so that
that information disclosure is seen as an influence of the the client will disclose sufficient information to the agent
likelihood that a sale will result, rather than negotiation so that the agent can consolidate the initial relationship
although social information is seen as an influence on into a sales transaction. As a result, the client is less likely
negotiation. Quigg and Wisner’s (1998) work needs to be to experience cognitive dissonance that will result in an
reworked to reflect the contributions of Levitt’s (1986) unsuccessful relationship establishment attempt.
consumer reference point thereby acknowledging attitu-
dinal psychology; Braun’s (1999) work needs to be mod- CONCLUSION
ified to reflect the substantial benefit to agent-principal
relationship establishment from identification of Kahne- A fresh view of the selling process as one of phase
man, Knetsch, and Thaler’s (1986) framing effects (iden- transitions offers great potential. At a practitioner level,
tification of the competition); and Jacoby, Chestnut, and the notion of “rapport” reaching as a prerequisite for a
Fisher (1978) and Marsh and Zumpano’s (1998) work phase transition to negotiation and then onto a sale offers
needs to be modified because these authors incorrectly empirical evidence not provided in this way before whilst
detail the role of information in agent-principal relation- not being a startling new insight for the sales profession.
ship establishment. Thus this research provides new support for existing
training methods. Gathering further data from the client
Although we do not specifically focus on communi- side of the agency transaction and replication in other
cation, interpersonal communication between agent and industries offers much scope for developing the model
client (in generally getting along and getting to know each into accountable marketing knowledge. The scope also
other) is shown to be paramount to a sale. Bagozzi (2000) exists for work across cultures and geographic locations
clearly acknowledges the role of emotions in consumer to even further refine and confirm the model. Interesting
purchase decisions and the level of rationality behind insights may result from the application of the model in a
agents’ efforts to avert cognitive dissonance on the part of cross-cultural setting. Importantly, this research is held
the purchaser (Bagozzi and Dholakia 1999; Bagozzi, out for comment, and contribution to the divide between

academic knowledge and practitioner experience and use that can assist in the production of research outcomes that
in the area of personal selling. In the theme of this are of assistance to both industry and academia.
conference, it is presented as a piece of marketing research
REFERENCES “Starting to Solve the Method Puzzle in Salesperson

Self-Report Evaluations,” Journal of Personal Sell-
Anderson, J.C. and D.W. Gerbing (1988), “Structural ing & Sales Management, 23 (4), 369–78
Equation Modeling in Practice: A Review and Rec- Kahneman, D., J.L. Knetsch, and R. Thaler (1986), “Fair-
ommended Two-Step Approach,” Psychological ness as a Constraint on Profit Seeking: Entitlements
Bulletinm, 103 (3), 411–23. in the Market,” The American Economic Review, 76
Arbuckle, J.L. and W. Wothke (1999), AMOS 4.0 User’s (4), 728–41.
Guide. USA, Smallwaters Corporation. Kline, R.B. (1998), Principles and Practice of Structural
Bagozzi, R.P. and U. Dholakia (1999), “Goal Setting and Equation Modelling. New York: The Guildford Press.
Goal Striving in Consumer Behaviour,” Journal of Kotler, P., S. Adam, L. Brown, and G. Armstrong (2001),
Marketing, 63, 19–32. Principles of Marketing. New South Wales: Pearson
____________, M. Gopinath, and P.U. Nyer (1999), Education.
“The Role of Emotions in Marketing,” Academy of Latz, M.E. (2001), “Give and Take,” American Bar Asso-
Marketing Science, 27 (2), 184–206. ciation, 87, 66.
____________ (2000), “On the Concept of Intentional Lee, S. and A.J. Dubinsky (2003), “Influence of Salesper-
Social Action in Consumer Behaviour,” Journal of son Characteristics and Customer Emotion on Retail
Consumer Research, 27 (3), 388–96. Dyadic Relationships,” The International Review of
Bergen, M., S. Dutta, and O.J. Walker, Jr. (1992), “Agen- Retail, Distribution, and Consumer Research, 13 (1),
cy Relationships in Marketing: A Review of the 21–36.
Implications and Applications of Agency Related Levitt, T. (1986), The Marketing Imagination. New York:
Theories,” Journal of Marketing, 56 (3), 1. The Free Press.
Braun, K.A. (1999), “Post experience Advertising Effects Marsh, G.A. and L.V. Zumpano (1998), “Agency Theory
on Consumer Memory,” Journal of Consumer Re- and the Changing Role of the Real Estate Broker,”
search, Gainesville, 25 (4), 319– 34. Journal of Real Estate Research, 13 (2), 151–64.
Churchill, G.A., Jr. (1979), “A Paradigm for Developing McFadden, D. (1976), “A Comment on Discriminant
Better Measures of Marketing Constructs,” Journal Analysis ‘Versus’ Logit Analysis,” Annals of Eco-
of Marketing Research, 60–73. nomic and Social Measurement, 5 (4), 511–23.
Dahlstrom, R. and R. Ingram (2003), “Social Networks McQuitty, S. (2003), “Statistical Power and Structural
and the Adverse Selection Problem in Agency Rela- Equation Models in Business Research,” Journal of
tionships,” Journal of Business Research, 56, 767– Business Research, 5641 (In Press).
75. Moore, G.S., G.E. Smolen, and L.V. Conway (1992),
Festinger, L. (1957), A Theory of Cognitive Dissonance. “The Effects of an Informational Disclosure Form on
Row, Peterson. the Real Estate Agency Representational Model,”
Figueredo, A.J., P.E. McKnight, K.M. McKnight, and S. Journal of Real Estate Research, 7 (2), 217–26.
Sidani (2000), “Multivariate Modeling of Missing Morgan, R.M. and S.D. Hunt (1994), “The Commitment-
Data Within and Across Assessment Waves,” Addic- Trust Theory of Relationship Marketing,” Journal of
tion, 95 (11 supp 3), 361–81. Marketing, 58 (3), 20–45.
Flury, B. and A. Zoppe (2000), “Exercises in EM,” Olekalns, M. (2002), “Negotiation as Social Interaction,”
American Statistician, 54 (3), 207, 3. Australian Journal of Management, 27 (Special Is-
Golis, C. (1991), Empathy Selling. Port Melbourne, Vic- sue 2002).
toria, Australia: Lothian Publishing. Pearl, J. (2000), Causality. United Kingdom: Cambridge
Jacobs, R.S., M.R. Hyman, and S. McQuitty (2000), University Press.
“Exchange-Specific Self-Disclosure, Social Self- Quigg, B. and B. Wisner (1998), Selling the Right Way.
Disclosure, and Personal Selling,” Journal of Mar- New Jersey: Prentice Hall.
keting Theory and Practice, (Winter). Reuschlein, H.G. and W.A. Gregory (1990), The Law of
Jacoby, J., R.W. Chestnut, and W.A. Fisher (1978), “A Agency and Partnership, 2nd ed. St Paul, MN: West
Behavioral Process Approach to Information Acqui- Publishing Co.
sition in Nondurable Purchasing,” Journal of Mar- Richins, M.C., W.C. Black, and C.F. Sirmons (1987),
keting Research, 15 (November), 532–44. “Strategic Orientation and Marketing Strategy: An
Jaramillo, F., F.A. Carrillat, and W.B. Locander (2003), Analysis of Residential Real Estate Brokerage Firm,”

Journal of Real Estate Research. Spekman, R.E. (1988), “Strategic Supplier Selection:
Shaw, M. (1981), “Sales Training in Transition,” Train- Understanding Long-Term Buyer Relationships,”
ing and Development Journal, (February), 74–83. Business Horizons, (July/August), 75–81.
Sheth, J.N. and A. Parvatuyar (2000), Handbook of Rela- Weitz, B.A. (1981), “Effectiveness in Sales Interactions:
tionship Marketing. California: Sage Publications. A Contingency Framework,” Journal of Marketing,
Singh, J. (2000), “Agency and Trust Mechanisms in 45 (Winter), 85–103.
Consumer Satisfaction and Loyalty Judgements,” ____________ and S.D. Jap (1995), “Relationship Mar-
Journal of Academy of Marketing Science, 28 (1), keting and Distribution Channels,” Journal of Acad-
150–67. emy of Marketing Science, 23 (4), 305–20.
Sitkin, S. and N. Roth (1993), “Examining the Limited Whittler, T.E. (1994), “Eliciting Consumer Choice Heu-
Effectiveness of Legalistic ‘Remedies’ for Trust/ ristics: Sales Representatives’ Persuasion Strategies,”
Distrust,” Organization Science, 4 (August), 367– Journal of Personal Selling & Sales Management,
92. XIV (4), 41–53.

Liz Hemphill
School of Marketing
University of South Australia
27 North Terrace
Adelaide, South Australia 5000
Phone: +61.8.8302.0623
FAX: +61.4.1788.2176
E-Mail: ehemphill@unisa.edu.au

COUPONS: THE INSIDE SCOOP
Somjit Barat, University of North Texas, Denton
SUMMARY The author also categorizes coupons into two major

classes: buy-one-get-one-free and half-off coupons, the
Coupons are distributed with the intent of promoting reason being, they influence the customer’s coupon-value
the product through an increase in sales or encouraging perception and savings-perception in different ways. The
new trials. Even though this widely popular medium of conceptual framework, therefore, may be laid out as
promotion has been extensively researched (Ailawadi follows: The perceived value of a coupon affects the
et al. 2001; Bawa et al. 1997; Garretson et al. 1999; direction and magnitude of income and substitution ef-
Heilman et al. 2002; Lichtenstein 1990; Mittal 1994; fects and are moderated by the type and face value of
Raghubir 1998; Reibstein 1982; Ward 1978; Taylor 2001; coupons, which, in turn, affects the price-perception of
Nevo 2002) in social science, this paper contributes to the the product. Based on this framework, eight hypotheses
body of knowledge by addressing at least one major are suggested out of which, five are strongly and two are
shortcoming, that coupons have some inherent disadvan- partially supported, while one is not supported.
tages. These latent weaknesses might eventually act as a
detriment to the “promoted” product. Such a possibility As far as implications are concerned, a good knowl-
arises from the disposition of customers towards coupons edge of how coupons might adversely affect sales pros-
in general and the promoted product in particular. pects is critical for managers, manufacturers, and retailers
alike. Even within a single product category, one should
Data was collected from a sample consisting of both note what kind of coupon is more appropriate for what
students and non-students at a large university in the kind of products. It also helps to address what face-value
southwestern part of United States. While students are a to put on a coupon so that it encourages the customer to
major segment of coupon-users, the analysis was also redeem the coupon as well as increases the bottom line of
restricted to convenience products because this category the coupon issuer. Results from this study can help the
is one of those where coupons are redeemed the most. The customer use coupons in a more efficient way. Moreover,
author attempts to buttress his point by investigating how given that the sample consists of both students and non-
customers evaluate a coupon when they see one. For students, the results can be generalized with reasonable
example, results showed that such an evaluation is based level of comfort.
on how strongly customers feel towards a coupon per se
as well as on how they view the savings that results from On the other hand, future studies can investigate how
redeeming a coupon. Similarly, on encountering a cou- customers actually utilize the “extra savings” they enjoy
pon, customers feel suspicious as well as confused about by redeeming coupons i.e., whether they just hold on to
the real price of the product. Finally, seeing a coupon also their savings, buy more of the same or of some other
influences the perception of the promoted product by the products. Information on redeeming specific type of cou-
customer. This perception can be operationalized through pons (e.g., half-off vs. buy-one-get-one-free) can be used
the income effect and substitution effect, both of which to target specific customers, thereby making the whole
are concepts well grounded in economic theory. process more efficient.

Somjit Barat
P.O. Box 311396
Denton, TX 76203–1396
Phone: 940.565.3121
FAX: 940.565.3837
E-Mail: BaratS@unt.edu

WHAT NEXT? EXPLAINING REPURCHASE DECISIONS AFTER
JOINING A LOYALTY PROGRAM
Shirley Y. Cheng, The Chinese University of Hong Kong, Hong Kong
Jessica Y. Kwong, The Chinese University of Hong Kong, Hong Kong
SUMMARY Thus, we propose that compared to someone who

does not join the LP, an LP participant is expected to
In the past few years, loyalty program (LP) started evaluate a repurchase decision with the LP-offering com-
drawing attention from researchers. However, extant lit- pany more favorably due to the prospect of reward. On the
erature focused mainly on consumer behavior before other hand, an LP participant may evaluate a similar
joining an LP (e.g., joining decisions and program evalu- transaction with a non-LP-offering company less favor-
ations); how consumers perceive and response to LPs ably, because the loss in prospect of reward would result
during the program is left largely unexplored. This paper in a deduction of acquisition utility. Due to risk aversion,
draws on mental accounting theory to explain how LP we predict that losing prospect of reward has greater
joining affects a consumer purchase decisions after join- impact than gaining the same prospect of reward on
ing an LP. This focus on purchase decisions subsequent to transaction.
program joining is crucial because getting the reward
requires a consumer to make a series of purchase with the In view of the effects of prospect of reward on
LP-offering company. This extended effort-investing pro- acquisition utility, this paper proposes that evaluation of
cess (Kivetz and Simonson 2003) implies that consumers acquisition utility can be influenced by an LP structural
can decide at any time whether to keep pursuing or not, features, which either: (1) make the value of the prospect
and the effectiveness an LP depends largely on whether it definite; or (2) segregate the prospect of reward from
can influence participants subsequent purchase decisions other components (e.g., price and value of products) of the
in favor of the LP-offering company. transactions.
How Does an LP Work? Making the Prospect of Reward Definite
According to Thaler (1985) mental accounting theo- To many participants of LPs, the program rewards are
ry, consumers evaluate the acquisition utility of transac- uncertain and delayed because they are contingent to a
tions, combining what is obtained relative to its price and series of subsequent purchases. In response to uncertainty
valuating the utility, in accordance to Kahneman and and delay outcomes, people discount their values and
Tversky (1979) prospect theory. Prospect theory holds impacts on present decisions (Frederick, Loewenstein,
that utilities are evaluated as gains and losses relative to and O’Donoghue 2003; Rachlin, Raineri, and Cross 1991).
some reference point, and while both gain and loss func- One could thus predict that the effect of reward prospect
tions are concave (i.e., displaying diminishing sensitivi- on acquisition utility can be very limited especially at the
ty), loss function is steeper (i.e., loss aversion). Based on early stage of program completion. A transparent medium
these features of value function, Thaler (1985) formulated system may counter the effects by making the prospect of
principles of hedonic framing, which specifies the way of reward definite at the time of transactions. We define the
evaluating joint outcomes to maximize resulting utility. transparency of a medium system as the extent to which
the monetary value of each unit of medium is made
When a consumer joins an LP, he/she has an expec- explicit to the consumers. A transparent medium system
tation of a future reward. Subsequently, every time the will make the reward prospect more definite in values,
consumer makes a purchase after he/she joined the LP, he/ thus enhance its effect on evaluation of acquisition utility.
she is approaching the program reward. We termed this
partial realization of reward rospect of reward Although Segregating the Prospect of Reward
the program reward is not obtained in a single transaction,
the prospect of the reward does carry utility that, as a part Because the gain function is concave, segregating the
of what is obtained in the transaction, represents a gain in prospect of reward from other components in utility
acquisition utility in every transaction with the LP-offer- evaluation enhances resulted acquisition utility. Offering
ing company, but a loss in every transaction with a non- incommensurate reward will prevent consumers from
LP-offering company. integrating the reward prospect with product value and

price when evaluating the acquisition utility. Incommen- Another way to segregate the reward prospect is to offer
surate rewards, in the context of LPs, are defined as multiple rewards. Suppose an LP offers three different
rewards that are of different units of measurement from gifts as reward, the prospect of each gift will be segregat-
that of the effort. Incommensurate reward relative to ed. The separate prospects of the three gifts will have
commensurate reward is expected to have higher value, greater value than the prospect of a single reward with
and thus, more positive impact on acquisition utility. same value.

Shirley Y. Cheng
Shatin, Hong Kong
Phone: 852.2609.7808
FAX: 852.2603.5473
E-Mail: shirleycheng@cuhk.edu.hk

NOT ALL DEALS ARE CREATED EQUAL: TWO DIFFERENT ROLES OF
SALES PROMOTION
Dongwoo Shin, Texas A&M University, College Station
James H. Leigh, Texas A&M University, College Station
SUMMARY Disturbance refers to a process whereby a sales

promotion influences a consumer’s buying behavior by
Shopping behavior is goal-directed and often planned. breaking his/her current implementation intentions. When
Shoppers seek desired outcomes (e.g., obtaining a desired a consumer encounters a promotion stimulus, it often
product) or positive experience (e.g., having a relaxing breaks his/her original implementation intentions and
vacation experience). A consumer often sacrifices other induces him/her to switch to purchase the promoted
alternatives to achieve his/her chosen goal and imple- product since it provides an alternative with additional
ments various self-regulatory strategies to prevent him/ benefits. In this disturbance process, the values (i.e.,
her from giving in to the alternatives. Many sales-promo- motivations) from the sales promotions should provide
tion strategies are designed to affect such self-regulatory sufficient force to break the current action plan. When the
strategies by providing economic and psychological in- incentive is sufficient, the shopper halts the execution of
centives. The action-oriented framework proposed in this his/her current implementation intention (i.e., it breaks the
paper emphasizes the influence of sales promotions in a cognitive link between the environmental cues and the
consumer’s goal-striving process and regulatory func- desired behavior), and activates a disturbance feedback
tions of implementation intention. The main premise of loop to reevaluate the current and new options.
the suggested framework is that sales promotion influenc-
es not only the process of formulating goal intentions but We introduce Kuhl’s (1994) action orientation ten-
also the process of creating and executing implementation dency as a moderating variable that influences the forma-
intentions (Gollwitzer 1999), the building blocks of an tion and disturbance processes. An action-oriented person
action plan, which play the key-mediating variable be- can maintain and execute goal intentions more efficiently
tween goal intention and buying behavior. The model than a state-oriented person because he/she focuses more
hypothesizes that a sales promotion stimulus influences a on a realistic action plan. In contrast, a state-oriented
consumer’s goal striving process via two routes: forma- person focuses more on the evaluation of the present state,
tion and disturbance processes. a past state, or a future state and lacks the flexibility of his/
her action plan. Since most sales-promotion stimuli con-
Formation refers to a process which a sales promo- tain the information required to create an implementation
tion motivates a consumer to form a new set of implemen- intention (when, where, and how), an action-oriented
tation intentions. The process of formation starts with consumer can quickly establish, maintain, and execute an
creating a purchase intention. When a consumer encoun- effective and well-developed shopping plan. However,
ters a sales promotion stimulus before developing a pur- since a state-oriented consumer tends to overanalyze
chase intention, he/she includes the additional benefits information about alternative goals (end states) even after
(i.e., informational, economic, and affective, Raghubir he/she created a goal intention and implementation inten-
et al. 2004) from the promotion to their decision process. tions, he/she is more easily distracted from the current
After considering the alternatives, the consumer makes up action plan. Therefore, a sales promotion strategy focused
his/her mind and creates an intention to purchase a certain on the formation process will influence action-oriented
product. The purchase intention summarizes the strength people more since they utilize the provided information
of motivation that provides the consumer energy to devel- and motivation better for creating implementation inat-
op, pursue, and realize his/her action plan. After the tentions. On the other hand, since state-oriented people
consumer created a purchase intention, he/she seeks in- are easily distracted from the current goal pursuit process,
formation concerning when, where, and how to formulate a sales promotion strategy focused on the disturbance
implementation intentions. The information from sales process will influence them more fully.
promotion provides not only value-related but also plan-
related content (i.e., the information about the store loca- It is also hypothesized that a state-oriented consumer
tion, the period of the promotion, etc.). Such plan-related will have a higher correspondence between his/her past
information is connected with appropriate purchase be- behavior with the current behavior since a state oriented
haviors, creating new implementation intentions leading person tends to delegate their regulatory functions to the
to a higher chance of realizing the purchase behavior. situational cues when they perform repeated behaviors in

a stable environment. A stable environment generates current behavior when performing a habitual behavior.
little variation in the past, present, and the future. When Because their main concern is in developing efficient and
state-oriented people experience such stability over a effective action plans, they will generate more contingen-
period time, they would have less motivation to evaluate cy plans in a stable environment, compared to state-
alternatives possible states. Consequently, they develop a oriented consumers. Instead of spending their cognitive
strong cognitive schema that connects the stable environ- resources evaluating alternative end states, action-orient-
mental cues with the routine behavior and will create ed people will spend their cognitive resources developing
strong correlations between past and future behavior as more flexible action plans because their main concern is
long as such routines can provide means to achieve their in developing efficient and effective action plans. There-
desired goals. On the other hand, an action-oriented fore, when thy encounter the right opportunities, they will
consumer, even though she/he usually has a higher corre- change their habitual behavior by activating another im-
spondence between intention and behavior, he/she will plementation intention.
have a lower correspondence between past behavior and
REFERENCES tion,” in Volition and Personality, Julius Kuhl and

Jurgen Beckmann, eds. Gottingen, Germany: Hogre-
Gollwitzer, Peter M. (1999), “Implementation Intentions: fer & Huber.
Strong Effects of Simple Plans,” American Psychol- Raghubir, Priya, J. Jeffrey Inman, and Hans Grande
ogist, 54 (7), 493–503. (2004), “The Three Faces of Consumer Promotions,”
Kuhl, J. (1994), “A Theory of Action and State Orienta- California Management Review, 46 (4), 23–42.

