Lim and Acito

Development of Archetypes of International Marketing Strategy
Author(s): Lewis K. S. Lim, Frank Acito, Alexander Rusetski

Source: Journal of International Business Studies, Vol. 37, No. 4 (Jul., 2006), pp. 499-524
Published by: Palgrave Macmillan Journals
Stable URL: http://www.jstor.org/stable/3875167
Accessed: 26/08/2010 04:55
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=pal.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Palgrave Macmillan Journals is collaborating with JSTOR to digitize, preserve and extend access to Journal of
International Business Studies.
http://www.jstor.org
of International
journal Studies
Business (2006)37,499-524
? 2006Academyof International
BusinessAllrightsreserved
0047-2506$30.00
www.jibs.net
Development of archetypesof international

marketing strategy
Lewis KS Liml'2, FrankAcito' Abstract

and Alexander Rusetski"'3 The extant business literaturecontains three separate characterizationsof
internationalmarketingstrategy: standardization-adaptation,concentration-
Schoolof Business,IndianaUniversity,
1Kelley dispersion, and integration-independence.These characterizationshave, for
Bloomington,Indiana,USA;2NanyangBusiness decades, informedresearchers,students, and practitionersalikeof the strategic
School,NanyangTechnological University, options a multinationalfirm might have in formulating its cross-border
Singapore;3Schoolof Administrative
Studies, marketingapproaches. Although useful, these characterizationshave yet to
YorkUniversity,
Toronto,Ontario,Canada be unifiedwithin an integrativeclassificationscheme that considersthe gestalt
combinatorialpatterns along multiple strategy dimensions. Towardcreating
Correspondence: Lewis Lim,Division of such a classificationscheme, this paperproposesa holisticconceptualizationof
Marketingand InternationalBusiness, internationalmarketingstrategygrounded in configurationaltheory, whereby
Nanyang Business School, Nanyang
Technological University,S3-B2C-95 strategiesare viewed as multidimensionalarchetypes.We present evidence of
Nanyang Avenue, Singapore 679798, three distinctinternationalmarketingstrategyarchetypesobtained through an
Singapore. exploratorycase coding/clusteringstudy. Afterdiscussingthe characteristics,
Tel: +65 6790 4095; possible drivers,and contingent performancepotentials of these archetypes,
Fax: +65 6791 3697; we offer directionsfor future research.
E-mail:akslim@ntu.edu.sg
Journal of International Business Studies (2006), 37, 499-524.
dol:I0. I057/palgrave.jibs.8400206
Keywords:international globalmarketing;
marketing; standardization
vs adaptation;
clusteranalysis
casesurveymethodology;
configurations;
Introduction
Classification is especially important to the study of organizational strategies;
strategies consist of the integration of many dimensions which, in turn, can be
configured in seemingly endless combinations. Without a classification
scheme, the strategyresearchermust deal individually with the many variables
of interest.., and must generally assume that all combinations are possible. A
strategy classification scheme helps bring order to an incredibly cluttered
conceptual landscape.
(Hambrick,1984, 27-28)
For the past four decades, business scholars have sought to

characterize and classify the international marketing strategies of
multinational firms (Buzzell, 1968; Keegan, 1969; Hovell and
Walters, 1972; Ozsomer and Prussia, 2000). Of ultimate concern
among these scholars is the performance potential associated with
any type of international marketing strategy. A more fundamental
goal of classifying these strategies, though, is simply to help
researchers,students, and practitioners in the field understand the
different strategic options a multinational firm might have in
Received: 25 November 2003
Revised: 14 October 2005 structuring its marketing approaches across country markets.
Accepted: 2 November 2005 For the most part, the literature has characterized international
Online publication date: 25 May 2006 marketing strategy from one of three perspectives (Zou and
International marketing strategy archetypes LewisKSLimet al
500
Cavusgil, 2002). The most common characteriza- keting strategy. Specifically, as the standardization-
tion of international marketing strategy is along the adaptation characterization is concerned with the
standardization-adaptation dimension (e.g., Jain, degree of harmonization of the marketing mix
1989). From this perspective, international market- elements, it captures the market offeringaspect of
ing strategies are differentiated according to the international marketing strategy. In comparison, as
degree of standardization (vs adaptation) pursued the concentration-dispersion characterization
with respect to one or more of the marketing mix deals with the geographical design of the marketing
elements (e.g., product, price, promotion). Thus, a value chain, it captures the structural/organizational
standardization strategy is characterized by the aspect of international marketing strategy. Finally,
application of uniform marketing mix elements as the integration-independence characterization
(i.e., product design, pricing, distribution, etc.) concerns the planning, implementation, and con-
across different national markets. Conversely, an trol elements of competing in a global marketplace,
adaptation strategy is characterizedby the tailoring it captures the competitiveprocessaspect of interna-
of marketing mix elements to the needs of each tional marketing strategy. Together, these charac-
market. terizations potentially provide rich descriptions of
A second way of characterizing international the ways in which a multinational firm can choose
marketing strategy stems from the concentration- to serve its customers, organize itself, and compete
dispersionperspective (e.g., Roth, 1992). This per- in the international marketplace.
spective, rooted in Porter's (1986) analysis of Unfortunately, the potential to richly describe
international competition and most recently holistic patterns of international marketing strate-
reflected in Craig and Douglas's (2000) theory of gies does not appear to be completely facilitated by
configural advantage, is concerned more with the conceptual and methodological advances in the
geographic design of the international marketing field. Until very recently, scholars have relied on
organization. The underlying premise of this per- unidimensional schemes to discuss international
spective is that a multinational firm should seek an marketing strategies, and/or have discussed them
optimal geographic spread of its value-chain activ- from a single perspective. Forexample, Jain's (1989)
ities such that synergies and comparative advan- treatment of the construct 'marketing program
tages across different locations can be maximally standardization' seems to be a general unidimen-
exploited. International marketing strategies, then, sional one, with complete standardization on one
are differentiated according to the extent to which end of the pole and complete adaptation on the
one or more aspects of the marketing value chain other.1 Likewise, Olusoga (1993) defines and mea-
are consolidated or 'concentrated' at particular sures 'market concentration' mostly in general
geographic locations, vs being scattered or 'dis- terms: that is, without distinction among varying
persed' across various country markets. degrees of concentration for different value chain
A third characterization of international market- activities. Consequently, one would suspect that
ing strategy is concerned with how competitive students and practitioners exposed to this literature
marketing activities across country markets are might be, at best, equipped to think of interna-
orchestrated. This perspective, referred to here as tional marketing strategies in terms of simple
the integration-independence perspective, is heavily categorical labels such as 'standardized', 'adapted',
influenced by the competitive 'warfare'description 'concentrated', etc., without a deep understanding
of Hamel and Prahalad (1985). The key question of what they imply at a holistic level.
here is whether a multinational firm treats its The recent effort by Zou and Cavusgil (2002) to
subsidiary units as standalone profit centers (i.e., model the construct of international (global)
independently), or as parts of a grander strategic marketing strategy represents a significant step
design (i.e., as integrated units). Accordingly, inter- toward a truly multidimensional approach to this
national marketing strategies should be differen- concept. Zou and Cavusgil (2002) propose a
tiated according to the degree of consultation and second-order factor construct, termed the 'GMS',
integrated action across markets, and the will- which overarches eight first-order dimensions of
ingness to which a performance outcome in any global marketing strategy spanning the three broad
one market is sacrificed in order to support the characterizations.2 Thus any multinational firm's
competitive campaigns in other markets. global marketing strategy, with given degrees of
Each of the above three major characterizations standardization, concentration, and integration,
captures an important facet of international mar- can be captured by a single GMS score. However,
journal of International Business Studies

501
such a second-order factor construct, while provid- marketing options that can be or have been
ing an aggregate measure of the degree of 'global- pursued by multinational firms, and allows a more
ness' of a multinational's marketing strategy, is not meaningful comparative evaluation of interna-
modeled to take into account possible interactions tional marketing strategies. Second, beyond identi-
among the first-order strategy dimensions. Specifi- fying strategy archetypes per se, our approach
cally, by virtue of its linear approach, it is not provides a starting point for inquiring into their
designed to capture qualitatively distinct patternsof evolution as well as for a contingent analysis of
strategy made up of different combinations of their performance potential. By virtue of the
strategy elements. This limitation is regrettable multidimensionality in their configurational pat-
because the various strategy elements are likely to terns, the archetypes contain rich information
interact and combine themselves into multidimen- about the marketing behaviors of multinational
sional 'gestalts' (Miller, 1981, 1986; Meyer et al., firms and about factors that might contribute to
1993). their effectiveness. This presents a valuable oppor-
All of the above problems point to the need for a tunity to consolidate our knowledge of interna-
more intricate, yet robust, method for describing tional marketing and international strategy. Third,
and classifying international marketing strategies. our approach has pedagogic value. By demonstrat-
To that end, we present a holistic and unified ing the presence of archetypes of international
approach to viewing international marketing strat- marketing strategies and discussing their contin-
egy, an approach that is grounded in the config- gent performance potential, we bring forth an
urational theory of organizations (Miller, 1981, important set of ideas and a useful framework for
1986, 1996; Meyer et al., 1993), and which sees future teaching of international marketing. Stu-
strategies as multidimensional archetypes. This dents and practitioners could then better appreciate
approach not only takes into account the different the multifaceted nature of international marketing
ways in which any given international marketing without having to rely on unidimensional labels or
strategy can be characterized (i.e., in terms of dichotomies.
standardization-adaptation, concentration-disper- The next section of this paper reviews the field's
sion, and integration-independence), but also con- past efforts in characterizing and classifying inter-
siders the overall configurational pattern of the national marketing strategy,and reiterates the need
strategy,in terms of its positions along the different for a more robust approach to delineating strate-
dimensions. Thus our approach offers significant gies. We then introduce our proposed archetype
advantages over any single characterization of approach and highlight its foundations in config-
international marketing strategy, or any single urational theory. Next, we illustrate the utility of
aggregate score of the 'globalness' of marketing our approach by examining data from our case
strategy (Zou and Cavusgil, 2002). coding study for the presence of archetypes. Based
To support our conceptualization, we report a on the findings, we explore the likely drivers of
case coding/clustering study in which we utilized a archetypes and potential archetype performance
taxonomic procedure for uncovering several dis- variations.
tinct archetypes of international marketing strat-
egy. Although the results cannot be taken as The study of international marketing
conclusive because of the exploratory nature of strategy: toward a unified multidimensional
our analysis, the presence of archetypes within a characterization
limited sample provides preliminary evidence for The marketing literature has dealt with interna-
the efficacy and theoretic value of the configurational advertising issues since at least the early
tional approach. 1960s (e.g., Elinder, 1961; Roostal, 1963; Fatt,
Our proposed approach makes three fundamental 1964), but it was Buzzell (1968) who offered the
contributions to the international marketing litera- first systematic discussion of standardization as a
ture. First, our notion of strategy archetypes type of international marketing strategy. A standar-
arguably represents the first truly multidimensional dization strategy was defined as the harmonization
way of describing and classifying international of the various marketing mix elements (e.g.,
marketing strategies. By differentiating strategies product design, pricing, distribution, etc.) across
in terms of their relative proximities in multi- different country markets. Conversely, a localiza-
dimensional space, our approach provides a novel tion or adaptation strategy would be the adoption
integrative perspective on the various strategic of a unique marketing mix in each market. Buzzell
Journal of International Business Studies