James H. Leigh
Texas A&M University
TAMU 4112
College Station, TX 77843–4112
Phone: 979.845.5893
FAX: 979.862.2811
E-Mail: jleigh@cgsb.tamu.edu

THE EFFECTS OF RESERVE PRICES ON BIDDING BEHAVIOR IN
ONLINE AUCTIONS
Marla Royne Stafford, The University of Memphis, Memphis
Ashley Kilburn, The University of Memphis, Memphis
Barbara B. Stern, Rutgers, The State University of New Jersey, New York
SUMMARY H2: The number of bidders participating in the auction

is positively related to the total number of bids
The influence of reserve prices on bidder behavior placed during the auction.
and bidding activity in online auction exchanges is exam-
ined. A reserve price is “the minimum price a seller is H3: The total number of bids placed throughout the
willing to accept for the item” (eBay 2002), and sellers use auction is positively related to the final bid price.
it to specify a hidden amount not disclosed to the bidder
that seller designates as a minimum acceptable bid price. H4: The presence of a reserve price is associated with
Despite the potential influence of reserves on online a higher final bid price.
bidding activity, there is little research about consumer
behavior in response to it. Insofar as general research on Method
consumer behavior in online auctions is quite sparse
(Chui and Zwick 1999; Wilcox 2000), the aim of our study To test the hypotheses, two sets of observational data
is to examine the impact of reserve prices on online bidder were collected from a list of Bay completed sales includ-
strategy, and in this way contribute to understanding ing all main eBay categories to ensure a variety of prod-
online consumer behavior. To do so, we investigated ucts and price ranges, but excluding real estate and cars.
buyers’ bidding responses to items with and without a Study 1 consisted of 198 listings (of which 14% had
reserve price. reserve prices) that were gathered over a one-month
period. To more adequately assess the impact of reserve
The theoretical basis of the study is grounded in the prices on bidding activity, a second stratified random set
risk perception literature, in which the level of perceived of data was collected with the intent of increasing the
risk is viewed as a powerful tool for predicting and number of items with reserve prices, while maintaining
explaining consumer behavior (Mitchell 1999). Under the sample characteristics from the first study (i.e., repre-
conditions of uncertainty (Bauer 1960; Cox and Rich senting all product categories and including those listings
1964) such as an online auction in which buyers must rely which generated activity). Therefore, Study 2 was de-
on a picture of the item and a verbal description in the signed to include at least one item with a reserve price in
absence of physical examination (Stern and Stafford, each selected subcategory (e.g., clocks, postcards and
forthcoming). Decision-making under uncertainty has paper, doll clothes and accessories, glassware, classic
been associated with risks of not only monetary loss but toys) resulting in a sample of 137 completed sales listings
also negative psychological and/or behavioral conse- each from different subcategories (of which 32.2% had
quences (Stone and Winter 1987; Sitkin and Pablo 1992). reserve prices). Two independent coders unaware of the
We propose that a seller’s decision to use a reserve price study’s purpose analyzed the items to determine the
diminishes the buyer’s level of perceived risk and out- presence or absence of a reserve price (coded as 1, 0,
come uncertainty by signaling to a potential bidder that respectively); the total number of bids placed; the total
until the reserve price is met, no contractual agreement number of different bidders; and the final bid amount.
obligates the bidder to buy. Inter-rater reliability was 100 percent, as expected in
analysis of observational data.
The study examines the following hypotheses:
Analysis and Results
H1A: The presence of a reserve price is positively related
to the number of bidders involved in an auction. AMOS 4.01 was used to test the hypotheses via path
analysis. The model fit was strong for Study 1. The
H1B: The presence of a reserve price is positively related Normed Fit Index (NFI) = .996, the Relative Fit Index
to the total number of bids placed during the (RFI) = .960, the Incremental Fit Index (IFI) = .997, and
auction. the Comparative Fit Index (CFI) = .997. The RMSEA =

.094, and the X2(1) = 2.75, p = .097. In short, the proposed robust, indicating once again that the proposed model
model seems to reproduce the covariance matrix well. reproduced the covariance matrix well; the NFI = 1.0, the
Further, four of the proposed hypotheses were supported RFI = .999, the IFI = 1.0, the CFI = 1.0. The RMSEA =
(H1A, H2, H3, H4). Results from Study 1 suggest that .000 and the X2 (1) = .072, p =.789. In contrast to the results
there is a positive direct effect of reserve prices on the from Study 1, all five hypotheses were supported included
number of bidders to participate in their auction, thus H1B (t = 2.6, p < .05), which suggests that the presence of
supporting H1A. Further, in accordance with H2, there is a reserve price is positively related to the number of bids
a positive direct effect of number of bidders on the number placed by an individual bidder. These findings suggest a
of bids placed. As suggested by H3, Study 1 found that an positive relationship between the use of reserve prices and
increased number of bids results in a higher final bid. number of bidders, number of bids, and final bid prices,
Finally, a significant direct relationship was found be- indicating that reserve prices may be an important tool for
tween the presence of a reserve price and a higher final bid online auction sellers. References available upon request.
price, showing support for H4. Study 2 results were more

Marla Royne Stafford
Fogelman College of Business and Economics
The University of Memphis
Memphis, TN 38152
Phone: 901.678.2499
FAX: 901.678.4052
E-Mail: mstaffrd@memphis.edu

ADVERTISING GOES MOBILE: EXPLAINING ATTITUDE
TOWARD M-ADVERTISING
Parissa Haghirian, Kyushu Sangyo University, Japan
SUMMARY H3b: The higher the informativeness of the mobile ad-

vertising message, the higher the perceived value
Devices and systems based on mobile technologies of m-advertising.
have become a commonplace in our everyday lives (Bal-
asubramanian, Peterson, and Jarvenpaa 2002) and have H4: The higher the irritation of mobile advertising
also changed the marketers’ world. Opportunities for message, the lower the perceived value of m-
direct contact with consumers are unprecedented. Con- advertising.
sumers can be provided with information they are interest-
ed in, which gives marketers the chance to build customer H5a: The higher the credibility of the mobile advertising
relationships of a new dimension. They also enable inter- message, the more positive consumers’ attitude
actions between customer and advertiser which become toward m-advertising.
increasingly rapid and easy. As a result, modern advertis-
ers are increasingly relying on various modes of interac- H5b: The higher the credibility of the mobile advertising
tive technology to advertise and promote their products message, the higher the perceived value of m-
and services (Pavlou and Stewart 2000). advertising.
Despite this, potential customers did so far not have H6a: The lower the age of the consumer, the more
many opportunities to signal their likes and dislikes with positive consumers’ attitude toward m-advertis-
marketing activities via mobile devices. This puts marketing.
ers at a high risk (Robins 2003). If marketers want to use
the communication channels that mobile media provide in H6b: The lower the age of the consumer, the higher the
an efficient way, they need to understand how mobile perceived value of m-advertising.
consumers perceive and evaluate mobile devices as a
source of advertising. A major issue in advertising re- H7a: Perceived value of m-advertising differs between
search is measurement of the effects of advertising on the men and women.
recipient. This paper aims to contribute to this objective
and presents results of a survey on consumer attitudes H7b: Attitude toward m-advertising differs between men
toward advertising via mobile devices (m-advertising). and women.
The objective of this study was to examine the relation-
ships between attitude toward m-advertising and key H8a: The lower the education of the consumer, the more
influence factors. positive consumers’ attitude toward m-advertis-
ing.
The following hypotheses were examined:
H8b: The lower the education of the consumer, the
H1: The higher the perceived value of m-advertising, higher the perceived value of m-advertising.
the more positive consumers’ attitude toward m-
advertising. The study was conducted over a six week period in
autumn 2003 during which 815 mobile phone owners
H2a: The higher the entertainment factor of the mobile were interviewed on their attitude toward m-advertising.
advertising message, the more positive consum- The interviewed persons have been selected on the basis
ers’ attitude toward m-advertising. of a quota sample that is representative for the Austrian
population. The Austrian market shows the highest pen-
H2b: The higher the entertainment factor of the mobile etration of mobile phone users in Europe and is therefore
advertising message, the higher the perceived val- very suitable for investigation on mobile marketing and
ue of m-advertising. advertising. All measures were assessed via a 5-point-
Likert-type scale ranging from “strongly disagree” to
H3a: The higher the informativeness of the mobile ad- “strongly agree.”
vertising message, the more positive consumers’
attitude toward m-advertising.

Evidence in support of hypotheses H4, H6a, H6b, Focus of the study are influence factors on consum-
H7a, H7b, and H8b could not be observed. Significant ers‘ attitude toward m-advertising. The results show that
relationships (at a .05 level) were observed between consumers’ attitude toward m-advertising but also per-
entertainment, informativeness, credibility, educational ceived value of m-advertising are strongly related to
level, and the dependent variable perceived value of m- advertising message content. Most surprisingly, consum-
advertising. Perceived value of m-advertising as well as ers‘ attributes (apart from educational level’s influence on
entertainment, informativeness, and credibility are re- attitude toward m-advertising) do not play such a domi-
flecting positively on attitude toward m-advertising. nant role regarding attitude toward m-advertising and
perceived advertising value. References are available
upon request.

Parissa Haghirian
Department of International Management
Kyushu Sangyo University
3-1-2 Matsukadai, Higashi-ku
Fukuoka 813–8503
Japan
Phone: 0081.92.673.5331
E-Mail: parissa@ip.kyusan-u.ac.jp

REGULATORY FOCUS AND RELATIONSHIP MARKETING SUCCESS
Maria Sääksjärvi, Swedish School of Economics and Business Administration, Finland
Johanna Gummerus, Swedish School of Economics and Business Administration, Finland
ABSTRACT tionship marketing tactics given that there is a fit between

the consumer’s inherent regulatory focus and that tempo-
Regulatory focus theory proposes consumers to nav- rally evoked by relationship tactics. Second, it affects a
igate towards goals using either a promotion or prevention customer’s relational orientation, thereby influencing
focus. We link regulatory focus to relationship marketing customer commitment indirectly. Acquiring relationally
success by proposing that the persuasiveness of relation- oriented customers is seen as the outcome of successful
ship tactics can be enhanced by inducing a regulatory fit, relationship tactics since they are likely to contribute to
and that regulatory focus impact relational orientation and firm profitability (Grönroos 1997). By recognizing what
thereby commitment. contributes to relational orientation and matching rela-
tionship tactics to fit a consumer’s inherent regulatory
INTRODUCTION focus, firms are likely to know “when and why consumers
respond favorably and strongly to companies’ relation-
Regulatory focus theory states that consumers pursue ship building efforts” (Bhattacharya and Sen 2003, p. 76).
goals with two different orientations: promotion or pre-
vention focus (Higgins 1987). Promotion-orientation in- The fit we are proposing contributes to existing
volves approaching pleasure, and relates to advancement, literature by suggesting that marketers that evoke tempo-
achievement, and aspirations. In contrast, prevention- ral regulatory focus and match it with a consumer’s
orientation involves avoiding pain in terms of negative inherent regulatory focus to yield optimal relationship
outcomes such as responsibilities, obligations, and secu- marketing strategy outcomes. We also contribute to the
rity (Higgins 2000). The type of focus evoked is of current literature by examining relational orientation,
importance as it has been found to affect goal pursuit, which despite its importance to relationship marketing
motivation, and perceived value (Higgins et al. 2003; has received little attention in academia, as well as its
Shah, Higgins, and Friedman 1998; Spiegel, Grant-Pil- impact on customer commitment.
low, and Higgins 2004).
We examine these issues by bringing forward a
Although regulatory focus has received attention number of propositions. Our conceptual framework de-
within social psychology, research addressing this con- picting these propositions is presented in Figure 1.
cept in marketing is still scarce (for exceptions see Aaker
and Lee 2001; Zhu and Meyers-Levy 2003). Linking the THE FIT BETWEEN CONSUMERS
concept of regulatory focus to relationship marketing is a AND RELATIONSHIP
fruitful avenue due to the established, but unexplored, MARKETING STRATEGIES
linkage between regulatory focus and relationship mar-
keting outcomes, such as the effectiveness of relationship Relational Mode
marketing tactics and commitment. Support for this link-
age comes, for example, from Fennell (1978) who postu- In this article, relational orientation is depicted by
lated that consumers faced with interest or pleasure op- means of a consumer’s relational mode, “a consumer’s
portunities (promotion goals) are more likely to explore relatively stable tendency to engage in a relationship with
brands, whereas those facing problems try to solve or a certain company” (Grönroos 1997). Relational mode is
prevent the problem (prevention goals). Of interest to an indicator of a consumer’s receptivity to relationship
marketers is also the proposition of an optimal fit between with a specific firm, and is characterized by consumers
personal goals and means to achieve these goals made in reacting favorably to relationship marketing activities
the regulatory focus literature. For instance, a person with (Grönroos 1997). Relational mode measures openness for
a promotion goal is more likely to exert behaviors that aim ongoing relationships and thus differs from a transaction-
for positive outcomes, whereas a person with a prevention al mode where consumers are only motivated to engage in
goal directs behaviors towards minimizing negative out- exchanges. It is thus seen as a pre-commitment phase that
comes. facilitates relationship maintenance and durability.
The central tenet of our paper is that regulatory focus As suggested by Grönroos (1997), the relational
impacts relationship marketing success on several levels. mode can be seen as a continuum ranging from passive to
First, regulatory focus enhances the effectiveness of rela- active. Consumers in an active mode seek contact, where-

FIGURE 1
P1 P6 Commitment
Regulatory Focus Relational Mode
- Affective, Normative,
Passive-Active
Calculative
P3 P2 P5 P4
Need for Connection/ Regulatory Focus
Belongingness (Temporally Induced)
Relationship
Tactics
as those in a passive mode rely on the existence of the We propose a consumer’s regulatory focus to be
relationship, considering it as a kind of safety net. More linked to his/her relational mode. More specifically, as
active consumers are often desirable from a marketing promotion-oriented consumers are consciously aspiring
point of view: the more active the customer’s role in towards goals, they believe in a relationship with a com-
service production is, the better the outcome. This is pany as it might enable goal achievement. They are thus
especially the case when the service requires customer likely to be more active than other consumers in the
participation or customer self-service reduces costs to a relationships they have as to ensure a successful outcome.
considerable extent. In contrast, prevention-oriented consumers are more re-
ceptive to establish relationships with firms to the extent
The relation between relational mode and regulatory they believe it helps them avoid losses or fulfill a duty or
focus is closely related to goal independence (Tjosvold obligation. This inclination, however, is likely to be
1993), which states that customer willingness to bond passive in nature, as it is based on vigilance and only doing
may depend on the extent to which customers believe they what is necessary (Shah and Higgins 1997). For example,
need the firm’s support to achieve important, personal a prevention-oriented consumer could be willing to be
goals. Regulatory focus is suggested to facilitate goal linked to a firm as a result of potential cost savings (avoid
pursuit (Shah, Higgins, and Friedman 1998) and is thus losing money) or out of a sense of duty (e.g., as a sense of
likely to relate to the effectiveness of a firm’s marketing reciprocity upon receiving excellent service), but is not
strategies. likely to actively contribute to relationship maintenance.
In essence, we propose promotion-orientation is likely to
Regulatory Focus contribute to an active willingness to engage in a relation-
ship, whereas prevention-orientation is likely to lead to a
Regulatory focus theory (Higgins 1987) brings for- passive one.
ward the idea that consumers not only care about what
goals they reach, but also how they reach them. It proposes P1a: Promotion-orientation has a positive impact on
consumers to utilize two different self-regulation strate- relational mode.
gies in goal striving: promotion or prevention orientation.
The promotion orientation emphasizes the pursuit of P1b: Prevention-orientation has a negative impact on
gains (or the avoidance of nongains) and aspirations relational mode.
toward ideals (Higgins 2000). It is related to eagerness;
consumers engage in tasks since they are motivated by Need for Connection
achievement and aspiration (Crowe and Higgins 1997).
The prevention orientation, on the other hand, emphasiz- Not all consumers, however, are interested in having
es the avoidance of losses (or the pursuit of nonlosses) and relationships with companies. Need for connection (Ander-
the fulfillment of obligations. It is related to vigilance sen and Chen 2002) or belongingness (Baumeister and
(Crowe and Higgins 1997). Consumers perform tasks as Leary 1995) describes “the need of people to form and
a sense of duty. They feel it is their responsibility, what maintain strong interpersonal relationships” and is mani-
they ought to do, but do not feel passionate about tasks the fested in “frequent, nonaversive interactions with an on-
same way as promotion-oriented consumers do. going relational bond” (Baumeister and Leary 1995).

Research in social psychology has also shown that the ceive they receive in return for their loyalty) (De Wulf,
mere belongingness of self to a certain object (thus refer- Odekerken-Schröder, and Iacobucci 2001). More inti-
ring to its ownership) leads to an increased liking of it mate, or higher levels of relationship tactics are signified
(Feys 1991). Given that relationships, as objects, can be by preferential treatment (the degree to which consumers
seen as extensions of the self (Andersen and Chen 2002) perceive that regular customers are being better treated
and that consumers are more likely to establish relation- than others) and interpersonal communication i.e., how
ships with service organizations than goods (Berry 1995) warm and personal the interaction between the company
we expect need for connection to have a positive effect on and the customer is (De Wulf, Odekerken-Schröder, and
relational mode. Iacobucci 2001).
P2: Need for connection has a positive effect on rela- Relationship tactics are likely to encourage custom-
tional mode. ers to become relationally involved and committed. For
example, warm and personal service, or participation in
We also expect a significant interaction effect to special events can show consumers the advantages of
occur between regulatory focus and need for connection. being actively involved with a company. Furthermore,
Although promotion-orientation is linked to eagerness customers may want to reciprocate and become more
and a more active relational mode, a consumer low in need active in the relationship if they perceive the relationship
for connection might not simply find relationships impor- tactics to offer value beyond the exchange. Companies
tant. As one of the main benefits of relationships to that use relationship tactics and thereby communicate a
customers is the perceived social bond the interaction long-term orientation are believed to restrain from oppor-
creates (Gwinner, Gremler, and Bitner 1997), those lack- tunistic behavior (Ganesan 1994) and are therefore likely
ing the need for such connections might not pursue them to be preferred by relation-oriented customers.
as means of achieving goals. For example, a promotion-
oriented consumer might purchase most items from a P4: Relationship tactics have a positive influence on
particular retailer, but might feel no need to belong or relational mode.
identify him/herself with it. In this case, the consumer is
more passive in the relationship. In contrast, need for The Perceived Fit Between Relationship Tactics and
connection might serve to make a prevention-oriented Regulatory Focus
consumer more active in the relationship. For example, a
consumer not wanting to avoid missing out on great Relationship tactics as other persuasive messages
promotions might be in active contact with the company usually involve some goal to be attained (e.g., an increase
if she/he feels a need to belong to their network of in the number of loyal customers). When firms employ
customers and interact with them on a frequent basis. In them, they present arguments in support of the advocated
this case, the need for connection heightens the consum- position, which can be framed as in terms of either gains
er’s activity level in the relationship while helping him/ or losses (Cesario, Grant, and Higgins 2004). For exam-
her achieve the same goal, thus contributing to a more ple, when trying to recruit customers for a loyalty pro-
active relational mode. gram, a firm could frame the arguments in terms of eagerly
ensuring gains (e.g., join our loyalty program – you will
P3a: For consumers high in need for connection, pre- receive special promotions) or in terms of losses (if you do
vention-orientation has a positive impact on rela- not join our loyalty program you’ll miss our special
tional mode. offers). Thus, the firm can temporally induce a regulatory
focus as means of influencing their customers.
P3b: For consumers low in need for connection, promo-
tion-orientation has a negative impact on relational If these arguments presented match the regulatory
mode. orientation of the customer, s/he should experience regu-
latory fit and feel right, which heightens his/her perceived
Relationship Tactics evaluation of an event and the value derived from it
(Higgins et al. 2003). Thus, we propose that if the relation-
Relationship tactics are means firms employ to in- ship tactics employed by a firm evoking a particular
crease customer retention. They can be viewed as a regulatory focus matches that of the customer they will be
continuum ranging from direct mail, tangible rewards, perceived as being more persuasive, enhancing consumer
and preferential treatment to interpersonal communica- willingness to engage in a relationship with a firm. Since
tion (Berry 1995). The lowest (i.e., least intimate) levels regulatory fit also enhances motivational strength (Spie-
of relationship tactics are represented by direct mail (the gel, Grant-Pillow, and Higgins 2004), we also propose
perceived degree of contact that companies have with consumers to take a more active role in the relationship if
their customer via direct mail) and tangible rewards (the there is a fit between the consumer’s and the firm’s
specific benefits e.g., savings in price that customer per- regulatory focus.