; LewisKSLimet al
International marketing strategy archetypes
502
(1968) provided several reasons for favoring a of view. He described four qualitatively different
standardization policy, including cost savings, con- strategies, which could arise from crossing the
sistency in customer dealings, and the exploitation product-standardization-versus-adaptation dimen-
of a universal appeal, and urged multinational sion with the promotion-standardization-versus-
executives to consider moving away from the adaptation dimension. In addition to these four
then-prevalent adaptation policy. types of strategy, Hovell and Walters (1972) sug-
Following Buzzell (1968), the international mar- gested including variations in terms of the other
keting literature continued to debate the merits of a marketing mix elements, such as distribution and
standardization strategy. Perhaps the most notable personal selling approaches. In the empirical realm,
proponent of standardization was Levitt (1983), Sorenson and Wiechmann (1975) found that multi-
who argued that diminishing cultural differences national firms vary their marketing approaches
across countries due to technological advance- across different country markets along as many as
ments necessitate a global (standardized) strategy 12 dimensions. Finally, Quelch and Hoff (1986)
that best captures worldwide economies of scale. discussed partial vs full standardization as well as
Other supporters of this view included Rutenberg partial vs full adaptation along more than 20
(1982), Henzler and Rall (1986), Jain (1989), and dimensions of business functions, products, mar-
Zou et al. (1997), who provided various arguments keting mix elements, and countries.
revolving around scale advantage and consistency Notwithstanding the momentum toward a multi-
in marketing planning and actions. On the other dimensional approach to describing and capturing
side of the debate were scholars such as Boddewyn international marketing strategies, the preponder-
et al. (1986), Kotler (1986), Douglas and Wind ance of studies until at least the mid-1980s had
(1987), Ohmae (1989), and Sheth (1986), who really focused only on the market offering aspect
variously pointed out the barriers to worldwide (i.e., the marketing mix) of international marketing
marketing standardization, including governmen- strategy. However, two streams of research were to
tal and trade restrictions, inter-country differences emerge to give rise to two additional strategy
in marketing infrastructure,and local management characterizations - description schemes based on
resistance. the structural/organizational and on the competi-
The debate between standardization and adapta- tive process aspects of international marketing.
tion remained largely unresolved through the As identified by Zou and Cavusgil (2002), the
1990s (see Theodosiou and Leonidou, 2003). More concentration-dispersion characterization of inter-
significantly, perhaps because of the intensity and national marketing strategy can be traced to
prominence of the debate, the literature in the Porter's (1986) 'design' framework. The focus of
1980s and 1990s began to treat standardization and analysis is on the structuring of value-chain
adaptation as fixed alternative options in interna- activities (e.g., R&D/product development, after-
tional marketing strategy. The titles of several sale service, logistics and distribution) across inter-
articles published during this period (e.g., Samiee national locations. The measurement scale implied
and Roth, 1992; Szymanski et al., 1993; Solberg, by this perspective is in terms of geographic
2000) suggested that the academic community had 'concentration' vs 'dispersion' of each of the
come to accept the concept of international value-chain functions (Roth et al., 1991; Roth,
marketing strategy itself as falling along a single 1992). Porter (1986) argued that multinational
continuum of standardization vs adaptation. More- firms should seek an optimal value-chain 'config-
over, several scholars had begun using unidimen- uration' such that scale and national comparative
sional scales to measure variations of international advantages are exploited, while balancing respon-
marketing strategies. For example, in a follow-up siveness to local needs. More specific to the
empirical study to Jain (1989), Samiee and Roth marketing area, Craig and Douglas (2000) expli-
(1992) used a single index to measure global cated how the spatial configuration of marketing
marketing standardization. value-chain activities, made up of differential
Yet a careful reading of the literature reveals that concentration-dispersion levels, could influence
the state of knowledge was also evolving toward a the tightness of the operational interlinkages across
multidimensional view of international marketing markets and the development of border-spanning
strategy. As early as 1969, Keegan (1969) proposed learning and market-sensing capabilities. These
looking at the issue of standardization vs adapta- outcomes in turn determine the 'configural' advan-
tion from both the product and promotion points tage of the firm.