P5: Relationship tactics employed by a firm that fit the activities. Indeed, it has been suggested that affective
regulatory focus of the consumer contribute posi- commitment leads to expansion and enhancement of the
tively to a more active relational mode. relationship, whereas calculative commitment supports
the maintenance of the relationship (Bendapudi and Berry
Building commitment in customer relationships is 1997). Thus, we propose:
one goal of relationship marketing. Commitment is con-
sidered as beneficiary to companies since it reflects a P6a: An active relational mode contributes to affective
customer’s lasting orientation to maintain a long-term commitment.
relationship (Dwyer, Schurr, and Oh 1987; Moorman,
Zaltman, and Deshpande 1992; Morgan and Hunt 1994), P6b: A passive relational mode contributes to normative
and is coupled with excluding other alternatives and being or calculative commitment.
resistant to change (Pritchard, Havitz, and Howard 1999).
Commitment differs from relational mode by being a CONCLUSIONS AND FUTURE RESEARCH
lasting tendency gradually built over time, instead of an
inclination or disposition towards relationship building. It The present article brought forward a number of
can thus be seen as an outcome of a customer’s relational propositions concerning the link between relationship
mode. marketing tactics and regulatory focus. Empirical testing
should be undertaken to verify the propositions we have
Commitment takes different forms depending on the presented. To assess fit, regulatory focus could be induced
customers’ reason for staying in a relationship. Previous by assessing consumer’s inherent regulatory focus (using
literature highlights three mains reasons as of why cus- the self-guide strength measure, see Higgins and Spiegel
tomers stay in relationships: (1) they want to (affective 2004) and then match it with a situationally induced
commitment), (2) they perceive they have no other choice marketing message framed as either involving a gain
(calculative commitment) or (3) they feel obliged to do so (promotion) or a loss (prevention).
(normative commitment). Affective commitment cap-
tures issues such as contentment and affection and is The linkage between relational mode and regulatory
based on emotional attachment to a firm (Allen and Meyer focus can be extended and studied at three different levels
1990). Calculative commitment in turn stems from an of relationships specified in relationship marketing liter-
acknowledgment of costs associated with leaving a firm ature, namely relationship initiation, maintenance and
(Meyer and Allen 1991) and is often characterized by “a dissolution (Grönroos 1990). For example, during rela-
sense of being locked in” (Meyer and Herscovitch 2001, tionship initiation, a consumer’s regulatory focus could
p. 307). Normative commitment can be best described as guide his/her relationship formation in the following
a feeling of obligation to stay in a relationship (Fehr 1999; manner. A consumer with a promotional focus is likely to
Meyer and Allen 1991). Obligation as a state of mind engage in relationship with a firm that emphasized the
could also be related to consideration of moral rightness positive outcomes of it. In opposition, a consumer with a
of maintaining a relationship (Marsh and Mannari 1977). prevention focus will choose a relationship that s/he
perceives as safe, i.e., encompassing a minimum likeli-
We suggest that affective commitment stems from hood of negative outcomes or avoid the obligations relat-
active relational mode that induces customers to bind ed to relationships and choose not to enter a relationship.
themselves to relationship-building efforts and thereby Similarly, a promotion focus may trigger relationship
deepens the relationship, whereas calculative and norma- dissolution since it is associated with openness for change,
tive commitment follow from passive relational mode in whereas a prevention focus is likely to prevent the con-
which the customer is more inclined to engage in relation- sumer from leaving an existing relationship, since it poses
ship maintaining activities than relationship enhancing a threat to stability (Liberman et al. 1999).
REFERENCES surement and Antecedents of Affective, Continu-

ance, and Normative Commitment to the Organiza-
Aaker, Jennifer L. and Angela Y. Lee (2001), “I Seek tion,” Journal of Occupational and Organizational
Pleasures and We Avoid Pains: The Role of Self- Psychology, 63 (1), 1–18.
Regulatory Goals in Information Processing and Per- Andersen, Susan M. and Serena Chen (2002), “The Rela-
suasion,” Journal of Consumer Research, 28 (1), 33– tional Self: An Interpersonal Social-Cognitive Theo-
49. ry,” Psychological Review, 109 (4), 619–45.
Allen, Natalie J. and John P. Meyer (1990), “The Mea- Baumeister, Roy F. and Mark R. Leary (1995), “The Need

to Belong: Desire for Interpersonal Attachments as a 319–40.
Fundamental Human Motivation,” Psychological ____________ (1997), “Beyond Pleasure and Pain,”
Bulletin, 117 (3), 497–529. American Psychologist, 52 (December), 1280–300.
Bendapudi, Neeli and Leonard L. Berry (1997), “Custom- ____________ (2000), “Making a Good Decision: Value
ers’ Motivations for Maintaining Relationships with from Fit,” American Psychologist, 55, 1217–30.
Service Providers,” Journal of Retailing, 73 (1), 15– ____________, Lorraine C. Idson, Antonio L. Freitas,
37. Scott Spiegel, Daniel C. Molden (2003), “Transfer of
Berry, Leonard L. (1995), “Relationship Marketing of Value from Fit,” Journal of Personality and Social
Services – Growing Interest, Emerging Perspectives,” Psychology, 84 (6), 1140–53.
Journal of the Academy of Marketing Science, 23 (4), ____________ and Scott Spiegel (2004), “Promotion and
236–45. Prevention Strategies for Self-Regulation: A Moti-
Bhattacharya, C.B. and Sankar Sen (2003), “Consumer- vated Cognition Perspective,” in Handbook of Self-
Company Identification: A Framework for Under- Regulation: Research, Theory and Applications, Roy
standing Consumers’ Relationships with Compa- F. Baumeister and K.D. Vohs, eds. New York: Guil-
nies,” Journal of Marketing, 67 (April), 76–88. ford Press, 171–87.
Cesario, Joseph, Heidi Grant, and E. Tory Higgins (2004), Liberman, Nira, Lorraine C. Idson, Christopher J. Cama-
“Regulatory Fit and Persuasion: Transfer From ‘Feel- cho, and E. Tory Higgins (1999), “Promotion and
ing Right,’” Journal of Personality and Social Psy- Prevention Choices Between Stability and Change,”
chology, 86 (3), 388–404. Journal of Personality and Social Psychology, 77
Crowe, Ellen and E. Tory Higgins (1997), “Regulatory (6), 1135–45.
Focus and Strategic Inclinations: Promotion and Pre- Marsh, Robert M. and Hiroshi Mannari (1977), “Organi-
vention in Decision-Making,” Organizational Be- zational Commitment and Turnover: A Prediction
havior and Human Decision Processes, 69 (2), 117– Study,” Administrative Science Quarterly, 22, 57–
32. 75.
De Wulf, Kristof, Gaby Odekerken-Schröder, and Dawn Meyer, John P. and Natalie J. Allen (1991), “A Three-
Iacobucci (2001), “Investments in Consumer Rela- Component Conceptualization of Organizational
tionships: A Cross-Country and Cross-Industry Ex- Commitment,” in Human Resource Management
ploration,” Journal of Marketing, 65, 33–50. Review, Vol. 1, Elsevier Science Publishing Compa-
Dwyer, F. Robert, Paul H. Schurr, and Sejo Oh (1987), ny, Inc.
“Developing Buyer-Seller Relationships,” Journal ____________ and Lynne Herscovitch (2001), “Com-
of Marketing, 51, 11–27. mitment in the Workplace: Toward a General Mod-
Fehr, Beverley (1999), “Laypeople’s Conceptions of el,” Human Resource Management Review, 11 (3),
Commitment,” Journal of Personality and Social 299–326.
Psychology, 76 (1), 90–103. Moorman, Christine, Gerald Zaltman, and Rohit Desh-
Fennell, Geraldine (1978), “Consumers’ Perceptions of pande (1992), “Relationships Between Providers and
the Product-Use Situation,” Journal of Marketing, Users of Market Research: The Dynamics of Trust
42, 38–47. Within and Between Organizations,” Journal of
Feys, Jos (1991), “Briefly Induced Belongingness to Self Marketing Research, 19 (3), 314–28.
and Preference,” European Journal of Social Psy- Pritchard, Mark P., Mark E. Havitz, and Dennis R. Howard
chology, 21, 547–52. (1999), “Analyzing the Commitment-Loyalty Link
Ganesan, Shankar (1994), “Determinants of Long-Term in Service Contexts,” Journal of the Academy of
Orientation in Buyer-Seller Relationships,” Journal Marketing Science, 27 (3), 333–48.
of Marketing, 58 (2), 1–19. Shah, James and E. Tory Higgins (1997), “Expectancy ×
Grönroos, Christian (1990), Service Management and Value Effects: Regulatory Focus as Determinant of
Marketing: Managing the Moments of Truth in Ser- Magnitude and Direction,” Journal of Personality
vice Competition. Lexington, MA: Lexington. and Social Psychology, 73 (3), 447–58.
____________ (1997), “Value-Driven Relational Mar- ____________, E. Tory Higgins, and Ronald S. Friedman
keting: From Products to Resources and Competen- (1998), “Performance Incentives and Means: How
cies,” Journal of Marketing Management, 13, 407– Regulatory Focus Influences Goal Attainment,” Jour-
19. nal of Personality and Social Psychology, 74 (2),
Gwinner, Kevin P., Dwayne D. Gremler, and Mary Jo 285–93.
Bitner (1998), “Relational Benefits in Services In- Spiegel, Scott, Heidi Grant-Pillow, and E. Tory Higgins
dustries: The Customer’s Perspective,” Journal of (2004), “How Regulatory Fit Enhances Motivational
the Academy of Marketing Science, 26 (2), 101–14. Strength during Goal Pursuit,” European Journal of
Higgins, E. Tory (1987), “Self-Discrepancy: A Theory Social Psychology, 34, 39–54.
Relating Self and Affect,” Psychological Review, 94, Tjosvold, Dean (1993), Teamwork for Customers: Build-

ing Organizations That Take Pride in Serving. San of Regulatory Focus on Consumer Information Pro-
Francisco: Jossey-Bass Publishers. cessing,” in Advances in Consumer Research, 30,
Zhu, Rui and Joan Meyers-Levy (2003), “The Influence 116–17.

Maria Sääksjärvi
Department of Marketing and Corporate Geography
HANKEN – Swedish School of Economics and Business Administration
Arkadiankatu 22
P.O. Box 479
00101 Helsinki
Finland
Phone: +358.9.4313.3227
FAX: +358.9.4313.3287
E-Mail: maria.saaksjarvi@hanken.fi

PREDICTING USAGE LEVEL AND UPGRADING BEHAVIOR OF
SERVICE CUSTOMERS: A MODEL FOR LIFETIME VALUE
ESTIMATION AT EARLY RELATIONSHIP STAGES
Florian V. Wangenheim, University of Dortmund, Germany
SUMMARY where
The concept of customer lifetime value (CLV) is CLV, = Lifetime value of customer I.
receiving increased attention in the current literature on
customer equity (CE) and on linking marketing to finan- TNT, = Total number of transactions made by
cial performance. Previous CLV research (e.g., Rust, customer I in period t.
Lemon, and Zeithaml 2004) has investigated predictabil-
ity of CLV and its drivers mainly in established customer CMNRT = Average contribution margin of base
relationships. The objective of the present paper is to transaction.
investigate how future customer transaction behavior and
lifetime value can be forecasted particularly at an early PUPit = Proportion of “upgrade” transactions
stage of the customer relationship. By differentiating by customer I in period t.
between frequency of customer transactions and upgrad-
ed transactions, not only CLV, but also the relevant CLV ∆CMUp = Additional contribution margin for high
drivers can be determined. In addition, for a subsample of value transactions.
customers, the paper links the derived CLV prediction
error to survey data, thereby determining whether predic- d = Discount rate.
tion inaccuracies can be partially explained by customer
attitudes that reflect the firms’ failure to retain a potential- t = 1. . .T = Number of time periods considered.
ly high-value customer.
CLV Driver Prediction
The key focus of the paper is relationship depth
(while relationship length is inherently considered due to As CLV predictors, customer channel usage, enable-
the longitudinal study design). In accordance with Bolton, ment of direct communication, competitiveness of choice,
Lemon, and Verhoef (2004), relationship depth is defined past behavior, and demographics are used. Channel choice
as the intensity or usage level of services over time, and refers to the differentiation between direct and indirect
specifically includes upgrading behavior (i.e., purchasing channel usage. Direct communication enablement is re-
of higher value or margin products instead of low cost lated to the decision whether to allow the firm to get into
“basic” alternatives). Consequently, in the present study direct contact with the customer. Competitiveness of choice
which focuses on the airline industry, relationship depth refers to the variety of providers from which customers
is conceptualized as two-dimensional: Usage level per choose. Past behavior is given by exhibited flying behav-
period is reflected by the number of flights per period. ior at the time the prediction is made.
Upgrading refers to the proportion of upgrades, i.e.,
flights in a “higher” booking class (where the baseline Model, Estimation, and Results
level is economy class) in the same period. Focusing on
those two drivers of the relationship depth dimension Data on the above described dependent and indepen-
rather than directly modeling the monetary transaction dent variables are taken from the frequent filer customer
value per period offers additional insights for firms and base of a large European airline. Two samples, one
academics, as it allows to differentiate between, e.g., consisting of 2,900 customers acquired in the first quarter
high-value customers whose value mainly derives from a of 1999, and one consisting of 3,200 customers acquired
large number of transactions, and another, potentially in the first quarter of 2000, the latter serving as hold-out
equally valuable group, which is valuable not because of sample, are analyzed. The predictive power of the inde-
a large number of transactions, but because it has a strong pendent variables on number of transactions and propor-
disposition to use high-value services (i.e., business class). tion of upgraded transactions is tested using a nested tobit
In general, the CLV of a customer in this context is given model. The model results yield strong support for an effect
by of all predictors on number of transactions, and also
T
TNTit [ CM NRT + (PUPit . ∆CM)] support for the effect of past upgrading behavior and
CLVi = Σ
t=1 (1 + d)t
(1) direct communication enablement on future upgrading

behavior. Fit of the estimated models is generally well, a high predicted CLV, and those could receive loyalty
and hold-out predictions do not perform much worse, program benefits before they have reached the formal
suggesting some stability in the relationships researched. “hurdle” set by the firm. Further, firms should try to move
their customers towards direct channels and direct com-
Further, customer satisfaction and share of wallet munication. Once customers have made those specific
data that are available for a subsample of those customers investments, firms are better able to increase usage fre-
are correlated with the standardized residuals from a quency and upgrading behavior.
regression model where predicted CLV is regressed on
actual CLV. Since negative residual values indicate that The fact that the study focused on two central drivers
CLV was overpredicted, both SOW and satisfaction are of CLV jointly (i.e., number of transactions and upgrad-
expected to be positively correlated with the residuals ing) offers additional insights for CLV-driver based mar-
from the regression. The results show that in all cases, ket segmentation. Based on the model, firms can early on
there is a positive and statistically significant correlation in the relationship predict which customer will be valu-
between the residuals and the share of wallet data, sug- able mainly due to which driver, and firms can make
gesting that indeed customers with a low share of wallet specific offers addressing the specific needs of “upgrad-
were systematically overvalued. Further, customer satis- ing” customers as opposed to “frequency” customers.
faction is also positively related to both the residuals and Customer segmentation solely based on CLV would not
the SOW data, suggesting that customers overvalued by be able to identify those distinct segments.
the model tend to be less satisfied with the airline, and that
increases in satisfaction would increase SOW, thereby Future research should compare industries with re-
decreasing the gap between predicted and actual CLV. gard to the effect of predictor variables and the degree of
CLV predictability. In various service settings, other CLV
Discussion, Implications, and Conclusion components than those addressed in the present applica-
tions, especially cross-buying, are important CLV driv-
The results of the study suggest that managers can, at ers. It would be interesting to extend the applied CLV
a fairly early stage in the customer relationship predict model to account for cross-buying and further CLV com-
their customer’s future CLV. There are a number of ponents. References available upon request.
opportunities for firms to concentrate on customers with

Florian V. Wangenheim
Service Management
Otto-Hahn-Str. 6
University of Dortmund
44227 Dortmund
Germany
Phone: +49.231.755.4611
FAX: +49.231.755.5254
E-Mail: Florian.Wangenheim@uni-dortmund.de

THE IMPACT OF PERCEIVED LANGUAGE STATUS ON PRODUCT AND
SERVICE QUALITY EXPECTATIONS
Melissa Maier Bishop, University of Texas at Arlington, Arlington
SUMMARY living in the United States. Specifically, Hispanics dem-

onstrated a higher affect for advertising that used at least
Decisions to standardize or localize international and some amount of English as compared to advertisements
cross-cultural advertising campaigns are usually made that used all Spanish when controlling for cultural sensi-
with a cost-benefit analysis (Onkvisit and Shaw 1987), as tivity of the advertiser. Extended to PQ and SQ, the
it is expensive to target every ethnic group within a region language used in communication and its perceived status
with an individualized campaign. Regarding ethnic lan- may affect formation of these expectations through the
guage, few advocate advertising in a single, more domi- phenomenon of the halo effect.
nant language (such as English or another Westernized
language) outside of the benefits achieved through costs- The halo effect has been defined as the tendency to let
savings. The approach to look at the possible beneficial our assessment of an individual on one trait influence our
effects of standardized language advertising or the poten- evaluation of that person on other specific traits (Blum
tial negative effects of localized language advertising that and Naylor 1968). Halos have been typically treated as
extend beyond the “bottom line” is not normally taken (for rater errors and viewed as something to overcome in the
a notable exception, see Koslow, Shamdasani, and Touch- research process (Murphy, Jako, and Anhalt 1993). How-
stone 1994). ever, they can be beneficial to advertisers that understand
how to effectively employ them. For example, halo effects
In terms of product quality (PQ) and service quality may contribute to increased or decreased quality expecta-
(SQ) expectations, it is not clear how perceptions of tions based on the perceived status of the language used in
language used in advertising affect their formation. In- advertising. That is, positive or negative perceptions of
creasing expectations regarding both PQ and SQ is impor- the status of the language used in advertising may transfer
tant, as individuals tend to interpret stimuli consistent to congruent expectations regarding PQ and SQ. Thus, the
with their initial expectations (Hawkins, Best, and Coney following proposition is proposed:
2004). Further, higher product or service quality expecta-
tions are generally believed to lead to higher patronage or Proposition: The relationship between product and
purchase intentions. For example, previous research ef- service quality expectations and language used in
forts have identified a direct relationship between service advertising depends on the perceived status of the
quality and patronage intentions (Sirohi, McLaughlin, language. That is, using a language in advertising that
and Wittink 1998; Zeithaml, Berry, and Parasuraman is viewed as relatively superior (inferior) in language
1996). It is possible that the perceived status of the status will result in higher (lower) product and ser-
language used in advertising may be an important moder- vice quality expectations among viewers.
ator in the formation of both PQ and SQ expectations.
If the above proposition holds, advertisers may gain
Language status refers to the perceived relative posi- increased insight into the need for standardized or local-
tion or standing of an ethnic language in a society. Strong ized campaigns that goes beyond issues related to costs. If
support has been found for the existence of the language- a foreign language is viewed as superior to the local one,
related inferiority complex (LRIC) in which one language advertisers may benefit due to halo effects, in which
is viewed to be inferior to another (see Haarmann 1986; positive perceptions of the language transfer to positive
Koslow, Shamdasani, and Touchstone 1994; Christie expectations of the product. These findings may be partic-
1990; Adegbija 1994). The LRIC phenomenon has been ularly relevant to advertised products that require low
found in many countries and in many cultural groups, and levels of information processing, such as commodities.
varies by severity. For example, Roos (1997) found that Advertising in a “superior” language may give relatively
although the government of South Africa officially recog- undifferentiated products such as these an advantage over
nizes eleven languages, the process of developing these products advertised in the “inferior” language, due to
languages “is complicated by negative attitudes which increased PQ expectations – even if slight. “Image” prod-
hold that indigenous languages are somehow ‘inferior’ to ucts or services may also benefit from using a foreign
English and other Western languages” (p. 171). Applied language that is perceived as having a higher status so to
to advertising, Koslow, Shamdasani, and Touchstone equate the image (or status) of the product with the image
(1994) found support for the LRIC among Hispanics (or status) of the foreign country. For example, Takashi

(1990) found this effect among younger Japanese con- a communicative tool. Ethnic language may also contrib-
sumers, who viewed products that were advertised with ute to image formation by cueing product and service
some amount of English as more sophisticated and mod- quality expectations. Empirical work should focus on
ern as compared to products advertised in Japanese. measuring perceived language status across a variety of
languages and ethnic groups to determine how it can
In conclusion, the primary purpose of this study is to strategically influence these expectations. References
encourage researchers to consider language as more than available upon request.