503
The other emerging characterization, the integra- Table 1 shows two somewhat distinct patterns of
tion-independence characterization, is grounded international marketing strategy, represented by
in the competitive 'warfare' description of Hamel variations along the first-order dimensions of the
and Prahalad (1985). This perspective is concerned GMS model. Using the factor loadings reported in
with the extent to which a multinational firm Zou and Cavusgil (2002) as weights, the overall
orchestrates its competitive moves on an interna- GMS scores computed for the two strategies turn
tional basis and leverages its competitive position out to be exactly equal.
in one market to achieve an advantage in other In summary, although the literature has begun to
markets (e.g., by cross-subsidizing its competitive acknowledge the need to integrate and unify the
campaigns across different countries). In other various characterizations of international market-
words, the key issue here is whether a multinational ing strategy in order to derive a rich description and
firm treats its subsidiaries as independent profit classification scheme, existing conceptual and
centers or as an integrated group of business units. methodological advances have yet to completely
Hamel and Prahalad (1985) illustrate the latter facilitate this endeavor. In particular, progress has
behavior with an example about Goodyear retaliat- yet to be made far beyond the use of unidimen-
ing against Michelin's incursion into the US tire sional labels such as 'standardized', 'adapted', and
market by launching an attack in Michelin's home 'concentrated', or the use of aggregate scores of
base, Europe, thereby tying up Michelin's resources globalness, to describe international marketing
and restraining its ability to compete. Such actions strategies. The field is still in need of a more
require close coordination among different country sophisticated and robust method for delineating
offices and compromises in country-level profit- patterns and types of strategy.
ability. This perspective therefore implies that
international marketing strategy should be mea- Strategy archetypes: a configurational
sured in terms of the degree of integration in approach
competitive moves and decision-making, the inte- The configurational theory of organizations (Miller,
gration of competitive response, and cross-unit 1981, 1986; Meyer et al., 1993) supplies the primary
communication and mutual consultation (Hout basis for conceptualizing our proposed holistic
et al., 1982; Yip, 1989). approach to viewing international marketing strat-
Noting the diverse (three-fold) conceptualiza- egy. Configurational theory holds that organiza-
tions of international marketing strategy, Zou and tional effectiveness arises out of superior
Cavusgil (2002) made the first important attempt at combinations of strategic and structural character-
the turn of the century to unify the concept and istics (Millerand Mintzberg, 1983; Doty et al., 1993;
develop a measure that reflects its multidimen- Ketchen et al., 1997). In keeping with the config-
sional nature. They conceived of a second-order urational perspective, our approach is grounded in
factor, labeled the 'GMS' (acronym for global the premise that any concept of strategy is inher-
marketing strategy), that incorporated eight first- ently multidimensional, and that various elements
order strategy sub-dimensions spanning standardi- of strategy can interact or combine differently in
zation-adaptation, concentration-dispersion, and multidimensional space. As Miller (1986, 235-236)
integration-independence (one of the dimensions, aptly puts it in his well-noted piece:
standardized price, was dropped during their The elements of strategy,structure,and environment often
empirical analysis). With the GMS construct, they coalesce or configure into a manageable number of
made it possible to capture the overall 'globalness' common, predictively useful types that describe a large
of a firm's international marketing strategy using a proportion of high-performing organizations. The config-
urations (or 'gestalts', or 'archetypes',or 'generic types') are
single score. The limitation of the GMS model, said to be predictively useful in that they are composed of
however, is that it is not able to detect the tight constellations of mutually supportive elements.
dominant combinatorial patterns in the data,
patterns that would reveal the true multidimen- Applied to the present context, it is the different
sional character of international marketing strate- 'constellations', or configurations, of strategy ele-
gies. Furthermore, to use the GMS score alone to ments that make up a 'universe' of international
account for variations in multinational firm per- marketing strategies. Thus, to richly describe inter-
formance may prove causally ambiguous, as several national marketing strategies, one must look
qualitatively distinct patterns of strategy could beyond single strategy dimensions for modal
technically share the same GMSscore. For example, combinatorial patterns across multiple dimensions.

504
Table 1 Exampleof two combinations of strategy ratingssharing the same GMSscore
Strategydimension Firmla Firm2a
GMSI: product standardization 3.86 -0.35

GMS2: promotion standardization 2.10 2.14
GMS3: standardized channel design 0.29 1.64
GMS4: concentration of marketing activities 0.78 -0.64
GMS5: coordination of marketing activities 1.65 5.32
GMS6: global market participation 2.32 1.86
GMS7: integration of competitive moves 1.29 2.63
Total GMS score 6.82 6.82
aFactorscores for strategy dimensions and for the GMSwere obtained by multiplyingstandardizedfactor loadings reported in Zou and Cavusgil(2002)
by corresponding standardized item scores from their original data set and summing up resulting products.
-.- Firm 2
2
2GMS- GS -uG-Firm
-1GMS1
GMS1 GMS2
GMS2 GMS3
GMS3 GMS4
GMS4 GMS5
GMS5 GMS6
GMS6 GMS7
GMS7
When patterns are distinctive and exemplary, they leaving a limited number of superior strategic
can be called the archetypes of strategy. When options. Second, borrowing from 'gestalt' principles
discovered, such archetypes may be represented (e.g., Kelly, 1955), organizations themselves tend to
using graphical snake-like line profiles that chart be drawn toward configurations of strategy ele-
different degrees of standardization, concentration, ments that are internally harmonious and mutually
and integration among themselves. However, reinforcing. Presumably,this could occur either as a
unlike the line profiles used by Wind (1986) and result of the organization's own strategic choice
Douglas and Wind (1987) to illustrate variations of (Child, 1972) or through industry mimetic actions
international marketing strategy, the archetypes and normative pressures (DiMaggio and Powell,
conceived here are not arbitrary line drawings or 1983), Third, as suggested by studies of organiza-
stylized prototypes. There is an implied assertion tional evolution (e.g., Miller and Friesen, 1984),
that archetypes are theoretically meaningful strate- organizational change often occurs either in small
gic forms that are at least viable and potentially incremental steps or in 'quantum leaps', implying
high-performing, if not approaching the 'ideal' that many hybrid forms are often avoided or left
types assumed by Doty et al. (1993) in their study unexplored. Consequently, even as the study of
of configurations. strategy moves onto a more complex, multidimen-
Miller (1981, 1986) offers three theoretical rea- sional plane, researchers need to deal with only a
sons for believing that only a few configurational limited number of strategy configurations out of
combinations would dominate any given strategy numerous technically possible combinations.
domain. First, from a population ecology perspec- Implementing the configurational approach to
tive (e.g., Hannan and Freeman, 1977), the envir- the study of international marketing strategy
onment tends to select out unviable, unsustainable, would involve three major steps. First, the specific
or otherwise uncompetitive strategies, thereby dimensions that collectively define the concept of
journalof InternationalBusinessStudies
505
international marketing strategy should be identi- taxonomic approach to develop international mar-
fied. This is in line with the recommendation of keting strategy archetypes. Our subsequent explora-
Ketchen et al. (1993), who find that configurations tion of contingent performance issues will,
derived from theory-based dimensions have greater however, leave scope for the future use of the
predictive validity. As the literature accepts three typological approach.
treatments of strategy - standardization-adaptation Most importantly, our proposed approach serves
(capturing the market offering aspect), concentra- to address current limitations in the characteriza-
tion-dispersion (capturing the structural/organization of international marketing strategy. First, our
tional aspect), and integration-independence approach represents a significant advancement
(capturing the competitive process aspect) - these from the traditional unidimensional way of classi-
three broad groups of dimensions should be used fying strategies based on any of the standardiza-
concurrently to identify international marketing tion-adaptation, concentration-dispersion, or
strategy archetypes. Second, a procedure for select- integration-independence perspectives alone. The
ing a suitable sample of international marketing archetypes conceived here are multidimensional in
strategies, to objectively quantify these strategies in nature, and they encompass strategy elements from
terms of the three sets of dimensions, and to all three perspectives. Second, our approach differs
statistically detect and delineate combinatorial from Zou and Cavusgil's (2002) second-order factor
patterns within the sample, should be performed. model in that it permits the identification of gestalt
To demonstrate this procedure, we describe below a patterns resulting from different combinations of
structured case coding/clustering methodology strategy dimensions, as opposed to a single aggre-
used for uncovering strategy archetypes. Third, gate measure of strategy 'globalness'. Overall, our
the derived archetypes should be meaningfully approach facilitates a more profound understand-
interpreted and analyzed for possible drivers and ing of international marketing strategy in terms of
contingent performance factors. To that end, we holistic combinatorial patterns.
examine the background characteristics of our
uncovered archetypes and look to the broader Evidence of archetypes: an exploratory case
international business literature to make predic- coding/clustering study
tions about their relative performance. To demonstrate the utility of our proposed config-
It should be noted that the approach outlined urational approach, we undertook an exploratory
above is closer to the taxonomicapproach espoused case coding/clustering study, which, we believe,
by McKelvey (1975) and Hambrick(1984), and used was the first of its kind in international marketing.
by Miller and Friesen (1977, 1978, 1980), Woo and Our study consisted of two phases. Phase I involved
Cooper (1981), and Hambrick and Schecter (1983) a 'case survey' methodology (Larsson, 1993), also
to empirically uncover configurational archetypes known as the 'structured case content analysis'
of organizational design or strategy. A taxonomic method (Jauch et al., 1980), to quantify the
approach is appropriate when extant theory does international marketing strategies of firms featured
not yet permit an a prioriidentification of superior in a sample of published cases. We believe that
strategy configurations (Meyer et al., 1993). How- published cases are a useful source of data for our
ever, in more mature domains, where typologies of research. Even though these cases are written
effective configurations have already been devel- mostly for teaching purposes, they contain factual
oped (e.g., the typology of Miles and Snow, 1978), and detailed information about the practices and
researchershave been able to employ a typologicalor strategic designs of actual firms. Such information,
'ideal profile' approach to test hypotheses about having been gathered through tedious fieldwork, is
certain configurational ideal-types (e.g., Gresov, extremely rich and not easily extracted through
1989; Doty et al., 1993; Vorhies and Morgan, quantitative surveys. Moreover, case writers nor-
2003). Those hypotheses are concerned mostly mally rely on multiple key informants and archival
with the effect of 'fit' among elements of strategy records to construct the cases (Jauch et al., 1980).
(measured in terms of deviation from ideal profiles) Much of the information contained in the cases
on firm performance. Related to fit is the issue of would have been cross-validated to the extent
equifinality among configurations: that is, whether possible (for prescriptions of case-writing practices,
different configurations achieving varying forms of see Corey, 1998; Roberts, 2001). Compared with
fit can be equally effective (Gresov and Drazin, using other forms of secondary data, such as annual
1997). In this paper, we primarily pursue a reports, the use of cases also affords the advantage