Melissa Bishop
University of Texas at Arlington
Arlington, TX 76019
Phone: 817.272.2330
FAX: 817.272.2854
E-Mail: mbishop@uta.edu

EFFECTS OF POSITIONING A FOREIGN BRAND AS A DOMESTIC
BRAND IN COUNTRIES WITH DEVELOPED (U.S.) VERSUS
TRANSITIONING (ROMANIA) MARKET ECONOMIES
Lada V. Kurpis, Gonzaga University, Spokane
Simona Stan, University of Oregon, Eugene
Carmen Barb, University Lucian Blaga, Romania
ABSTRACT association of the custom brands with their foreign owner

is not mentioned in marketing communications and, pre-
When companies enter foreign markets they usually sumably, it remains unknown to or it is overlooked by the
market their brands by openly communicating the “for- majority of their consumers.
eign-owned” status of their brands. However, these com-
panies sometimes prefer to develop custom brands for a Creating the appearance that a foreign brand is do-
particular international market and position these brands mestic might enhance or lower brand attitudes when
as seemingly domestic for that market. This papers reports consumers become aware of the true origin of the brand.
the results of an exploratory study conducted in the U.S. On the one hand, consumers may perceive the foreign
and Romania which suggest that, in a transitioning econ- company’s practice of LCCP as a positive effort to cus-
omy, consumers are likely to reward a foreign marketer’s tomize the brand to their specific market. On the other
effort to customize the brand so that it seems domestic. On hand, consumers may perceive the practice as a manipu-
the contrary, in a developed economy, consumers are lative attempt. Whether a positive response prevails over
likely to penalize a marketer’s attempt to position a a negative response may depend on the international
foreign brand as a seemingly domestic one. The paper market’s characteristics, such as level of economic devel-
provides theoretical arguments based on differences in opment. Research suggests that there is a generalized
consumers’ persuasion knowledge, inferences of unfair preference for nonlocal brands in developing economies,
marketing tactics, and perceptions of expended marketing due mainly to symbolic, status-enhancing reasons (e.g.,
effort. Batra et al. 2000). However, it is not clear whether this
preference would be affected by a foreign company’s
INTRODUCTION explicit use of a LCCP or FCCP strategy. No known
research has addressed these issues, in spite of the signif-
When established companies enter foreign markets icant attention paid to branding in international markets.
they usually market their existing brands by openly
communicating the “foreign-owned” status of their brands. The purpose of this paper is to explore the following
Such brands often employ a “foreign consumer culture research questions. First, does the practice of positioning
positioning” (FCCP), in which the brand is associated a foreign brand as a seemingly domestic brand, as in a
with a specific foreign culture, through symbols, images, LCCP strategy, have advantages over openly communi-
and/or language (Alden, Steenkamp, and Batra 1999). cating the foreign origin of the brand, as in a FCCP
However, these companies sometimes prefer to develop strategy? Second, will this practice have different results
custom brands for a particular international market and in a developed free market versus a transitioning econo-
position these brands as seemingly domestic for that my? The paper starts with a brief review of several
market. These brands would employ what Alden et al. call theoretical perspectives, followed by hypotheses and the
a “local consumer culture positioning” (LCCP), by using description of an exploratory study conducted in the U.S.
symbols of the local consumer culture. For example in and in Romania.
Russia, in addition to its traditional brands, Swiss-based
Nestle Corporation produces and markets Rossia choco- THEORETICAL BACKGROUND
lates and Choc candy bars. The advertising themes for
both brands are very “a la Russ” (Russian-like) styled. Perceived Marketing Effort
Similarly, tobacco factory RJR-Petro in St. Petersburg,
Russia, owned by RJR Nabisco Holdings, produces and Kirmani and Wright (1989) demonstrate that under
markets Peter I cigarettes along with Winstons and Cam- certain conditions consumers use the perceived effort a
els (The Peter I brand bears the name of the first Russian company invests in a persuasion attempt (e.g., the cost of
emperor Peter the Great, which confers a positive appeal advertising) as a signal of their strength of belief in their
to the glory of the Russian past for most Russians). The product quality, resulting in more favorable brand atti-

tudes. Kirmani and Wright (1989) conceptualize per- developed and more accessible body of persuasion knowl-
ceived effort in terms of marketer’s investment of scarce edge than consumers in a transitioning economy. In par-
resources such as money, managerial time, thought, etc. ticular, consumers in a developed economy might have
They propose that spending more than other brands in the better understanding of the tactical meanings of messag-
product category or spending large amounts relative to the es. The PKM posits that if a consumer becomes aware of
company’s available resources would be considered ef- the tactical meaning of messages (e.g., use of celebrities in
fortful. advertising), natural processing of a persuasion message
might be disrupted and intended persuasion will be under-
Consumers in developed economies, such as the mined (Friestad and Wright 1994). Hence, it is expected
U.S., are used to the foreign manufacturers who are ready that persuasion is undermined more for consumers in a
to go at great lengths in customizing their brands specif- developed economy than in a transitioning economy. All
ically for the lucrative American market, with examples of these theoretical considerations conduct to the following
such brands ranging from Toyota Tacoma trucks to the hypotheses:
Swiss-owned Poland Springs water. Therefore, it is likely
that American consumers will not perceive efforts associ- H1: In a developed free market economy, a foreign brand
ated with developing yet another customized brand as positioned as a seemingly domestic brand will evoke
something out of the ordinary. On the contrary, consum- less favorable responses (brand attitudes and pur-
ers in transitioning economies are not “spoiled” by having chase intentions) than the same foreign brand posi-
too many foreign brand offerings customized for their tioned as a foreign brand.
local markets. Consequently, they are more likely to
perceive the company’s marketing effort as substantial H2: In a transitioning economy, a foreign brand posi-
and positive. tioned as a seemingly domestic brand will evoke
more favorable responses (brand attitudes and pur-
Inference of Unfair Marketing Tactics chase intentions) than the same foreign brand posi-
tioned as a foreign brand.
Although appealing to the values of the local con-
sumers seems to be an advantageous strategy for a foreign METHODOLOGY
brand, marketing literature contains some indications that
the favorable effect might be weakened, and even re- The hypotheses were tested using a between-subjects
versed, if the true origin of the brand becomes known to experimental design in which subjects were exposed to
the consumer. Campbell (1995) demonstrates that if con- three conditions: a foreign brand positioned as a seeming-
sumers infer a manipulative intent on the part of the ly domestic brand using LCCP (custom foreign); a foreign
advertiser, advertising persuasion is lowered. We suggest brand positioned as a foreign brand using FCCP (foreign
that the likelihood of consumers’ interpreting the posi- “as is”); and a truly domestic brand (used as a base level).
tioning of a foreign brand as a seemingly domestic brand The experiment was performed in the U.S., a developed
to be a manipulative intent (i.e., deception) is higher in a market, and in Romania, a developing Eastern-European
developed than in a transitioning economy. The rationale market which continues to struggle with the transition
is that consumers in developed markets have more expe- process from a centralized to a free market economy.
rience with marketing, have benefitted from more con- Certainly, country-of-origin effects (e.g., Batra et al.
sumer education, and therefore have more persuasion 2000; Maheswaran 1994) are capable of augmenting or
knowledge. attenuating the responses to a foreign brand positioned as
a seemingly local one. As this research is not interested in
Persuasion Knowledge COO effects, efforts were made to identify a product
category and a country-of-origin for the foreign brand that
Lay persons’ persuasion knowledge as described by would have similar COO effects both in Romania and in
Friestad and Wright (1994, 1995) consists of a system of the U.S. Based on interviews and pretests, chocolate was
beliefs that consumers hold about marketers’ persuasion chosen as the product category and Belgium as the foreign
goals and tactics, as well as about their own coping country of origin. Both Romania and the U.S. have strong
strategies in dealing with persuasion attempts delivered domestic chocolate brands and both recognize Belgium,
by the marketers. The Persuasion Knowledge Model along with Switzerland and Germany as “chocolate coun-
(PKM) posits that lay persuasion knowledge is develop- tries.”
mentally contingent and improves with frequent practice
(Friestad and Wright 1994). Since consumers in a devel- Subjects
oped economy have to cope with a larger number of
persuasion attempts (e.g., telemarketing, advertising) it Subjects were recruited among the students of two
seems plausible that these consumers will have a better mid-sized public universities in the U.S. (n = 99) and

Romania (n = 97) to participate in a marketing product brand attitude index. Purchase intentions regarding Rhap-
concept test for a fictitious brand of boxed chocolate. sody were measured on a 5-point scale with response
Respondents in both samples did not differ significantly options from (1) definitely will not buy to (5) definitely will
in terms of age (mean = 22.5 in Romania; mean = 23.3 in buy. An open-ended question: “What do you think the
the U.S., F(1,207) = 2.4, n.s.), or gender composition company is trying to tell you by developing a special
(42.7% female in Romania, 50.5% female in the U.S., design of packaging featuring [name of the landmark] for
χ2(209) = 1.26, n.s.). this market?” was used to assess the cognitions underly-
ing the responses. Company’s sincerity was measured
Procedure using two 7-point semantic differentials (insincere – sin-
cere, dishonest – honest). Country-of-origin effects were
A fictitious brand name Rhapsody (in English) and measured by asking the subjects to rate perceived quality
Rapsodia (in Romanian) was selected on the basis of pre- of chocolate from a number of countries, including the
testing for use with the American and Romanian samples, target countries of Belgium, USA, and Romania on a 7-
respectively. All subjects saw the same picture of fine point scale from 1 (very poor quality) to 7 (very high
chocolate confections and read the same paragraph de- quality) with an additional response category 0 (“I don’t
scribing the test brand as a high-quality brand of gourmet know”). To test whether subjects believed that featuring a
chocolate that is owned and manufactured by a family- national landmark on the packaging would be sufficient
owned company with a long successful history of making for misleading the customers about the origins of the
chocolate confections. The brand’s owner was described brand, respondents were asked: “If an average consumer
as either domestic (American or Romanian, respectively) just sees the packaging of this chocolate brand, how likely
or foreign (Belgian, for both the U.S. and Romanian is it that he/she will perceive it as a genuine American
samples). Different pictures featured on chocolate pack- [Romanian] brand?” (1–very unlikely genuine; 7–very
aging were used to manipulate the three conditions. In the likely genuine).
truly domestic brand condition and the custom foreign
(foreign positioned as domestic) brand condition, the Manipulation Check
design of the chocolate box featured a landmark from the
target market country. The American subjects saw a box As expected, the subjects in both countries thought
featuring the Golden Gate Bridge and the Romanian that the custom brand was just as likely to be considered
subjects saw a box featuring the Romanian Atheneum. In domestic as a true domestic brand (t(68) = .16, n.s. in
the foreign “as is” brand condition, the chocolate box Romania; t(64) = -.41, n.s. in the U.S.). The respondents
featured a Belgian landmark, a Canal in Bruges (see also believed that the foreign brand was much more likely
Table 1). to be considered domestic when it featured a landmark
from the target market country (Foreign – Custom Posi-
Measures tioning condition) relative to the case when it used a
landmark from the country of origin (Foreign – No Cus-
Brand attitude was measured using three 7-point tom Positioning): t(63) = 4.3, p < 0.001 in Romania;
semantic differential scales (bad – good, poor quality – t(61) = 9.7, p < 0.001 in the U.S. These results suggest that
high quality, not appealing – appealing). Reliability of subjects believed that the packaging for the custom brand
the three-item brand attitude scale reached .81 in Romania was likely to mislead the uninformed consumer about the
and .86 in the U.S. The items were averaged to form a brand’s true origin.
TABLE 1
Summary of Experimental Conditions
Brand Origin Brand Positioning
Romanian Romanian Local: Romanian landmark (Atheneum) picture on the package

Sample Belgian Custom foreign (seemingly local): Romanian landmark (Atheneum)
Belgian Foreign “as is”: Belgian landmark (The Grand Canal in Bruges)
American American Local: American landmark (The Golden Gate) picture on the packaging
Sample Belgian Custom foreign (seemingly local): American landmark (The Golden Gate)
Belgian Foreign “as is”: Belgian landmark (The Grand Canal in Bruges)

RESULTS positioning a foreign brand as seemingly domestic re-
mains advantageous even after target consumers learn
First, the direction and magnitude of the country-of- about the true origin of the brand, the attitudes toward the
origin effect was evaluated. For both the Romanian and custom foreign brand were compared to the attitudes
the U.S. sample, Belgian chocolate, in general, is per- toward the foreign brand positioned as foreign (“as is”).
ceived as of significantly better quality than the domestic These differences reached marginal significance for both
chocolate (mean = 5.4 versus 3.7, t = 5.4, p < 0.001 for the the Romanian (t(62) = 1.47; p = .07, one-tailed test) and
Romanian sample; mean = 6.2 versus 4.5, t = 6.6, p < 0.001 the U.S. (t(61) = -1.4, p = .08, one-tailed) sample. As
for the U.S. sample). Hence, both the Romanian and the expected, the Romanian subjects liked more the foreign
U.S. samples showed country-of-origin effects for Bel- brand positioned as a seemingly domestic brand, while the
gium in the same direction and of similar magnitude. American subjects liked more the foreign brand posi-
tioned as foreign. These findings suggest some support
T-test analyses were used to test the directional hy- for the advanced hypotheses. A comparison of the pur-
potheses regarding brand attitudes and purchase inten- chase intentions failed to render any significant differenc-
tions. The results are presented in Table 2 and Figure 1. To es. However, the pattern in purchase intentions was sim-
test whether the strategy of developing custom brands by ilar to the pattern in brand attitudes.
TABLE 2
Means for the Dependent Variables by Country and Brand Positioning
Positioning Brand Attitude Index Purchase Intentions
Romania Local 5.15 3.6
Custom Foreign 5.7 3.9
Foreign “as is” 5.2 3.8
U.S.A. Local 4.9 2.7
Custom Foreign 4.7 2.5
Foreign “as is” 5.2 2.8
FIGURE 1
Brand Attitude Index Means by Brand Positioning and by Country
5.8
5.6 Romania
5.4
5.2 USA
5.0
4.8
4.6
4.4
4.2
Local Custom Foreign “Foreign as is”
(seemingly local)

In order to explore in more depth the reasons why nian culture and tradition; they used the picture of the
consumers evaluated a foreign brand positioned as seem- Atheneum because it is an important Romanian landmark
ingly domestic differently than a foreign brand positioned and therefore it would give the product a Romanian
as foreign, several other analyses were performed. First, image”); (2) thoughts that infer an explicit attempt by the
we wanted to know whether consumers would evaluate marketer to persuade the customer about the quality and
the custom foreign brand differently than a truly domestic uniqueness of the product (e.g., “they are trying to give the
brand. Since these conditions differed only in the disclo- product a true European feel, to make their chocolate look
sure of the marketer’s true origin (domestic versus for- more authentic and of higher quality”); (3) thoughts that
eign), any difference in brand attitudes would be more infer the marketer’s attempt to persuade customers to buy
likely due to the interpretation of the marketer’s brand the product (e.g., “the company wants to attract more
position as a persuasion/manipulation attempt or as a customers to buy the chocolate”); and (4) thoughts that
positive customization effort. Romanian respondents’ infer a deceptive manipulation attempt by the marketer
attitudes toward the custom foreign brand were more (e.g., “they trick me; they make me think it is an American
favorable than their attitudes toward the true Romanian product”). The first category seems to be related to “per-
brand (t(67) = -1.8, p = 0.03, one-tailed). This difference ceived marketing effort” (Kirmani and Wright 1989),
was not significant in the U.S. (t(64) = 0.43, n.s.), although while the other three correspond to Friestad and Wright’s
respondents seemed to evaluate the true domestic brand (1994) persuasion knowledge categories represented by
slightly more favorable. It seems that consumers in Roma- beliefs about marketers’ tactics and persuasion goals.
nia, a transitioning economy, were not concerned with the Finally, some respondents provided seemingly neutral
deceptive positioning of a foreign brand as domestic. On thoughts (e.g., “the Bruges picture shows the origin of the
the contrary, in this case, they tended to evaluate the Belgian company”) which were not classified.
custom foreign brand more favorably, which is consistent
with the direction of country-of-origin effect. This phe- The results are summarized in Table 3. Overall, it
nomenon seems to be reversed for the consumers in the seems that respondents’ thoughts were consistent with the
developed country. pattern of brand evaluation, in which Romanian respon-
dents were most favorable to the foreign brand positioned
Second, we investigated respondent’s perception of as a seemingly domestic brand while the U.S. respondents
company’s sincerity in the different conditions. The pat- were least favorable to this customized foreign brand.
tern of company sincerity perceptions followed the pat- Romanians seemed more likely than Americans to inter-
tern of brand attitude ratings in Romania, but not in the pret a foreign brand’s position as seemingly domestic as
U.S. Sincerity ratings of the Belgian company using a positive customization effort (44% of the total number
custom foreign positioning (essentially, using deception) of thoughts provided versus 37%, respectively). In con-
were significantly higher than sincerity ratings of the trast, Americans seemed more likely than Romanians to
domestic company in Romania (t(68) = -2.6, p < 0.01). interpret it as a deceptive manipulation attempt (19% of
Sincerity of the Belgium company not using custom the total number of thoughts provided versus none).
positioning was still higher relative to the local Romanian Interestingly, both the U.S. and Romanian respondents
company (t(61) = -1.9, p = 0.06). One possible explana- were most likely to interpret the communication of the
tion is that Romanian subjects are chronically suspicions foreign brand as being foreign as an attempt to persuade
of everything that is domestic. The foreign company customers about the quality and uniqueness of the product
might have benefitted from a halo effect. Interestingly, (41% of Romanians’ thoughts and 58% of Americans’
there were no significant differences in the U.S. sample. thoughts). These results support our arguments that con-
sumers in a developed country differ from consumers in
Third, in order to gain an insight into respondents’ a transitioning country in terms of persuasion knowledge
thinking processes, while evaluating the brand, we ana- and perceptions of a marketer’s effort to customize a
lyzed the open-ended responses to the question about the foreign brand to appeal to the domestic market.
company’s rationale to feature the specific country land-
mark. The coding scheme was based on the theoretical CONCLUSIONS
perspectives described in Friestad and Wright’s (1994)
classification of lay persuasion beliefs. Two judges coded In summary, the results of the reported exploratory
the open-ended responses for the number of thoughts study suggest that, in a transitioning economy, consumers
expressed and the category to which each thought be- are likely to reward a foreign marketer’s effort to custom-
longed. ize the brand so that it seems domestic. On the contrary, in
a developed economy, consumers are likely to penalize a
The categories that emerged from the data coding are: marketer’s attempt to position a foreign brand as seeming-
(1) explicit inference about the company’s attempt or ly domestic, instead of communicating it as a foreign
effort to customize the brand in order to have a domestic brand. Due to lower exposure to marketing tactics and less
appeal (e.g., “the Belgian firm learned about the Roma- choice of high quality products, consumers in transition-

TABLE 3
Thoughts Concerning Company’s Intent in Brand Communication
Local Brand Custom Foreign Foreign Brand

(Seemingly Local) Brand (“as is”)
Romania U.S. Romania U.S. Romania U.S.
Sample size 39 36 39 30 37 33
Total number thoughts 33 38 36 32 29 33
Customizing effort/
domestic appeal 8(24%)* 13(34%) 16(44%) 12(37%) – –
PK: Product quality

inferences 12(36%) 11(29%) 7(19%) 3(9%) 14(41%) 19(58%)
PK: Sales/Customer
attraction inferences 1(3%) 1(3%) 9(25%) 2(6%) 5(17%) 5(15%)
PK: Manipulation/
deception inferences – – – 6(19%) – –
*Note: Percentages are computed out of the total number of thoughts.
ing markets are less likely to be suspicious of a foreign of college students, from only two universities, one in a
marketer’s persuasion attempts and more likely to appre- developed and one in a transitioning country. Further,
ciate the marketer’s effort to appeal to the domestic only one product, chocolate, was used. Future research
consumer. It seems that these consumers are likely to be should use multiple products from different product cat-
more suspicious of marketing actions undertaken by do- egories and more representative samples in order to draw
mestic producers than the actions undertaken by foreign more definite conclusions. Future research should test
producers. Evidently, this situation holds if the country- whether the effect is moderated by product category. In
of-origin effect is favorable to the foreign producer. In particular, it is plausible that effects for the “good” prod-
contrast, an international company attempting to enter a ucts, such as chocolate, might differ from the effects for
developed market with a strategy of positioning the for- the “stigmatized” products, such as cigarettes. The current
eign brand as seemingly domestic will encounter consum- study opens a program of research that intends to replicate
ers’ suspicions and lack of acceptance. In this case, the and extend reported findings. The study needs to be
company may benefit from explicitly communicating the replicated with various brand and packaging stimuli.
foreign origin of the brand as a signal for product quality. Further exploration is also needed to understand the
mechanism that brings about the effects of custom posi-
Evidently, the results of this exploratory study should tioning a foreign brand as domestic.
be interpreted with caution. The study used small samples
REFERENCES keting, 63 (1), 75–87.

Campbell, Margaret C. (1995), “When Attention-Getting
Alden, Dana L., Jan-Benedict E.M. Steenkamp, and Ra- Advertising Tactics Elicit Consumer Inferences of
jeev Batra (1999), “Brand Positioning Through Ad- Manipulative Intent: The Importance of Balancing
vertising in Asia, North America, and Europe: The Benefits and Investments,” Journal of Consumer
Role of Global Consumer Culture,” Journal of Mar- Psychology, 4 (3), 225–54.

Batra, Rajeev, Venkatram Ramaswamy, Dana L. Alden, About the Psychology of Advertising,” Journal of
Jan-Benedict E.M. Steenkamp, and S. Ramachander Consumer Research, 22, 62–74.
(2000), “Effects of Brand Local and Nonlocal Origin Kirmani, Amna and Peter Wright (1989), “Money Talks:
on Consumer Attitudes in Developing Countries,” Perceived Advertising Expense and Expected Prod-
Journal of Consumer Psychology, 9 (2), 83–95. uct Quality,” Journal of Consumer Research, 16,
Friestad, Marian and Peter Wright (1994), “The Persua- 344–53.
sion Knowledge Model: How People Cope with Maheswaran, Durairaj (1994), “Country of Origin as a
Persuasion Attempts,” Journal of Consumer Re- Stereotype: Effects of Consumer Expertise and At-
search, 21, 1–31. tribute Strength on Product Evaluations,” Journal of
____________ and ____________ (1995), “Persuasion Consumer Research, 21, 354–65.
Knowledge: Lay People’s and Researchers’ Beliefs

Lada V. Kurpis
School of Business Administration
Gonzaga University
502 E. Boone Ave.
Spokane, WA 99258–0009
Phone: 509.323.7033
E-Mail: kurpis@jepson.gonzaga.edu

CAN SELF-AFFIRMATION REDUCE PREJUDICE EXPRESSION
TOWARD STEREOTYPED BRANDS?
Huimin Xu, University of Arizona, Tucson
ABSTRACT consumption. Toward this end, the validity of the meta-

phor of brands as people is investigated. (2) Borrow
Prejudice toward brands is recognized as a distinctive relevant theories from social psychology, and conceptual-
construct. Holding prejudice against stereotyped brands ize brand prejudice expression as a means of embracing
serves to defend one’s worldview and to maintain self- certain worldview and maintaining self-esteem. This study
esteem. However, when a chance of self-affirmation is proposes that self-affirmation decreases the tendency of
provided, the need to maintain self-esteem through prej- brand prejudice. A 2x2 factorial experiment is conducted
udice expression is reduced. The implication for effective to examine the proposition. In doing so, the current study
marketing communication is discussed. takes the lead in empirically studying the antecedents of
brand prejudice. The empirical result might help assess
INTRODUCTION the effectiveness of some relationship marketing tech-
niques that provide consumers with opportunities of self-
The equation of consumption with choice and wants/ affirmation.
desire has long pervaded consumer research. This domi-
nant paradigm treats wants/desire as the only, or at least BRAND PREJUDICE
the main motive in consumption. This equation biased
various fields of consumer research. For example, it Consumer research has long been focused on acqui-
disregards the pushing force of brand loyalty, and ne- sition behavior. How consumers identify themselves with
glects the negative form of relationship with brands (e.g., certain brands has been intensively studied, while how
Fournier 1998). Consumers’ prejudice and other negative consumers deny association with brands remains under-
emotions at the brand level are vastly ignored, though explored. This biases research on consumer emotions.
there are some works on emotions in ads. watching and Reducing the multi-dimensional consumption emotions
product consumption (e.g., Richins 1997). Realizing that to a range of satisfaction, researchers frequently ignore
there has been almost no systematic empirical work on the fact that there are a variety of both positive and
consumer prejudice, the current study stresses that preju- negative emotions (Westbrook and Oliver 1991). Though
dice is another way that consumers relate to the commod- some scholars have paid attention to emotions in ads.
ity world on a daily basis. Some scholars in social psychol- watching, product consumption, and service encounter
ogy and consumer research even take a more assertive (e.g., Richins 1997), one specific category of negative
stance, suggesting that the avoidance of an undesired state consumption emotions, namely, consumer prejudice, has
is a stronger driving force of behavior than the pursuit of been largely overlooked. Nonetheless, its existence is
a desired state (Wilk 1997), and a better predictor of abundantly documented in mass media. The intellectuals
satisfaction (Ogilvie 1987). Their view suggests that dismiss rap music as vulgar. Abstract art may be accused
studying consumer prejudice is indispensable. as pretentious by the working class. “The worst dressed
celebrities” exposes and ridicules fashion mistakes. Con-
It might be due to the following two reasons that scant sumer prejudice permeates our daily conversations.
attention has been paid to consumers’ prejudice: (1)
Consumers’ prejudice come side by side with anti-con- The brand level of consumer prejudice is termed as
sumption, which usually leaves no material traces. If brand prejudice. Brands prejudice is the emotional antip-
prejudice is not expressed explicitly, there is no visibility athy toward brands. This emotional antipathy is mainly
at all. Understandably, it is more difficult to study preju- based on stereotypic perception of the symbolic meaning
dice than to study more visible consumption; (2) Out of of the brands. It is not merely involuntary visceral disgust,
managerial interest, it is more profitable to focus the though it could be accompanied by the latter. Instead, it
marketing effort on current and potential consumers, than often reflects deliberation. To give an example, the fol-
on people who show resistance against that product/ lowing is an excerpt from interviews conducted by Ban-
brand. ister and Hogg (2001).
The current study tries to fill the gap by doing the “The typical Sharon is . . . that girl, she had this tiny
following: (1) Advance the notion of consumers’ preju- little dress on and she was absolutely hammered
dice against brands, and address its important role in (English slang for drunk ) with her mate and she was

trying to chat up this group of lads . . . because she had a similar way they perceive people. Brands are perceived
obviously decided that everybody else wore the short to possess demographic characteristics such as age, gen-
strappy little dresses, and that was what she was der, and socioeconomic class (Levy 1959). Country-of-
wearing . . . she looked dreadful. . . .” Another origin renders brands a nationality. Besides demographic
example is the prejudice against Starbucks. This characteristics, brands also have human-like personali-
brand is stigmatized as yuppie haven: rushed, imper- ties. Brand personality is a well-accepted concept, which
sonal, overly analytical, and sterile. Plus, as a domi- refers to the set of human characteristics associated with
nant chain store brand, it is charged as contributing to a brand (e.g., Aaker 1997). Paralleling the “Big Five”
corporate spread and cultural homogenization (Th- dimensions of human personality, a reliable, valid, and
ompson et al. 2004). generalizable scale was developed to measure the five
dimensions and 42 personality traits of brand personality
Brand prejudice is a construct distinctive from con- (Aaker 1997). Sound analogy can be drawn between three
sumer dissatisfaction. Unlike consumer dissatisfaction, of the five dimensions and their counterparts from human
brands prejudice does not need to be preceded by direct personality scale (i.e., sincerity with agreeableness, ex-
experiences of purchase or consumption. Rather, it often citement with extroversion, competence with conscien-
comes together with anti-consumption, which refers to tiousness) (Aaker 1997). Furthermore, just like a person
the deliberate avoidance of products and services. Even in is perceived as a knot in a interpersonal relationship
the absence of direct consumption experience, one’s ob- network, a brand is perceived to be embedded in its
servation is sufficient to form stereotypes as the cognitive relationships with other brands, e.g., competing brands,
basis of prejudice. Second, brand prejudice is based alliance brands, and ingredient brands (Keller 2003). It is
mainly on a brand’s unpleasant symbolic implication, argued that the personification of brands origins from the
while dissatisfaction frequently stems from the utilitarian universal human need for animism (Gilmore 1919), which
dimension of brands. The third thing that differentiates gives rise to the tendency to imbue brands with a variety
brand prejudice from dissatisfaction is that prejudice of human traits. Both marketer and consumers contribute
holders demand no constructive change to the targeted toward this state, with the former consciously spending
brands, while dissatisfied consumer might require the efforts to enhance certain personification, and the latter
problems to be corrected, performance to be improved, or willingly accept such personification. Marketers might
the brand to be removed from the market. Why don’t employ techniques including anthropomorphization, cre-
brand prejudice holders demand constructive change? ation of user imagery, celebrity endorsement, event spon-
Anthropologists contend that the existence of prejudice sorship, and corporate image presentation. In addition,
target brands might be of positive value to prejudice human-like traits are also associated with a brand in a less
holders, by serving as “a reminder of the war against the direct way through product-related attributes, product
others and their despised way of life” (Douglas 1997). category associations, association with other brands, brand
Finally, more than what they do with dissatisfaction, name, symbol/logo, packaging, advertising style, price
people express their prejudice to seek affiliation and and distribution channel.
maintain their social identity (Tajfel and Turner 1986).
This is why brand prejudice exists abundantly and vocally Prejudice’ Role in Consumption as Another Motive
at group level (Muniz and O’ Guinn 2001; Thompson
et al. 2004), besides the private, individual level. The dominant paradigm treats wants/desire as the
only motive for consumption. However, some research
Brand prejudice is an ongoing phenomenon. Most suggests that consumer prejudice, including brand preju-
people reported that their experiences of disgust and dice, constitutes another motive. Prejudice may be even
revulsion had changed a great deal during their lifetimes more important and effective than consumption in form-
(Wilk 1997). Greater education, moving into adulthood, ing a consumer’s identity. This might be particularly true
increased wealth, and foreign travel are correlated with a in contemporary developed societies, where status sym-
greater diversity of likes, a broadening of tastes, and fewer bols are rapidly disappearing in the face of abundance. In
hate (Belk et al. 1982; Wilk 1997). These evidences Wilk’s study (1997), it was found that most people,
demonstrate consistency with findings in social psychol- regardless of class and ethnicity, seemed to like approxi-
ogy: Education, especially specific intercultural educa- mately the same foods, the same TV shows, and the same
tion, helps engender tolerance (Allport 1954) music. People rarely got excited about things they liked.
Few relationships surfaced between specific tastes and
Both stereotype and prejudice are concepts that orig- social class, income, or ethnicity. In contrast, distaste and
inated from psychology and are usually used on people. prejudice are much more powerful social predictors. This
The metaphor of brands as people needs to be established finding is consistent with Bourdieu’s observation that
to justify the construct validity of brand prejudice and any consumption choices are often constituted in opposition
further discussion on this topic. People perceive brands in to the choices of other social groups. It is through negation

of others’ choice, that one claims his/her position (Bour- which is not found in animals. This awareness, combined
dieu 1979). Being more articulate, people in Wilk’s inter- with the animal instinct for self-preservation, creates
view (1997) would often get deeply involved in discuss- terror of mortality. To buffer the fear, we immerse our-
ing the things they hated or disliked and the reasons of selves in a cultural worldview, which “imbues life with
their hates/dislikes. In his interview with Belizean high meaning, order and permanence, and the promise of
school students, it was found that only a tiny minority safety and death transcendence to those who meet the
pursued the most fashionable, while most students’ con- prescribed standards of value” (Greenberg et al. 1997). By
sumption choices were driven by anxiety over consuming living up to these cultural standards, one gains self-
the wrong things. That means, most were more concerned esteem, the belief that he/she is “an object of primary
with being out of fashion than they were with being in the value in a meaningful universe” (Greenberg et al. 1997).
latest style. It appeared that the more important daily Terror management theory has two implications: first, we
consumption skill is that of knowing what not to wear. need to have faith in a particular cultural worldview,
Literature in social psychology provides an explanation defending it when threatened; Second, we need to believe
for this phenomenon, if we think of prejudice as vocaliz- that we live up to the cultural standards. The need for self-
ing the undesired self and desire/preference reflecting the esteem drives both noble and contemptible human behav-
ideal self. It was argued that the undesired self is less iors. A point relevant here is that holding and expressing
abstract than the content of the ideal self, since it is more prejudice is fundamentally motivated by the striving for
experience based than the ideal self. Thus the undesired self-esteem. This causal link has garnered empirical sup-
self is a more stable standard against which one judges his/ port. For example, when subjects faced self-esteem threat
her present level of well-being. The distance between the in the form of receiving negative bogus feedback on their
real self and the undesired self is a better predictor of life intelligence, they expressed more prejudice toward a
satisfaction than is the distance between the real self and negatively stereotyped ethnic group, compared to sub-
the ideal self (Ogilvie 1987). Similarly, of what to con- jects receiving neutral feedback. Such derogating in re-
sume and what not to, prejudice/avoidance may be a sponse to self-image threat was of positive value to the
stronger motive than wants. It may be the more frequently prejudice holder, since it mediated increase in self-esteem
used anchor against which we gauge our consumption (Fein and Spencer 1997).
satisfaction. The desire and prejudice toward brands are
closely related with each other. Learning good tastes is Holding prejudice against out-groups helps to guard
largely a matter of learning what bad tastes are. Philoso- one’s cherished worldview (Allport 1954). The question
phers like Bataille and Derrida have argued that all desire arises as to whether brand prejudice in particular serves to
is ultimately rooted in disgust or rejection. Choices based defend one’s worldview. It has long been proposed and
on wants are always choices based on aversions and empirically supported that personal values and consump-
distaste as well. tion choices are consistent (e.g., Homer and Kahle 1988;
Pitts and Woodside 1983). By contrast, the relation be-
Why Brand Prejudice? Can Self-Affirmation Reduce tween values and anti-consumption choices has been
Prejudice Expression? escaping most consumer researchers’ attention. The cur-
rent study would argue for this statement’s validity in the
Although there are works which recognize the exist- arena of consumer behavior. Not only consumption choic-
ence of brand prejudice and its important role in consump- es, but also anti-consumption choices announce cultural
tion choice, the following questions remain understudied: allegiance. This relation has been documented through
Why do consumers develop prejudice against brands that out various historical eras. Particular kinds of food were
have not done any economic or physical harm to them? consciously renounced by medieval monks, for whom
Under which condition are people more likely to express rejecting meat and other luxuries was a repudiation of the
prejudice toward particular brands? The current study values that placed eating first among worldly values. It
tries to investigate these research questions by conceptu- was once a chic protest against U.S. involvement in the
alizing brand prejudice as a means of self-esteem mainte- Vietnam War to drive a Swedish-made Volvo, because
nance. Cognitive, socio-cultural, and motivational factors the Swedish government was vocally criticizing U.S.
all promote and maintain prejudice. The need for self- involvement at that time. The resistance to IBM and
esteem has frequently been mentioned as the motivational Microsoft’s overwhelming dominance coexists with the
factor of prejudice (Allport 1954; Tajfel and Turner ideology of democracy. Basically, brand prejudice re-
1979). Many researchers contend that there is a universal flects prejudice against the cultural worldview brands
need for self-esteem (Greenberg et al. 1997, 1999; Heine represent. Prejudice expression serves two purposes in the
et al. 1999). Terror management theory proposes that process of boosting self-esteem. First, by belittling the
humans use self-esteem to bring the fear of mortality ideology underlying an out-group’s consumption, preju-
under control (Greenberg et al. 1997). Terror of mortality dice expression affirms the validity of our own cultural
is engendered by awareness of the inevitability of death. worldview, or at least one aspect of that worldview.
Humans’ cognitive capacity gives rise to this awareness Second, by contrasting ourselves with the out-group along