SInternational marketing strategy archetypes LewisKSLimet al
506
of enabling measures of process variables, that is, moderately successful adapters in population ecol-
variables dealing with the processes of strategy ogy terms (Hannan and Freeman, 1977). Conse-
formulation and implementation. Such processes quently, taxonomic archetypes found among these
often are not publicly observable but are central to companies would represent viable strategic forms.
at least one dimension of international marketing (This does not imply, however, that any of these
strategy (the integration-independence dimen- archetypes is maximally effective. Archetype per-
sion). Finally, with cases, unlike with survey formance potential has to be examined in relation
respondents, the data source (i.e., the case docu- to contingent fit, a subject we shall address later in
ments themselves) resides permanently in our this paper.)
premises. It is always possible to return to the The process of translating the qualitative descrip-
source for clarification of item ratings and/or for tions in the case into quantitative measures was
exploration of additional items. facilitated by a detailed coding scheme. Appendix B
To overcome the problem of comparability across shows the coding scheme that was used for
cases due to different units of analysis and different measuring the international marketing strategies
case types, we select only cases that describe, in a featured in our sample of cases. The coding scheme
reasonably detailed manner, the actual interna- contained 16 scale items: seven measuring the
tional marketing strategies of firms at the business standardization-adaptation dimensions, five mea-
unit level at any given point in time. More suring the concentration-dispersion dimensions,
precisely, each case should: and four measuring the integration-independence
dimensions of the featured firm's international
(1) be concerned with the strategic marketing issues
of a firm operating in an international environ- marketing strategy. Each scale item was an 11-point
ment; bipolar scale with extreme descriptors at both
(2) relate to an identifiable business unit for which ends.3 To ensure that the coders fully understood
a single, distinct marketing strategy is formu- the meanings of the coding dimensions, we also
lated and executed; provided the coders with detailed definition cards
for all of the coding scheme scale items and an
(3) report the key aspects of the firm's overall
international marketing strategy (as opposed instruction booklet that they could refer to as they
to being focused on a particular international undertook the coding task. Besides rating the scale
items, a coder attending to a given case was
marketing event or decision, or a particular
country strategy); and required to reference particular sections of the case
that contained information supporting his/her
(4) describe an actual, rather than a planned,
international marketing strategy that has been ratings. In undertaking the assignment of numer-
in implementation for a reasonable period of ical values along various dimensions based on case
time. narratives, the coder effectively replaced the role of
the key informant (as in a typical survey) on behalf
In accordance with these principles, we developed a of the featured firm.
case qualification checklist (see Appendix A) and We recruited and trained two successive groups of
stringently applied the criteria to our selection of student assistants to serve as coders in our project.
cases. Furthermore, for each case selected for All coders were provided with a detailed instruction
coding, we clearly specified the level of analysis at sheet, thoroughly briefed on the project require-
which the coder should interpret the international ments, and taken through several practice examples
marketing strategy of the featured firm (e.g., of item-coding procedures before being put on an
company, product division, or brand level corre- independent trial. After successfully completing the
sponding to a strategic business unit), as well as the trial and being debriefed on particular procedural
geographic scope of the firm's international mar- issues and techniques, the coders began a semester-
keting strategy (e.g., worldwide or regional). long coding job. Each week, the coders were
Our case selection process thus generated a assigned specific cases taken from our case pool.
sample of exemplars of international marketing Every case was coded by two coders, who subse-
strategies that existed at some point in time. To quently met to compare their item ratings and
the extent that companies chosen to be featured in resolve any discrepancies. We defined a discrepancy
published cases are usually notable industry players as a difference of more than two points on an
with a certain level of financial stability and 11-point scale between the two coders' ratings.
solvency, it can be argued that they are at least Whenever there was a discrepancy, the coders

507
Rescaled Distance Cluster Combine

concerned had to discuss their respective reasons
CASE 0 5 10 15 20 25
for the rating given, come to an agreement Label Num +---------+---------+---------+------------ -+
regarding the source of the difference (which was Daewoo 4

17
usually due to differences in the interpretation of Singer
P&G Europe 7
the text), and then voluntarily adjust their ratings R&A Bailey 37
Loctite 25
to within a two-point difference. Only in rare Schering AG 49
instances were the coders unable to successfully DHL

Hewlett Packard
9
34
resolve a discrepant item. In those rare instances, Nando's 38
42
we stepped in and arbitrated the differences by L'Oreal
PolyGram
Elseve
Classics
Brand
10
recoding the items ourselves. Haier

Citibank
40
32
The above steps executed over an 8-month period Mary Kay 35
resulted in a coded case sample of about 80 cases. Toyota

Bausch & Lomb
39
1
Many cases inevitably contained some items that Gallo Rice 2
31
could not be coded because of incomplete informa- Ben & Jerry's
Utex 47
tion in the case text. To reduce the number of Fike 33
Selkirk 44
missing values in our data set, we removed cases Microsoft 26
that did not have at least 10 out of the 16 items Montgras

BRL Hardy
28
16
coded. The resultant sample came to 51 cases. With AXA 21
Euro RSCG 46
this final sample, we executed Phase II of our Nestle 13
taxonomic procedure. ICI Paint 20
Ikea 8
Phase II of our procedure involved the prepara- Dendrite 18
tion and analysis of our data set. To prepare our Kikkoman 30

48
Sargan
data set for analysis, we first removed five items for Tesco 3
Carrefour 5
which ratings were not given in more than 30% of Sony Europa 6
the cases. For the remaining 11 items, we then Godiva Europe 51
23
Henkel
imputed all missing values with their respective Samsung 43
item means. We believed these two steps collec- PSA Peugeot

Zara
Citroen 50
14
tively helped us maintain the integrity of our data Jurassic Park

Disco
24
15
(by focusing only on commonly rated items with
Supermecados
Dunhill Holdings 36
fewer missing values) while preserving our sample Barco 22
12
Murphy Brown
size (by not having to perform listwise or pairwise WebEx 19
Singapore Airlines 29
deletion).4 Rochas 45
To statistically uncover archetypes of interna- Benetton 11
Stella Artois 27
tional marketing strategy from our data set, we Wal-Mart 41
utilized cluster analysis because of the technique's

Figure 1 Dendrogram from hierarchicalclustering procedure
'unparalleled ability to classify a large number of via Ward's method.
observations along multiple variables' (Ketchen
and Shook, 1996, 453). Following Punj and Stew-
art's (1983) recommendation, we applied a two-
stage procedure to cluster-analyze all 51 cases in the solution seemed acceptable. For example, from
data set along the 11 remaining dimensions. In the the dendrogram, we noted three relatively dense
first stage, the sample was subjected to hierarchical branches, indicating the presence of three 'natural'
clustering via Ward's method, which generally clusters (Ketchen and Shook, 1996). On the other
produces clusters whose centroids differ maximally hand, the statistics favored a two-cluster solution,
based on minimum within-cluster variance. At this with peak values of the pseudo-F,the CCC, and the
point, we relied on multiple criteria as suggested by agglomeration distance-change found at two
Milligan and Cooper (1985) and Ketchen and (followed by three) clusters. In deciding between a
Shook (1996) to determine the appropriate number two- and a three-cluster solution, we therefore
of clusters in our data set. Based on an inspection of considered the interpretability and meaningfulness
the dendrogram (see Figure 1) and an evaluation of of each solution (Hair et al., 1998). A two-cluster
the pseudo-F, the cubic clustering criterion (CCC), solution gave very little information, distinguish-
and the agglomeration distance-change statistics ing clusters as merely high or low along all
(see Table 2), either a two- or a three-cluster dimensions. Such a pattern also implied a high

SInternational marketing strategy archetypes LewisKSLimet al
508
Table 2 Pseudo-F, cubic clustering criterion (CCC), and agglom-

(see Figure2). We complement our analysis of these
eration distance-change statistics for possible cluster solutions
plots with what we call 'pseudo-t' statistics (see also
Statistic Numberof clusters Table 3), each calculated as the absolute value of the
ratio of the difference between each pair of centroid
2 3 4 5 values along each dimension to the estimated
Pseudo-F 14.49 10.33 8.67 8.11
standard error of this difference (cf. Lattin et al.,
CCC 5.003 3.061 2.066 1.953 2003). Largervalues of the pseudo-t indicate larger
Agglomeration 644.48 280.00 201.18 168.92 differences in centroid values with tighter distribu-
distance-changea tions.6
aEachcolumn refersto the number of clusters priorto the change.
Explanatorynotes on the statistics: ArchetypeA
The pseudo-F statistic is intended to capture the 'tightness' of clusters, This archetype follows a comparatively more
and is in essence a ratio of the mean sum of squares between groups to
the mean sum of squares within group (Lattin et al., 2003: 291). The standardized market offering policy. Companies
value reported is obtained from SASPROCFASTCLUS and is calculated as grouped under this archetype display, on average,
Pseudo-F=T - P/(G
higher degrees of standardization in product
- 1)
PcG/(n- G) design, advertising theme, and pricing as compared
where G is the number of clusters, Tis the total sum of squares, and PG is with the other archetypes (t> 2.584). On the
the within-groupsum of squares. Largernumbers of the pseudo-Fusually dimensions of brand name and sales promotion
indicate a better clustering solution.
The Cubic Clustering Criterion (CCC) was developed by SAS (Sarle, tactics, this archetype is also arguably more stan-
1983) as a comparative measure of the deviation of the clustersfrom the dardized than at least one other archetype
distribution expected if data points were obtained from a uniform
distribution.The criterion is calculated as (t, 3.346). Only on the dimension of channel
design is this archetype not evidently the most
CCC =In[1 - E(R2)]xK standardized (average item rating: 5.8). However,
where E(R) is the expected RL, RLis the observed RL,and K is the this archetype adopts a more concentrated market-
variance-stabilizing transformation (see Sarle, 1983). Larger positive ing value chain design, in terms of its product
values of the CCC indicate a better solution, as it shows a larger
difference from a uniform (no clusters) distribution. However, the CCC design and development (t> 2.349) and advertising
may be incorrect if clustering variablesare highly correlated. and promotional planning functions (t> 3.716). Its
The agglomeration distance refersto the distance between clustersbeing logistics and distribution planning function is also
merged in each step. A largerdistance indicates highly dissimilarclusters more concentrated than that of at least one other
being joined, suggesting that not joining them will preserve the natural
structure of the data. As distance-always grows, it is convenient to archetype (t=4.391). Moreover, this archetype
compare the incremental distance-changes, with a large change appears to be more integrated in its competitive
indicating that an appropriate solution has been found at the number
of clusters priorto the change (Ketchen and Shook, 1996: 446). process in terms of competitive decision-making
(t~ 4.655). On the dimension of communication
and mutual consultation across country units, this
archetype is evidently more integrated than one
correlation among all clustering variables, which other archetype (t=3.294). In view of its greater
we knew was not true. In contrast, a three-cluster degree of standardization, concentration, and inte-
solution appeared to offer richer insights into the gration, we label this archetype the Global Market-
possible ways various dimensions of international ers. The case that is most representative of this
marketing strategy could be combined, thereby archetype, based on the least sum of squared
illuminating subtler configurational patterns.5 For deviations from the cluster centroid, is Godiva
these reasons, we chose a three-cluster solution. Europe (Lambin, 2004), a producer of premium
Next, in the second stage of the clustering proce- chocolates. At the time of writing, Godiva's inter-
dure, the centroid values from the hierarchical national marketing strategy typified what many
clusters were used as seeds in an iterative K-means scholars would regard as a global or near-global
algorithm to recluster the observations into exactly strategy. The following quotes from the case
three clusters. This was to ensure reliable cluster illustrate the relatively more standardized market
groupings (Punj and Stewart, 1983; Ketchen and offering policy, especially in the areas of product
Shook, 1996). The final group assignments and design and advertising theme (average item ratings
centroid values are shown in Table 3. of 8.5 and 8.0, respectively):
The three clusters now represent three interna-
The Godiva facility in Belgium produces chocolates for the
tional marketing strategy archetypes. To interpret entire world, with the exception of the United States.
the clusters, we rely on plots of the centroid values Products exported from Belgium are identical for all

509
Different Same
brandname p brandname
in each market ineach market
.
Product . Same product
designed for * in each
each market I market
Localizedad Same advertising

theme for ... ....in each
each market market
Differentmix of Same sales

promotiontools promotiontools
in each market in each market
in
unique
Channelstructure
each market
structure
in
Same channel
each market
1
Prices set Same price
according to situation position
in each market in each market
Each country . Productdesign

responsible for product and development
design and development / in a single location
"
Each country Distributionand

responsible for logistics logistics consolidated
and distribution \ at a single location
Each country \ Advertisingand promotion

responsible for planningconsolidated
andadvertising.
promotion . . . . I at a single location
Each country I Marketingplanning

formulates tightly coordinated
own marketingplan I across all markets
..
Countrymanagers Countrymanagers
do not share . frequentlyshare
information [ information
...
0 1 2 3 4 5 6 7 8 9 10
Infrastructureminimalists
TacticalCoordinators..........
Global Marketers
Figure 2 Centroid values of three archetypes derived from case coding/clustering procedure.
countries, but sales by item are different... Today, the US The Belgian consumer is the reference point: 'Shouldn't a
factory still produces a slightly different and more limited product that has passed the test of the Belgian consumer, a
assortment of chocolate pralines. These differences will fine connoisseur of chocolate and a demanding customer,
progressively vanish, and the trend is toward similar be assured of success throughout the world?' (p. 332).
production (p. 332). Over the course of the past year van der Veken [Presidentof
Today, Godiva does not need to make itself known on the Godiva Europe, based in Brussels]had completely restruc-
international level: Its brand name is already globally tured the company. He startedby firing the marketing and
recognized. Its current concern, in line with the policy that sales staffand then changed the retaildistributionnetworkby
has been pursued for the past several months, is to create a removing Godiva'srepresentationfrom numerous stores. He
common advertising message for the entire world (p. 336). then completely rethought the decorationand design of the
remainingstores,and establishedpreciserulesof organization
Similarly, the quotes below illustrate the relative and functioning applicableto those stores (p. 336).
high geographical concentration of Godiva's mar-
keting value chain, notably the product design/ Events described in the case also suggest a tight
development and the distribution/logistics plan- worldwide (triadic) coordination in Godiva's mar-
ning functions (averageitems ratings of 8.5 and 7.5, keting planning and competitive decision-making
respectively): process (average item rating: 8.0).

Table 3 Clusterassignments and centroid values
Case Publisher/source Case number Industry Strategylevel
Cluster1: archetypeA
Barco HBS 9-591-133 Professionalequipment ProjectionSystems Division
Benetton HBS 9-396-177 Fashionapparel Corporation
Daewoo HBS 9-598-065 Diversified Automobile business
Dunhill Holdings CranfieldUniv. 588-002-1 Fashion and tobacco Corporation
Godiva Europe Kerinand Peterson (2004) - Chocolates Corporation
Henkel HBS 9-585-185 Diversified Adhesives Group
Loctite HBS 9-594-021 Adhesives and sealants Corporation
Murphy Brewery UCC 597-029-1 Breweries Murphy brands
JurassicPark HBS 9-596-014 Entertainment Licensing business
P&G Europe HBS 300-085-1 Consumer packaged Ariel Ultra brand
goods
PSAPeugeot Citroen Jain(2001) 20 Automobiles Corporation
Rochas CardiffBusiness School 594-006-1 Perfumes Rochas brand
Samsung ICFAI 503-055-1 Electronics Consumer electronics
Singapore Airlines HBS 9-504-025 Airlines Corporation
Singer HBS 9-804-001 Sewing machines Corporation
Sony Europa IMD IMD-5-0488 Electronics Consumer electronics
Stella Artois Ivey 9BOOA019 Breweries Stella Artois brand
Supermercados Disco HBS 9-599-127 Grocery retailing Disco supermarketchain
Utex MI 396-160-1 Industrialsealing devices Corporation
Wal-Mart ICFAI 304-1 38-1 General retailing Corporation
WebEx Stanford SM-121A Web-based communications Web conferencing products
Zara HBS 9-503-050 Fashion apparel Apparel stores
Cluster2: archetypeB
Bausch & Lomb HBS 9-594-056 Eye care/lenses Corporation
Ben & Jerry's Ivey 9A99A037 Ice cream Corporation
Citibank HBS/Wharton 9-395-142 Banking Corporation
Dendrite HBS 9-594-048 Sales automation systems Corporation
DHL HBS 9-593-011 Expressdelivery Corporation
Fike Jain(2001) 19 Industrialcontrol devices Corporation
Gallo Rice HBS 9-593-018 Rice production and Gallo brand
marketing
Haier ICFAI 304-264-1 Electricappliances Corporation
Hewlett-Packard HBS 9-501-053 Computers Home Products Division
Ikea ICFAI 303-112-1 Furnitureretailing Corporation
L'Oreal CEMS 501-011-1 Beauty products Elseve brand
501-012-1
Mary Kay HBS 9-594-023 Beauty products Corporation
Nando's WBS WBS-1999-4 Restaurants Corporation
Table 3 Continued
Case Publisher/source Case number Industry Strategy level
s
R&ABailey UCD 501-044-1 Wines and spirits Bailey's brand W
Schering AG BSE 303-221-1 Specialty pharmaceuticals Corporation W
Selkirk Ivey 9A99M003 Building materials Bricksbusiness A
Cluster3: archetypeC
AXA HBS 9-793-094 