the scale defined by our cultural criterion, we strengthens pared to a brand stereotype that does not reflect
the belief that we fulfill the cultural standard well enough, worldview incongruence.
since apparently there are some people worse off than
ourselves. H2 For incongruent brand stereotype, when individuals
are given an opportunity to self-affirm, they are less
Fortunately, prejudice is not the exclusive means to likely to express prejudice toward the stereotyped
promote self-esteem. An individual may also increase brands, compared to individuals who are not given
self-esteem by enhancing the belief that he/she lives up to such an opportunity.
the cultural standards. The stronger the belief that he/she
lives up to the cultural standards, the smaller the need to When the focal brand is not incongruently stereo-
defend his/her cultural worldview by prejudice expres- typed, this study does not expect prejudice to exist.
sion. In earlier social psychology research, this substitut- Whether the self-affirmation is available or not should not
ability is found between non-prejudiced self-affirmation affect the likelihood of prejudice expression.
and prejudice (Fein and Spencer 1997). Self-affirmation
is the tendency that people avoid to confront a threat, H3 When a brand stereotype does not involve worldview
instead, they affirm some other important aspect of the incongruence, lack of self-affirmation doesn’t in-
self to reinforce one’s overall self-adequacy (Steele et al. crease the tendency of prejudice expression.
1993). The substitutability exists because they both serve
a common, higher-order goal: Self-esteem. Self-affirma- METHODOLOGY AND INITIAL EMPIRICAL
tion is a relatively general substitute for many self-esteem SUPPORT
processes. It means self-affirmation can turn other means
of self-esteem maintenance on and off, while other pro- The pretest was a 2x2 factorial between-subjects
cesses are able to turn self-affirmation on and off. experiment and was of preliminary nature. Seventy-eight
undergraduate students participated. The IVs are avail-
In Fein and Spencer’s study, Ss were either given an ability of self-affirmation and availability of brand stereo-
opportunity for non-prejudiced self-affirmation or not. type incongruence. Manipulation of self-affirmation is
All Ss were asked to rank order 11 personal characteris- borrowed from psychology, as described earlier. The
tics/values according to their personal importance. Ss in other IV has two levels: whether the brand stereotype is
the self-affirmation conditions were then asked to write perceived to be incongruent with the specific sample’s
down the nunber one personally important value, and worldview, or there is no such perceived incongruence.
describe a time in his/her life when it has been important. Since two real wine cooler brands were used as stimuli,
Research has shown that causing Ss to think about a brand familiarity was measured as a covariate. Conven-
personally important value is an effective means of pro- tional attitude semantic differentials, combined with six
ducing self-affirmation (Steele 1999). Ss in the no self- newly written items that purportedly captured “perceived
affirmation conditions were asked to indicate the number value incongruence” and “refusal of association,” were
nine personally important value, and write about a time used as the measure of brand prejudice. ANCOVA shows
when it may be important to a typical student. The second significant interaction between the two independent vari-
independent variable was the target’s ethnicity. Targets ables (p = 0.019). For Boone’s Farm (a brand supposedly
were from either a negatively stereotyped ethnic group or involving stereotype incongruence in the subjects’ per-
not. It was found that Ss given the opportunity to self- spectives), self-affirmation decreases brand prejudice
affirm expressed less prejudice toward members from the expression (p = 0.064). Therefore, hypothesis 2 is mar-
negatively stereotyped ethnic group, compared to Ss who ginally supported. For Seagram’s (a brand supposedly
were not given the self-affirmation opportunity. This involving no stereotype incongruence), there is no signif-
finding has potential managerial implication in market- icant difference across the two levels of affirmation avail-
ing. If self-affirmation can substitute for prejudice expres- ability (p = 0.124). Hypothesis 3 is thus supported. H1 is
sion, providing a source of self-affirmation can reduce the not supported. The author is fully aware of the limitation
holder’s prejudice and lessen the negative influence on of the pretest design. Since real brands were used, alterna-
others through interpersonal interactions. tive explanations can not be ruled out. More rigorous
design should be used in the future.
To sum up, this study is interested in replicating this
relation when the prejudice targets are stereotyped brands. FUTURE DIRECTION: THE POWER OF
Specifically, this study investigates the following hypoth- PREJUDICE
eses:
Once we conceptualize attitudes with certain charac-
H1 When a brand stereotype reflects worldview incon- teristics as prejudice, a new area is open up for study.
gruence, more brand prejudice would occur, com- Singling out certain attitudes as prejudice has proved to be

profitable in social psychology. Prejudice definitely adds to consumers and help them maintain a positive self-
beyond the more aged construct of attitude. For example, regard. On the other hand, communications utilizing fear
prejudice predicts discrimination far better than heavily appeal or making upward social comparison salient threaten
cognitive attitude (Dovidio et al. 1996); It is held more consumers’ self-esteem. Therefore, such communications
firmly (Abelson et al. 1982), and is more effectively land on the lower end of the self-affirmativeness contin-
altered by affective rather than cognitive persuasion (Ed- uum. Communication high in self-affirmation, via preju-
wards and von Hippel 1995). It would be interesting to dice reduction, enhances consumers’ trust and contributes
find out whether these findings could be replicated in the to a better buyer-seller relationship.
arena of marketing, after we conceptualize certain brand
attitudes as brand prejudice. Service marketing managers can reduce prejudiced
expression against their brands, utilizing the abundance of
Based on shared stereotype, brand prejudice is also interpersonal communication. The concrete implementa-
culturally shared. When stereotype and prejudice toward tion is: service personnel employ face-to-face communi-
brands are recognized as socio-cultural phenomenon, we cation to affirm consumers’ self-esteem. Provision of self-
can go on to examine the societal factors that lead to brand affirmation is not restricted to one-on-one service en-
prejudice, instead of being confined to the paradigm of counters. It can also be delivered through mass media. For
individual information processing. In such inquiry, vari- example, advertisement of both services and commodities
ous social psychology theories can be utilized. This would can convey pleasing messages such as “you are worth it”
give rise to a new scope of questions which have not been or “you can do it.” Sometimes, provision of self-affirma-
fully explored before. For example, does excessive homo- tion goes farther than pleasant verbal or nonverbal com-
geneity in demographics and psychographics lead to munications. Helping some consumers to achieve major
greater brand prejudice, according to social identity the- life goals or resolve troubling issues, and hopefully boost-
ory (Tajfel and Turner 1986)? Can we predict the content ing their self-esteem subsequently, has long been used by
of brand prejudice once we know the associated user business. Walmart offers school supplies to needy chil-
group’s niche in the social structure? Particularly, does dren. Beverly Hill stylists make over consumers into more
higher social class unavoidably hold paternalistic preju- glamorous people, who are finally happy with them-
dice toward brands associated with lower class as a means selves. Negatively stereotyped brands and professions
of system justification (Glick and Fiske 2001)? Do dom- have improved their images by improving consumers’
inant corporate brands in different societies possess sim- self-images first. In such cases, consumers who witness
ilar stereotypes and prejudice held against them, accord- self-affirmation provision, though they do not directly
ing to the transcultural principle of stereotyping (Stephan experience it, might feel self-affirmed too. At present, this
and Rosenfield 1982)? Are brands associated with con- is only a speculation. Whether vicarious self-affirmation
flicting user groups derogatorily viewed by each other, in could actually occur constitutes a topic for future investi-
line with the realistic conflict view of prejudice (Taylor gation. Another subtle issue that awaits future research is
and Maghaddam 1987)? To embark on these intriguing the distinction between prejudice and prejudice expres-
issues more effectively, researchers need to recognize the sion. As a value, prejudice is stable and resistant to
distinctive contribution of the construct of brand preju- attempted influence. By comparison, expression of prej-
dice. udice is more variable, since it is subject to the salience of
prejudice content and social norm. Thus the question
DISCUSSION arises as to self-affirmation’s impact on these two differ-
ent constructs. Does providing self-affirmation really
Provision of self-affirmation, via prejudice reduc- reduce prejudice, or does it merely temporarily hold back
tion, eventually strengthens buyer-seller relationship. the expression of it? Plus, how long will the curbing effect
Buyer-seller relationships are a focal concern of market- last? These questions remain for exploration later on.
ing as a discipline (Day and Montgomery 1999). Commu-
nication between the two parties, among others, is seen as Although brand managers do not always worry about
one antecedent to good buyer-seller relationship (e.g., reducing prejudice toward their brands, this does become
Nonaka 1994). Particularly, communication increases a concern when a brand attempts to penetrate into a new
trust and cooperation (Morgan and Hunt 1994). However, segment which holds prejudice against the brand. The
this line of research has been largely confined to commu- current study, though preliminary in nature, points out
nication frequency as the independent variable. The spe- self-affirmation provision as one way to bring down brand
cific characteristics of communication content, among prejudice. Besides its obvious managerial benefit, self-
which being self-affirmative is one, have been ignored. affirmation provision caters to the universal need for self-
This study contends that communication could be of esteem, delivering genuine value to consumers. There-
variable magnitude of self-affirming from a consumer’s fore, this research quest fits in the recent trend of stressing
angle. Highly self-affirming communications pay respect communication, value and customer relationship, as re-

flected in the American Marketing Association 2004 experimental design, would be worthwhile to pursue this
definition of marketing. More research, with rigorous research stream.
REFERENCES ski (1997), “Terror Management Theory of Self-

Esteem and Cultural Worldviews: Empirical Assess-
Aaker, Jennifer (1997), “Dimensions of Brand Personal- ments and Conceptual Refinements,” Advances in
ity,” Journal of Marketing Research, 34 (August), Experimental Social Psychology, 29, 61–139.
347–56. ____________, ____________, and ____________
Abelson, Robert. P., Donald R. Kinder, and Mark D. (1999), “Why Do People Need Self-Esteem? Con-
Peters (1982), “Affective and Semantic Components verging Evidence That Self-Esteem Serves an Anx-
in Political Person Perception,” Journal of Personal- iety-Buffering Function,” in The Self in Social Psy-
ity and Social Psychology, 42 (4), 619–30. chology, Roy F. Baumeister, ed. Philadelphia, PA:
Allport, G.W. (1954), The Nature of Prejudice. Cam- Psychology Press, 105–22.
bridge, MA: Addison-Wesley. Heine, Steven J., Darrin R. Lehman, Hazel Rose Markus,
Banister, Emma N. and Margaret K. Hogg (2001), “Map- and Shinobu Kitayama (1999), “Is There a Universal
ping the Negative Self: From ‘So Not Me’. . . To ‘Just Need for Positive Self-Regard?” Psychological Re-
Not Me’,” Advances in Consumer Research, 28, view, 106 (4), 766–94.
242–48. Homer, Pamela M. and Lynn R. Kahle (1988), “A Struc-
Bourdieu, Pierre (1979), “The Aesthetic Sense as the tural Equation Test of the Values-Attitude-Behavior
Sense of Distinction,” in The Consumer Society Read- Hierarchy,” Journal of Personality and Social Psy-
er, Juliet B. Schor and Douglas B. Holt, eds. New chology, 54 (4), 638–46.
York: The New Press. Keller, Kevin (2003), “Brand Synthesis: The Multidi-
Day, George S. and David B. Montgomery (1999), “Chart- mensionality of Brand Knowledge,” Journal of Con-
ing New Directions for Marketing,” Journal of Mar- sumer Research, 29 (March), 595–600.
keting, 63 (special issue), 3–13. Levy, Sidney (1959), “Symbols for Sales,” Harvard Busi-
Douglas, Mary (1997), “In Defense of Shopping,” in The ness Review, 37 (4), 117–24.
Shopping Experience, Pasi Falk and Colin Campbell, Morgan, Robert M. and Shelby D. Hunt (1994), “The
eds. Sage Publications, 15–30. Commitment-Trust Theory of Relationship Market-
Dovidio, John F., John C. Brigham, Blair T. Johnson, and ing,” Journal of Marketing, 58 (July), 20–38.
Samuel L. Gaertner (1996), “Stereotyping, Preju- Muniz, Albert, and Thomas O’ Guinn (2001), “Brand
dice, and Discrimination: Another Look,” in Stereo- Community,” Journal of Consumer Research, 27
types and Stereotyping, Neil Macrae, Charles Stang- (March), 412–32.
or, and Miles Hewstone, eds. New York: Guilford, Nonaka, Ikujiro (1994), “A Dynamic Theory of Organi-
276–319. zational Knowledge Creation,” Organization Sci-
Edwards, Kari and Williams von Hippel (1995), “Hearts ence, 5 (February), 14–37.
and Minds: The Priority of Affective Versus Cogni- Ogilvie, Daniel (1987), “The Undesired Self: A Neglect-
tive Factors in Person Perception,” Personality and ed Variable in Personality Research,” Journal of
Social Psychology Bulletin, 21 (10), 996–1011. Personality and Social Psychology, 52 (2), 379–85.
Fein, Steven and Steven Spencer (1997), “Prejudice As Pitts, Robert E. and Arch G. Woodside (1983), “Personal
Self-Image Maintenance: Affirming The Self Through Value Influence on Consumer Product Class and
Derogating Others,” Journal of Personality and So- Brand Preferences,” The Journal of Social Psychol-
cial Psychology, 73 (1), 31–44. ogy, 119 (1), 37–53.
Fournier, Susan (1998), “Consumers and Their Brands: Richins, Marsha L. (1997), “Measuring Emotions in the
Developing Relationship Theory in Consumer Re- Consumption Experience,” Journal of Consumer
search,” Journal of Consumer Research, 24 (March), Research, 24 (September), 127–46.
343–73. Steele, Claude M., Steven J. Spencer, and Michael Lynch
Gilmore, George (1919), Animism. Boston: Marshall Jones (1993), “Self-Image Resilience and Dissonance: The
Company. Role of Affirmational Resources,” Journal of Per-
Glick, Peter and Susan T. Fiske (1996), “The Ambivalent sonality and Social Psychology, 64 (6), 885–96.
Sexism Inventory: Differentiating Hostile and Be- ____________ (1999), “The Psychology of Self-Affir-
nevolent Sexism,” Journal of Personality and Social mation: Sustaining the Integrity of the Self,” in The
Psychology, 70 (3), 491–512. Self in Social Psychology, Roy F. Baumeister, ed.
Greenberg, Jeff, Sheldon Solomon, and Tome Pyszczyn- Philadelphia, PA: Psychology Press, 372–90.

Stephan, W.G. and D. Rosenfield (1982), “Racial and “Realistic Group Conflict Theory,” in Theories of
Ethnic Stereotypes,” in In the Eye of the Beholder: Intergroup Behavior, New York: Praeger, 33–57.
Contemporary Issues in Stereotyping, A.G. Miller, Thompson, Craig J., Aric Rindfleisch, and Zeynep Arsel
ed. New York: Praeger, 92–136. (2004), “The Doppelganger Brand Image and Con-
Tajfel, Henri and John Turner (1979), “An Integrative sumers’ Motivations for Brand Avoidance: The Case
Theory of Intergroup Conflict,” in Intergroup Rela- of Starbucks,” Working Paper.
tions: Essential Readings, Michael A. Hogg and Westbrook, Robert and Oliver Richard (1991), “The
Dominic Abrams, eds. New York: Psychology Press Dimensionality of Consumption Emotion Patterns
2001, 94–109. and Consumer Satisfaction,” Journal of Consumer
____________ and John Turner (1986), “The Social Research, 18 (June), 84–91.
Identity Theory of Intergroup Behavior,” in The Wilk, Richard R. (1997), “A Critique of Desire: Distaste
Psychology of Intergroup Relations, S. Worchel and and Dislike in Consumer Behavior,” Consumption,
W.G. Austin, eds. Chicago: Nelson-Hall. Markets and Culture, 1 (2), 175–96.
Taylor, David M. and Fathali M. Maghaddam (1987),

Huimin Xu
University of Arizona
1130 E. Helen St.
Tucson, AZ 85721
Phone: 520.621.7485
FAX: 520.621.7483
E-Mail: huiminx@bpa.arizona.edu

SELF AND BRAND IMAGE CONGRUENCE: DRIVING
CONSUMER VALUE
Adam Marquardt, University of Oregon, Eugene
SUMMARY along a given set of specific image traits and characteris-

tics, and the individual’s perceived categorization of a
As firms attempt to create value for consumers, many product along the same dimensions increases in congru-
rely on brand image as an effective tool to reach and ence, that consumer value will also increase.
penetrate their target markets (Park, Jaworski, and MacIn-
nis 1986). This study explores how consumers relate to When areas of customer-desired value are success-
their chosen brands, specifically investigating the degree fully identified, integrated and communicated, consumers
to which a consumer’s actual and/or ideal self-image is will be more likely to buy the brand, often paying a
congruent with the images of his or her preferred brands, premium in the process (Romaniuk 2001; Keller 1993).
and how this impacts the consumer’s level of perceived Keller (1998) posits the importance of congruence be-
value. This research combines qualitative interviews with tween user imagery and brand personality in building
exploratory empirical studies to examine this relationship brand image, and suggests that this congruence is partic-
from the consumer’s viewpoint. Preliminary findings ularly “relevant” regarding the more extrinsic benefits
support the role of congruence between the individual’s associated with symbolic brands. Brand personalities and
self-image and the brand image, predicting that as congru- images appealing to a consumer’s actual or ideal self-
ence increases, consumer value (customer-based brand image can help to create consumer value, and thus a
equity) will also increase. deeper level of rapport (Fournier 1998; Aaker 1997). For
illustrative purposes, inner-city basketball players pro-
A strong brand presents consumers with consistent vide a good demonstration of how this process operates.
and positive messages, differentiating it from competitive For many “streetballers” their chosen brands are not only
offerings and assisting in the consumer decision-making functional, but also symbolic. Some of these basketball
process (McDonald, de Chernatony, and Harris 2001; players select athletic apparel and footwear based on their
Blankson and Caliphates 1999). Firms can create favor- actual level of performance, reputation and status, choos-
able positions in the minds of their customers by presenting brands that provide a reflection of their actual self.
ing attractive brand images. Park, Jaworski, and MacInnis Other less accomplished and respected players may select
(1986) argue that firms that are able to introduce, elabo- the same merchandise based on their desire to project an
rate, and fortify the brand’s concept (brand image) will be image or attitude consistent with their ideal self-image
better able to attract and retain customers, creating stron- (projection of who they want to be). Still others select their
ger and more profitable brands. In doing such, brands are apparel and footwear utilizing a combination of their
able to generate value for the consumer (customer-based actual and ideal self-images.
brand equity), and comparative advantages for the firm
(McAlexander, Schouten, and Koenig 2002; Keller 1993). This study utilized a two-part methodological ap-
proach. The first approach involved conducting explor-
The consumer’s self-image plays a critical role in atory interviews with subjects from differing demograph-
perceiving brand image, and can be described in terms of ic and geographic environments. The purpose of these
the imputed brand personality (Aaker and Biel 1993). The interviews was to determine if a brand’s image was
consumer’s self-image refers to an individual’s ideas and important to these consumers and if it played a role in their
feelings about one’s self (Underwood, Bond, and Baer decision-making processes. It was determined that many
2001), and the basic premise held within this study is that consumers value a brand’s image, and in fact make the
consumers’ possessions contribute to and reflect their brand image an important factor within their decision-
identities. This study explores the individual’s actual and making process. The second methodological approach
ideal self-images, where the individual’s actual self-im- involved an exploratory online survey based on informa-
age refers to how respondents relate their current self- tion extracted from Study 1. The survey provides evi-
image along predetermined self-concept scale measures, dence that consumers are inclined to place higher levels of
while the ideal self-image refers to how respondents relate value on brands that exhibit the image traits that they have,
their image perceptions of an ideal person along the same or desire to have. Perhaps more importantly, the survey
scale dimensions (utilizing projection theory to derive indicates that brand image plays an important role in
characteristics consistent with the individual’s ideal self- consumers’ decision-making processes and in their pur-
image). This study proposes that as the degree of per- chasing intentions.
ceived fit between an individual’s self-categorization

The definitive goal of a brand is to signal added value image (projection of an ideal self), or some combination
to the consumer such that they look to purchase and/or of the two has broad implications for branding, marketing
repurchase the product, thus developing a competitive and strategic management. This is particularly true if the
advantage for the provider. Determining the degree to degree to which self-image and brand image align truly
which consumers place value on brands that embody their does create superior consumer value (customer-based
actual self-image (reflection of self), their ideal self- brand equity). References available upon request.

Adam Marquardt
Charles H. Lundquist College of Business
205B Peterson Hall
University of Oregon, Eugene
1208 University of Oregon
Eugene, OR 97403
Phone: 541.346.1438
FAX: 541.346.3341
E-Mail: amarquar@uoregon.edu

THE SPILLOVER EFFECTS OF PRODUCT-HARM CRISES IN
A BRAND PORTFOLIO
Jing Lei, Universiteit Maastricht, The Netherlands
Niraj Dawar, University of Western Ontario, Ontario
Jos Lemmink, Universiteit Maastricht, The Netherlands
SUMMARY over to other brands in a portfolio. This is primarily due to

the lack of understanding about how brand portfolio
In the current marketplace, complex brand portfolios information is cognitively organized and represented in
with multiple brands and sub-brands are often developed consumers memory. The cognitive representation of brand
by companies to fulfill market segmentation and brand information is of central importance to the understanding
strategy objectives (Aaker and Joachimsthaler 2000b). of consumer behaviour and marketing. As evidenced in
For instance, Laforet and Saunders (1994) state that, previous literature, many different conceptualizations have
nowadays, the conventional single-brand company is rare been used to study this construct, including nodes in an
and 95 percent of companies have more than one brand for associative network (Alba and Chattopadhyay 1985, 1986;
their products. Our starting point is that brands in a Nudungadi 1990), semantic categories (Dawar 1996;
portfolio are not cognitively independent in the way that Park, Milberg, and Lawson 1991), schemas (Sujan and
consumers store them in memory. Rather, due to a variety Bettman 1989), and bundles of associations (Keller 1993;
of reasons, including common corporate logos, similar Krishnan 1996; Lawson 2002). Common to these concep-
trade dress and design, similar or related advertising, tualizations is that they place the representation of indi-
complementary usage, and even common shelf location, vidual brands in the context of a larger cognitive structure
consumers may establish associations and linkages be- (e.g., a category), and attempt to understand how such
tween brands in a portfolio to help them navigate the representation influences the processing of new informa-
randscape in which they consume. tion. However, despite examining brands as part of larger
structures, past research has largely neglected the interde-
Once these linkages are established in consumers pendency between brands. In particular, there has been
memory, a brand cannot be isolated from exposure to little research on networks of brands where brands are
information about related brands in the same brand port- associated with each other (as opposed to cognitive net-
folio. This phenomenon, termed pillover effect (e.g., works that link brands to associations, products, and
Balachander and Ghose 2003), refers to the extent to product categories). As a spillover effect is expected to
which external information (e.g., a product-harm crisis) occur only if these brands are in someway interrelated in
about a brand changes attitudes about other brands that are consumers cognitive memory, the unawareness of possi-
not directly involved. For instance, if Kellogg Cornflakes ble linkages between brands in a portfolio might have
suffer severe quality problems, consumers might update hidden the spillover effect from attention in previous
their attitudes not just about Cornflakes, but also about studies.
SpecialK and other Kellogg brands. Therefore, compa-
nies with a brand portfolio need to look beyond the impact Therefore, the first objective of this paper is to ex-
of external information on the directly involved brands to plore the cognitive representation of brand portfolios in
other affiliated brands. In some cases, spillover effects consumers memory. Second, we expect to use this knowl-
may even explain more variance in attitude change on a edge to better understand the spillover effect of external
brand than external information that directly addresses information, especially the impact of product-harm cri-
this brand. For instance, in the previous example about ses, in brand portfolios. Based on the network association
negative information at Cornflakes, the spillover effect on theory, we conceptualize the representation of brand
SpecialK might overwhelm the effect of an advertisement portfolio information in consumers memory as an asso-
on SpecialK in altering consumers attitude towards Spe- ciative network, where a parent brand, sub-brands, and
cialK. Hence, neglecting spillover effects provides a linkages between them form the basic elements. We
limited view of the antecedents of brand equity (of which examine the structure of this network and the consequent
brand attitudes are one component) and may lead to spillover effect by introducing negative information at
inappropriate managerial actions. one brand node and studying the differential effects on
related nodes. We suggest that the primary characteristic
Yet to our knowledge no research has yet addressed of a brand network, the strength of association between
the impact of external information on a brand portfolio, brands, determines the magnitude of spillover effects.
and critically, if and how the impact on one brand spills Three consecutive experiments are developed to test these

hypotheses. The first experiment measures the strength of the parent brand and from one sub-brand to other sub-
association between brands in a portfolio, and the second brands. The results of Experiment 2 indicate that the
experiment manipulates the strength of association and increased strength of association between a sub-brand and
validates the findings of Experiment 1 in a different brand the parent brand leads to greater spillover effects from the
portfolio. On the basis of the findings from the first two sub-brand to the parent brand. Furthermore, the increased
experiments, we test the differences between the pro- strength of association between the sub-brands and the
posed brand network model and the existing categoriza- parent brand also leads to a greater spillover effect be-
tion model in explaining the spillover effect in Experi- tween sub-brands. The results of Experiment 3 show that,
ment 3. in the same product category, brands in the same portfolio
will have a stronger influence on each other than brands
In the results of Experiment 1, the evidence of spill- across brand portfolios. This finding suggests that the
over effects supports our proposition that there are asso- proposed network model is a stronger explanation for the
ciative linkages between brands in a portfolio forming a spillover effect in a brand portfolio than the categorization
brand network. The results also suggest that strength of model, even if all sub-brands in the portfolio are in the
association is a significant predictor of the spillover ef- same product category. References available upon re-
fects of a product-harm crisis, both from the sub-brand to quest.

Jing Lei
Universiteit Maastricht
P.O. Box 616
NL–6200MD Maastricht
The Netherlands
Phone: +31.43.3884945
FAX: +31.43.3884918
E-Mail: Jill.Lei@mw.unimaas.nl

MARKET DRIVEN INTANGIBLES AND SUSTAINABLE
PERFORMANCE ADVANTAGES
Matti Tuominen, Helsinki School of Economics (HSE), Finland
Sheelagh Matear, University of Otago, New Zealand
Sami Kajalo, Helsinki School of Economics (HSE), Finland
Saara Hyvönen, University of Helsinki, Finland
Arto Rajala, Helsinki School of Economics (HSE), Finland
Kristian Möller, Helsinki School of Economics (HSE), Finland
Gordon E. Greenley, Aston University, United Kingdom
Graham J. Hooley, Aston University, United Kingdom
SUMMARY be examined in more depth (Weerawardena and O’Cass

2004). This paper takes this argument and expands the
In recent years there has been considerable debate on framework of Day (1994). Our conceptual model is ground-
the role of marketing in firm strategy and competitive ed on the theoretical foundation of the RBT. We argue that
performance. Several scholars have argued that one way superior performance is initiated by the degree of market
for marketing to make a more significant contribution to orientation and market driven capabilities developed.
the development of a managerially oriented theory of the Performance advantage is defined in this study as the
firm is by better articulating the constructs of market achievement of overall profit levels, profit margins, and
orientation, marketing capability, innovativeness and com- ROI, emphasizing especially financial outcomes, while
petitive superiority, and their interrelationships as well sustainable edge is specified here as a durable and defen-
(for an in-depth discussion see e.g., Stoelhorst and van sible competitive posture a firm possesses. We forward
Raaij 2004). From marketing point of view, the emerging the premise that market driven intangibles facilitate firm
resource based theory (RBT) of the firm declares the competitive performance, and suggest that the link ap-
importance of intangible resources – in particular capabil- pears to be even stronger in small open economies, where
ities and relationships – in explaining firm competitive the highly “marketized“ environmental setting – regard-
superiority. However, there seems to be a lack of under- less of its location in the globe – is determined by an
standing when certain types of marketing intangibles are increased degree of customer sophistication and technol-
most critical, even though the marketing literature has for ogy proliferation, just, like in the case of New Zealand and
long argued that by leveraging these resource endow- Finland. The rationale behind our argument is grounded
ments firms could be in a stronger position to succeed in on the “rules of the game in different institutional set-
the marketplace (e.g., Day 1994). tings.”
This paper adopts a capabilities view and introduces The study involved three main phases. First, in-depth
a conceptual model consisting of three types of market interviews were conducted with senior marketing execu-
driven intangibles, namely, outside-in capability, inside- tives in 24 organizations. Thereafter, a questionnaire was
out capability, and spanning capability grounded on the developed and piloted. Finally, a representative-mailed
theoretical foundation supplied by Day (1994), and ar- survey was undertaken. Our survey was carried out in
gues that these firm intangibles contribute differently to 2001–2002 (coordinated by Aston Business School, U.K.).
sustainable competitive edge and superior performance Following the SIC (Dun and Bradstreet), our sample
advantages. Our argument, thus, follows the path depen- covered small (20–99 employees), medium (100–499
dence assumptions of the RBT traced back to the work of employees), and large (500 or more employees) firms
Penrose (1959). Market driven intangibles can be defined representing business products, consumer products, busi-
as any attribute, intellectual or relational, that can be ness services, and consumer services. The sampling frame
deployed advantageously in the marketplace. Following was supplied by national research institutes in both coun-
this conceptualization, we distinguish between two key tries involved. In total, 799 usable responses were re-
marketing intangibles: market orientation with a market ceived: 472 in New Zealand and 327 in Finland, and a
driven emphasis (Jaworski et al. 2000) and market driven response rate over 20 percent. No significant differences
capabilities (Day 1994). It has been argued, that as part of in means were found between early and late respondents
enhancing our understanding of market driven intangi- on the scales studied (t-tests at .05 level), indicating that
bles and performance, the relationship among innovative- non-response bias is unlikely to be a problem. Similar
ness, market orientation, and marketing capabilities should studies are underway in other countries (e.g. the U.K.,

Ireland, Australia, Austria, Germany, Greece, Hungary, recommended levels, respectively of: 0.60 or greater for
The Netherlands, Poland, Slovenia, China, Hong Kong, the former, while 0.50 or higher for the latter, providing a
and Brazil) and at various stages of completion to allow set of reliable and valid metrics for the constructs in-
the international robustness of the scales to be gauged. volved. Our hypotheses were tested simultaneously with
LISREL 8.51 (Jöreskog and Sörbom 1996). Modeling
Market orientation was captured deploying 14 items was undertaken deploying covariance matrix and the
from the scale first reported by Narver and Slater (1990). maximum likelihood estimation procedure. The fit index-
The measures of market driven intangibles, sustainable es for the structural model (χ2 =0.96; df = 3; p = 0.811;
edge, and performance advantage were developed for the RMSEA = 0.000; CFI = 1.000; NNFI = 1.022; CFI =
research questions at hand. In this respect, the proxies 1.000) indicate that the model fit is very good. To summa-
comprised new items and were initially developed by rize, we found generally good empirical evidence for the
identifying and creating questions on the basis of the hypotheses stated and all the relationships analyzed were
literature review, expert opinions, and field-based inter- significant.
views. Following the analysis of the pilot data, the seminal
questionnaire was further refined. Subsequently, the final The study makes a number of academic and manage-
questionnaire was developed deploying 20 capability rial contributions. Although the literature on the role of
items generated through the above, hypothesized as three market driven intangibles in competitive superiority has
separate factors following the three proposed by Day grown in significance over the last decade, few efforts
(1994). Besides, the questionnaire consisted of 19 advan- have been made to operationalize the key constructs and
tages items, hypothesized as two separate factors based on synthesize the arguments in a single rationale or frame of
the taxonomy of Day and Wensley (1988): one set for reference. This study contributes to the capabilities view
positional advantages and one set for performance advan- on competition by developing metrics for the two con-
tages. structs of market driven intangibles, namely, outside-in
capability and inside-out capability, and, thereafter, eval-
Confirmatory factor analysis (CFA) was deployed uating them in international context. In a nutshell, the
for scale construction and validation. Overall, the fit structural model demonstrates the importance of market
indexes for the measurement model (χ2 = 788.40; df = 174; driven intangibles in contributing to the explanation of
p = 0.000; RMSEA = 0.067; CFI = 0.914; NNFI = 0.927; superiority in firm competitive performance. By elaborat-
CFI = 0.940) indicate that the scale structures fit the data ing the competitive nature of market driven intangibles
acceptably and the developed proxies perform well in the this approach may help business executives better under-
context concerned. Composite reliability values (ρc) and stand the resource endowments a firm possesses, and
values of average variance extracted (ρv) were calculated thereby align competitive moves within the market pos-
following the general instructions, and all exceed the ture adopted. References available upon request.

Matti Tuominen
Helsinki School of Economics (HSE)
POB 1210 (Lapuankatu 6)
FIN-00101 Helsinki
Finland
Phone: +358.9.43138735
FAX: +358.9.43138660
E-Mail: matti.tuominen@hkkk.fi

UNDERSTANDING CREATIVE CAMPAIGN IMPLEMENTATION:
AN INVESTIGATION OF ITS ANTECEDENTS
Atlanta L. Stoyle, University of Queensland, Australia
SUMMARY and perceptions of campaign success might be conceptu-

alized as stages of an implementation process (Noble and
The process of strategy development and implemen- Mokwa 1999). Campaign implementation is clearly an
tation within the context of buyer–supplier relationships important theoretical and managerial construct and fur-
is an important yet under-researched area of marketing ther research on its antecedents and outcomes is warrant-
theory. Much of the strategy literature examines imple- ed.
mentation from a macro perspective, for example, exam-
ining the effects of market orientation, positional advan- The potential antecedents of commitment to cam-
tages, and environmental factors on strategy implementa- paign implementation and success are drawn from the
tion and firm performance (Im and Workman 2004). relationship marketing, fit, and innovation literatures.
Absent from these theories, however, is a detailed consid- Trust is conceptualized as the confidence that an adver-
eration of implementation issues from a more micro tiser places in the reliability and integrity of an advertising
perspective. While several recent studies examine micro agency (Ganesan 1994; Morgan and Hunt 1994). Rela-
characteristics, such as organizational commitment and tionship commitment is defined as the degree to which an
“buy-in” to a proposed strategy, their focus is on strategy advertiser perceives a relationship with an advertising
making within the context of organizational settings agency as valuable and worth working on to ensure that it
(Menon et al. 1999; Noble and Mowka 1999). Few studies continues (Garbarino and Johnson 1999; Moorman et al.
consider strategy commitment and implementation within 1993). Trust and commitment are drawn from the rela-
the context of buyer–supplier relationships. Both inter- tionship marketing literature and are key characteristics of
organizational factors (e.g., relationship commitment and effective relationships. The fit literature emphasizes the
trust) and strategy-specific factors (e.g., strategy fit) are importance of goal congruence or “fit” as an explanatory
important when strategy implementation is viewed from variable (Chatman 1991; Netemeyer et al. 1997). Cam-
this perspective. The specific context in which this re- paign fit is the degree to which a creative campaign
search is set is advertiser–advertising agency relation- proposed by an agency matches with the goals, objectives,
ships. Drawing on several diverse literatures, a conceptual and vision of an advertiser (Noble and Mokwa 1999).
framework of the potential antecedents and outcomes of Campaign fit exists when the themes and objectives of a
commitment to campaign implementation is advanced. proposed creative campaign are in alignment with the
strategic direction and goals of a client. Campaign charac-
Strategy implementation is posited as a key driver of teristics include creativeness and radicalness (cf., Amabile
marketplace and financial performance (Bonoma 1985; 1983; Im and Workman 2004). Campaign creativeness is
Cespedes 1991; Walker and Reukert 1987). The contin- conceptualized as the level of novelty and newness of a
ued success of firms depends not just on recognizing campaign; whereas a radical campaign represents a “break
strategic alternatives (e.g., positional advantages), but on from the past.” Both concepts may be important in the
committing to strategy implementation. Creative cam- process of campaign implementation.
paign implementation is operationalized as an advertis-
ers’ commitment to a campaign developed and advanced A number of theoretical and practical implications
by an agency specifically for an advertiser and an adver- can be drawn from the conceptual framework, including
tisers’ perceptions of the success of a campaign (Menon directions for further research. Perhaps the main theoret-
et al. 1999; Noble and Mokwa 1999). An advertiser must ical contribution is to cast the strategy implementation
commit itself, as well as necessary resources such as time issue within the context of buyer–supplier relationships.
and money, to campaign implementation in order to From this perspective, relational factors such as trust and
ensure maximum effectiveness (Menon et al. 1999). Thus, commitment, which are drawn from the relationship mar-
campaign commitment is a fundamental prerequisite of keting literature, emerge as a possible a starting point in
campaign effectiveness; that is, the commitment of an the process of strategy implementation. The framework
advertiser to a campaign partly determines the effective- has implications for agencies and advertisers. Sophisticat-
ness of the campaign itself. Implementation is successful ed agencies should recognize that relationships with cli-
when a client perceives a campaign as effective and ents are a valuable strategic asset. The ability of an agency
fulfilling its stated objectives. Commitment to a campaign to develop and maintain high levels of trust and to build an

advertiser’s commitment to a relationship is important. within the context of advertiser–agency relationships.
Furthermore, advertisers might consider trust as an im- Although much of the supporting discussion contextual-
portant screening criterion. In addition to providing eval- izes the model within advertiser–agency relationships, the
uative criteria for selecting potential partners, the frame- model may generalize to other contexts. The model may
work might suggest useful metrics for managing relation- help to inform strategy implementation across a range of
ships with agencies, for example, managing fit through inter-organizational contexts, for example, relationships
the communication of goals and objectives. Implicit in between marketing strategy and research firms and their
this paper are a number of directions for further research. clients, and more generally, the strategy implementation
A first step is to empirically test the proposed framework process within networks of marketing relationships.

Atlanta L. Stoyle or Leonard V. Coote
UQ Business School
Brisbane, Qld 4072
Australia
Phone: +61.7.3365.9721
FAX: +61.7.3365.6988
E-Mail: a.stoyle@business.uq.edu.au

VALUE-BASED DIFFERENTIATION IN BUSINESS RELATIONSHIPS:
GAINING AND MAINTAINING KEY SUPPLIER STATUS
Wolfgang Ulaga, ESCP-EAP European School of Management, France
Andreas Eggert, University of Paderborn, Germany
SUMMARY Lohmöller’s (1989) partial least square (PLS) latent path

modeling (see Figure 1). Our findings suggest that each
Many business customers today consolidate their value driver contributes to some extend to explain ob-
supply bases and implement preferred supplier programs. served variance in overall value. Though none of the
As a consequence, vendors increasingly face the alterna- differentiators should be neglected, the potential for dif-
tive of either gaining a key supplier status with their ferentiation varies considerably.
customers, or being pushed into the role of a back-up
supplier. As product and price are no longer key differen- In our study, relationship benefits display a stronger
tiators, suppliers search for new ways to differentiate potential for differentiation than cost considerations when
themselves in a buyer-seller relationship. it comes to deciding which supplier to name “primus inter
pares” among a set of available suppliers (77% of ob-
This research investigated avenues for differentia- served variance versus 17% of observed variance in our
tion through value creation in business relationships. study). At first sight, this result appears contradictory to
Building on Cannon and Homburg’s (2001) categoriza- conventional wisdom emphasizing costs in general and
tion of relationship costs, we first developed an integrated price in particular as key variables in purchasing deci-
framework of relationship value. In our research, we sions. A closer look at the sampling procedure employed,
conceptualized value in business relationship in terms of however, resolves the apparent contradiction. As we
a two-by-three matrix distinguishing between two funda- intended to understand what sets a main supplier apart
mental dimensions of value creation, i.e., relationship from competition, we focused on incumbent suppliers
benefits and costs, and three levels at which these value that already made their way on the shortlist. Hence,
drivers operate, i.e., the supplier’s core offering, the whereas cost factors often serve as qualifiers to get a
sourcing process, and the customer firm’s internal opera- supplier on the shortlist of those vendors considered by a
tions. customer for business, benefit dimensions put a vendor in
the key supplier’s position.
Next, we designed a qualitative study to identify and
describe customer-perceived key value drivers in each Looking at individual value drivers, we identify
category of our framework. Using a grounded-theory service support and personal interaction as core differen-
approach, we conducted depth-interviews with ten senior tiators. In the present study, these sourcing benefits offer
purchasing managers in nine U.S. manufacturing compa- the strongest potential for differentiation, explaining more
nies. The results of this exploratory step suggested that than half of the observed variance (52%). They are fol-
nine core drivers allow suppliers to set themselves apart lowed by operation benefits. A supplier’s know-how and
from competition in buyer-seller relationships. These key its ability to improve a customer’s time-to-market explain
differentiators are product quality, delivery performance, 21 percent of the observed variance. Operation costs
and direct product costs at the level of a supplier’s core (10% of observed variance), along with core benefits, i.e.,
offering; service support, personal interaction and acqui- product quality and delivery performance (8% of ob-
sition costs in the sourcing process; and supplier know- served variance), and acquisition costs (7% of observed
how, time-to-market, and operation costs at the level of variance) display each a moderate potential for helping a
the customer’s operations. Our depth-interviews further firm gain and maintain key supplier status. Finally, direct
allowed to probe into the various facets of each value product costs, i.e., price, show the weakest potential for
driver and to provide directions for collapsing the dimen- differentiation (3% of observed variance).
sions into an overall value index.
Overall, these findings provide strong empirical ev-
We finally conducted a nationwide cross-sectional idence that the core product and its price are no longer key
survey among 400 senior purchasing managers in U.S. differentiators in customer-supplier relationships. Rather,
manufacturing companies to specifically assess how the offering value through personal interaction and service,
value drivers identified in the previous qualitative stage access to know-how and increased time-to-market, have
contribute to differentiating key suppliers in a business become key to securing a key supplier position.
relationship. Empirical data were analyzed using

FIGURE 1
Formative Higher-Order Measurement Model
R2 = 0.73
Relationship
Value
0.77 -0.17
Relationship Relationship
Benefits Costs
0.11 0.68 0.27 0.22 0.44 0.61
Core Sourcing Operations Direct Acquisition Operation

Benefits Benefits Benefits Costs Costs Costs
(N.B.: All parameter estimates are significant at the 5 percent level. The variance explained by each formative value
dimension is obtained by multiplying the path coefficient of the second-order dimension with the coefficient of the
respective first-order dimension. Example: variance explained by sourcing benefits = 0.77 x 0.68 = 0.52).

Wolfgang Ulaga
ESCP-EAP European School of Management
79 avenue de la République
75543 Paris Cedex 11
France
Phone: +33.1.49.23.26.10
FAX: +33.1.49.23.22.48
E-Mail: wulaga@escp-eap.net

AFFECT AND CONATION IN BUSINESS-TO-BUSINESS
RELATIONSHIPS: AN EMPIRICAL ANALYSIS OF
LOYALTY LIFECYCLE SEQUENCE
Chad Ruel Allred, Brigham Young University, Provo
SUMMARY ship maturity: relational governance (Heide 1994; Mac-

neil 1980; Rindfleisch and Heide 1997), core loyalty
Loyalty measures in early loyalty studies were often (Dick and Basu 1994), and perceptual orientation in
overly simplistic, concerned primarily with repeat pur- marketing exchange (Konovsky 2000). Contrary to theo-
chase behaviors (Jacoby and Chestnut 1978). Researchers ry, results show greater relationship maturity to be more
now believe loyalty to be a more complex phenomenon, indicative of affective loyalty than conative loyalty. High-
manifest through both attitude and behavior (Day 1969), er levels of relational governance increase the likelihood
and that the attitudinal components of loyalty may be of membership in the affective phase of the lifecycle. Core
characterized by the psychological processes of cogni- loyalty is greater for members of the affective phase. Core
tion, affect, and conation (Dick and Basu 1994). Recent loyalty has a greater marginal effect on exchange satisfac-
theory further suggests that these processes constitute a tion for members of the affective phase. Members of the
loyalty lifecycle: cognitive loyalty affective loyalty con- affective phase place comparatively less weight on distri-
ative loyalty (Oliver 1999). Little has been done to corrob- butional outcomes in marketing exchange.
orate or refute this perspective of loyalty. This study seeks
to (1) classify business-to-business exchange participants Results suggest that a naïve form of conative loyalty
into either affective or conative phases of the loyalty can develop in business-to-business relationships prior to
lifecycle, (2) identify the sequencing of these two loyalty the development of affective loyalty. It is likely that naïve
phases, and (3) determine how loyalty phase might influ- conative loyalty is fostered by socio-emotional cues with-
ence performance evaluations of complex business-to- in the business-to-business “consumption community”
business marketing exchange. that are telegraphed by firm reputation and brand image
(Friedman, Vanden Abeele, and De Vos 1993). A similar
Three thousand four hundred eighty six business- phenomenon occurs in consumer markets where brands
customers of a large systems software producer represent- are routinely extended to embrace new or unfamiliar
ing 2,730 institutional buyers and channel firms from 46 products. For these products, naïve conative loyalty drives
countries are classified into affective or conative phases initial purchase and use, which ultimately leads to expe-
of the loyalty lifecycle using a finite mixture model. rience and the formation of affective loyalty. In business-
Results show repeat purchase intent to be highly correlat- to-business relationships, naïve conation and affect may
ed with cognitive and affective attitude. Latent class then interact to create the more complex form of conative
regression shows these measures to explain 67 percent of loyalty characterized in loyalty lifecycle theory.
the variance in repeat purchase intent. For members of the
affective class, repeat purchase intent is most highly Loyalty-to-action defines to the focal objective of a
correlated with overall product and service satisfaction. loyalty behavior, such as loyalty-to-purchase or loyalty-
For members of the conative class, repeat purchase intent to-recommend. Study results also show loyalty lifecycles
is most highly correlated with a strategic commitment to to be influenced by loyalty-to-action. In the modeled
the producer. business-to-business relationships the transition from naïve
conation to affection occurs more rapidly in the loyalty-
Affective and conative lifecycle phase sequence is to-purchase lifecycle than in the loyalty-to-recommend
manifest through correlations between business-custom- lifecycle. References available upon request.
er loyalty phase membership and three proxies of relation-

Chad Ruel Allred
Brigham Young University
683 TNRB
Provo, UT 84602
Phone: 801.422.3540
FAX: 801.422.0108
E-Mail: cra@byu.edu

THE EFFECT OF TOTAL AND ASYMMETRIC SPECIFIC ASSET
INVESTMENT ON SUPPLIER-BUYER RELATIONSHIP:
A STRUCTURAL MODEL
Taewon Suh, Texas State University, San Marcos
Henry Yu Xie, Saint Louis University, St. Louis
Ik-Whan G. Kwon, Saint Louis University, St. Louis
SUMMARY also found for the effect of asymmetric specific asset

investment on conflict among the partners to a relation-
Although a body of research on specific asset invest- ship. The asymmetric specific asset investment, as a result
ment has been built and conducted on channel relation- of the difference of specific asset investment among
ship, to the best of our knowledge, few studies have partners in a relationship, was found to be positively
examined the direct impact of specific asset investment related to conflict among the partners. Asymmetry in
(SAI) on commitment, trust, and conflict in an exchange specific asset investment by partners in the relationship
relationship. More specifically, magnitude and symmetry denotes lower reciprocal engagement in the relationship,
of specific asset investment hasn’t received considerable therefore leading to possible rise of conflict and decline of
attention in the literature yet. The current study attempts trust among the partners. This finding is in conformity to
to extensively analyze the effect of specific asset invest- the previous findings that power asymmetry in channel
ment in supplier-buyer relationship by dealing with mag- relationship is likely to generate lower trust and continuity
nitude and symmetry of specific asset investment. This than symmetry in the relationship. In summary, this cur-
study is posited to fill the void in the following research rent study confirms our observation on the focal effects of
areas. First, this study measures and uses SAI from both specific asset investment on business relationship/part-
sides of a channel relationship to investigate more detailed nership. The results also reveal that the effect of specific
effect of SAI on variables of relationship, i.e., commit- asset investment is mediated by decision-making uncer-
ment, trust, and conflict. Second, this study also intends to tainty. That is, the magnitude and symmetry of the specific
test the mediating role of decision-making uncertainty asset investment in a channel relationship is closely asso-
between SAI and the three dependent variables. The ciated with uncertainty in decision-making, which inten-
purpose of this study is to build upon previous studies on sifies their effect on commitment, trust, and conflict.
specific asset investment and relationship and to further
our understanding of specific asset investment and ex- Some interesting managerial implications are evident
change relationship. The rest of the paper is organized as in this current study. Firms should work to strengthen
follows: First, a brief literature review on specific asset their business relationship by increasing the total specific
investment (SAI) is presented. Second, we propose a asset investment from all the partners in the relationship
conceptual framework based on the literature review and with aim to increase the trust and commitment. Alterna-
theoretical discussion. Third, we discuss data analysis tively, firms can also strive to avoid potential conflict by
using structural equation modeling as well as results. reducing the asymmetric specific asset investment. Al-
Lastly, we discuss the conclusions and directions for though specific asset investment is likely to yield stronger
future research. relationship, asymmetric specific asset investment, as a
result of difference in partners’ investment, can potential-
The results of this current study have indicated that ly increase the conflict among the partners. Therefore,
both total specific asset investment and asymmetric spe- firms should specially pay attention to the level of asym-
cific asset investment have exerted considerable impacts metry in the specific asset investment among the partners
on facets of relationship, i.e., commitment, trust, and in the relationship. Efforts should be taken to reduce the
conflict. We found the evidence of a positive effect asymmetry in specific asset investment in case of increas-
between specific asset investment and commit, as well as ing asymmetric specific asset investment.
between specific asset investment and trust. Total specific
asset investment is positively related to commitment and As the case for any study in marketing context, the
trust in the relationship. When all the partners to the generalizability of our findings beyond our specific sam-
relationship mutually commit specific asset investment in ple may be limited. Our sample might not be representa-
the relationship, the interdependence between them is tive of all firms in supply chain as a whole. Further
subsequently boosted, resulting in a more solid and inter- research is needed to determine whether similar findings
dependent relationship. On the other hand, support was can be obtained. Further research is needed to enrich the

literature on specific asset investment. The interaction The counterbalance of effect of increasing total specific
between total and asymmetric specific asset investment asset investment and decreasing asymmetric specific as-
and its effect on the relation should be explored further. set investment is worth noticing in future studies as well.

Taewon Suh
McCoy College of Business Administration
Texas State University – San Marcos
601 University Drive
San Marcos, TX 78666
Phone: 512.245.3239
E-Mail: ts21@txstate.edu

AUTHOR INDEX
Adams, Russell 147 Dickinson, John R. 22
Allred, Chad Ruel 296 Dorsch, Michael J. 98
Antia, Kersi D. 111 Dubelaar, Chris 242
Arnold, Mark J. 234 Duffy, Cameron 18
Arnold, Todd 130 Echambadi, Raj 177
Atuahene-Gima, Kwaku 185 Eggert, Andreas 294
Aubert, Benoit 136 Evanschitzky, Heiner 63, 220
Baack, Daniel W. 159 Feick, Lawrence 125
Bae, Mueun 234 Foley, Janice 132
Baldauf, Artur 87 Foley, Linda M. 138
Bansal, Harvir 187 Fürst, Andreas 89
Barat, Somjit 253 Gangadharbatla, Harsha 24
Barb, Carmen 272 George, Morris K. 65
Barczak, Gloria 185 Geursen, Gus 242
Bates, Kenneth W. 134 Gold, Melanie A. 216
Bearden, William O. 59, 157 Goodman, Steven 242
Bischof, Christian 87 Gooner, Richard A. 32
Bishop, Melissa Maier 270 Greenley, Gordon E. 290
Blankson, Charles 13 Griffith, David A. 77
Blocker, Christopher P. 47 Griffiths, Merlyn A. 209
Blunch, Niels J. 198 Grove, Stephen J. 98
Bottomley, Paul 168 Grünhagen, Marko 132
Brady, Michael K. 91 Gummerus, Johanna 262
Briggs, Elten 130 Haghirian, Parissa 260
Brocato, E. Deanne 128 Hardesty, David M. 157
Brown, James R. 200 Harmon, Robert 238
Bush, Victoria D. 138 Harmon, Tracy R. 7
Cable, Daniel M. 11 Hausman, Angela 207
Calantone, Roger J. 30, 140 Helm, Sabrina 202
Carlson, Jay P. 157 Hemphill, Elizabeth 242
Carlson, Les 98 Hennig-Thurau, Thorsten 136
Cavusgil, Erin 94 Hirunyawipada, Tanawat 73, 183
Chan, Haksin 61 Hollmann, Thomas 218
Chang, Joseph W. 75 Holzmüller, Hartmut H. 96
Chao, Paul 240 Homburg, Christian 89, 179, 222
Chaudhuri, Arjun 161 Honebein, Pete C. 136, 194
Chen, Qimei 77 Hooley, Graham J. 290
Cheng, Shirley Y. 254 Hopkins, Christopher D. 98
Childers, Terry L. 154 Horowitz, David M. 91
Chmielewski, Danielle A. 79 Hoyer, Wayne D. 222
Choi, Pilsik 105 Huggins, Kyle A. 134
Conant, Jeffrey S. 177 Hyvönen, Saara 290
Coote, Leonard V. 28. 112, 292 Ingram, Kendra 20
Coulter, Keith S. 155 Jarvis, Cheryl Burke 218
Coulter, Robin A. 155 Jensen, Ove 179
Cravens, David W. 87 Johnson, Devon 58
Crittenden, Victoria L. 18 Johnson, Jean L. 236
Cui, Anna Shaojie 30 Jones, William J. 154
Daugherty, Terry 24 Kachroo, Pushkin 200
Dawar, Niraj 288 Kahn, Barbara E. 196
de Almeida, Stefânia Ordovás 36 Kajalo, Sami 290
de Gregorio, Federico 26 Kalafatis, Stavros P. 13
DelVecchio, Devon S. 154 Kaleka, Anna 32
Dev, Chekitan S. 200 Kilbourne, William 132
Dholakia, Ruby Roy 204 Kilburn, Ashley 258

Kim, Seung H. 234 Rice, Marshall 1
Kleiser, Susan B. 128 Rindfleisch, Aric 111
Koschate, Nicole 222 Rupp, Deborah 181
Krishen, Anjala S. 200 Russ, Courtney 20
Kurpis, Lada V. 272 Sääksjärvi, Maria 262
Kwon, Ik-Whan G. 298 Sauer, Paul L. 240
Kwong, Jessica Y. 254 Sebastiao, Helder J. 114
Lam, Desmond 9 Shin, Dongwoo 256
Lambert, Claire 101 Singh, Jagdip 96
Landry, Tim 130 Sirdeshmukh, Deepak 96
Leach, Mark P. 126 Srivastava, Mona 238
Lee, Alvin 9 Sit, Alfred Y. 61
Lei, Jing 288 Slotegraaf, Rebecca J. 107
Leigh, James H. 256 Spotts, Harlan 168
Lemmink, Jos 288 Stafford, Marla Royne 258
Lentz, Patrick 96 Stan, Simona 272
Lim, Lewis K.S. 107 Stephen, Andrew T. 28, 112
Liu, Annie H. 126 Stern, Barbara B. 258
Lou, Yung-Chien 75 Stoyle, Atlanta L. 292
Lu, Min 148 Suh, Taewon 234, 298
Luce, Mary Frances 196 Sung, Yongjun 26
Luenemann, Ulrich F. 223 Tang, Felix T. 92
Lukas, Bryan A. 79 Thota, Sweta Chaturvedi 103
Malhotra, Neeru 15 Till, Brian D. 159
Mani, Sudha 111 Townsend, Janell D. 94
Marquardt, Adam 286 Troy, Lisa C. 117
Martin, Kelly D. 236 Tu, Yanbin 148
Matear, Sheelagh 290 Tuominen, Matti 290
Menon, Kalyani 187 Tzeng, Dale Taoping 168
Miller, Elizabeth G. 196 Ulaga, Wolfgang 294
Milward, Mary Anne 20 Vahie, Archna 183
Mintu-Wimsatt, Alma 20 Voorhees, Clay M. 91
Mitchell, Vincent-Wayne 44 Vorhies, Douglas W. 138
Mittal, Vikas 125 Voss, Glenn B. 11
Mizerski, Dick 9 Voss, Zannie Giraud 11
Mizerski, Katherine 99 Walsh, Gianfranco 44, 220
Mizerski, Richard 99, 101 Walters, Rockney G. 107
Möller, Kristian 290 Wangenheim, Florian V. 63, 268
Money, R. Bruce 59 Weinberger, Marc 168
Morgan, Neil A. 32 White, J. Chris 177
Mukherjee, Avinandan 15 Widing II, Robert E. 79
Nevins, Jennifer L. 59 Wilson, Elizabeth J. 18
Nijssen, Ed 96 Winsor, Robert D. 126
Nique, Walter Meucci 36 Wong, Andrew K.C. 34
Noble, Stephanie M. 77 Wood, Charles 130
Nyaga, Gilbert N. 140 Xie, Henry Yu 298
Page, Thomas J. 140 Xu, Huimin 279
Parker, Andrew M. 216 Ye, Yun 118
Paswan, Audhesh K. 117 Yeniyurt, Sengun 94
Punj, Girish N. 65 Yim, Frederick Hong-Kit 46
Qiu, Tianjiao 181 Zahay, Debra 206
Qualls, William 181 Zhang, Yinlong 125
Rajala, Arto 290 Zhao, Hongxin 234
Rajamma, Rajasree K. 109 Zhao, Miao 204
Reimann, Martin C. 223 Zolfagharian, Mohammadali 73, 109
Rexha, Dafina 99

3 PDF

Uploaded by

Copyright:

Available Formats

3 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

3 PDF

Uploaded by

Copyright:

Available Formats

2005 AMA Winter Educators’ Conference

Marketing Theory and

311 S. Wacker Drive • Chicago, Illinois 60606 • (312) 542 - 9000

Printed in the United States of America

Publications Director: Francesca Van Gorp Cooley

All rights reserved. No part of the material protected by this copyright

Consumer Behavior Jeff Inman, University of Pittsburgh

Kathleen Seiders Glenn B. Voss

A Mary J. Bitner, Arizona State

PREFACE AND ACKNOWLEDGMENTS iii

TABLE OF CONTENTS viii

COMPANY-CONSUMER AND CONSUMER-CONSUMER SOCIAL EFFECTS

Cultural Influence on Word-of-Mouth Communication

ORGANIZATIONAL IMPACT ON SERVICE DELIVERY

The Effects of Authenticity Rift on Firm Performance

Positioning in Service Firms: Model Development and Some Basic Normative

Antecedents and Consequences of Role Clarity in Explaining Employee-Perceived

INSTRUCTIONAL INNOVATION IN MARKETING EDUCATION

The Status of Cross-Functional Education in Undergraduate Marketing Curricula

On Different Teaching Pedagogies: What Happens to Your Course Evaluations?

An Assessment of a Consumer Behavior Multiple-Choice Question Taxonomy

PRODUCT PLACEMENTS EFFECTS

A Comparison of Consumers’ Responses to Traditional Advertising and Product

New Brand Worlds: A Comparison of College Student Attitudes Toward Brand

PROCESS AND BEHAVIOR

The Marketing Cocompetition Process and Strategic Alliance Instability: A System

Focal Supplier Opportunism in Retailer Category Management

AFFECT-RELATED CONSUMER RESEARCH

“No Greater Satisfaction Than to Vindicate Expectation” How Affective Expectations

Customer Delight: An Attempt to Comprehend the Dimensions That Compose the

Towards a Conceptual Framework of E-Confusion

UNDERSTANDING CUSTOMER-FIRM EXCHANGE PROCESSES

Differential Impacts of CRM on Consumer Responses: Toward an Integrative

Exploring the Phenomenon of Buyer-Seller Mismatches in Business-to-Business

Consumer Trust Norms in Multi-Channel Firms: The Role of Trust in Technology

EXPLORATIONS OF CULTURE AND ETHNOCENTRISM DIMENSIONS

Validation and Application of a Bi-Dimensional Long-Term Orientation Scale

Consumer Socialization of Third Culture Kids in a Cosmopolitan City

Consumer Ethnocentrism in the German Market

NEW PRODUCT CONSUMER DECISION PROCESSES

Identifying Information Search Patterns in a Web-Based Environment: Development

Perceived Entitativity as a Moderator of Family Brand Evaluations

MARKET-BASED DRIVERS OF FIRM PERFORMANCE

The Performance Implications of Synergistic Knowledge Resource Effects in

The Impact of Market Characteristics on Order-of-Brand Entry Strategy: An Empirical

Organizational Culture Antecedents of Market-Driven Positional Advantage and

DEFENSIVE AND PROACTIVE RESPONSES TO SERVICE FAILURE AND

The Rest of the Iceberg: An Examination of Noncomplaining Service Customers

From Empathy to Forgiveness: A Prosocial Perspective in Service Failure and Recovery

STRATEGIC CONCERNS IN GLOBAL MILIEUS

International Marketing Alliance Dynamics: Empirical Findings from the

A Cross-National Study of Consumer-Firm Exchange Relationships Within the

CONVERGENT COMMUNICATIONS: MOVING BEYOND ADVERTISING

COMPETITIVE INTERACTION AND STRATEGIC DECISIONS

The Nature of Co-Opetition: Literature Review and Propositions

Technology Versus People: Two Schools of Thought on Pricing Capability Development

Dimensions and Outcomes of Relational Exchange in a Business-to-Business Context:

Entry Mode and Level of Equity: A Simultaneous Examination of Foreign Direct

Towards an Understanding of the Governance of Complex Networks of Relationships: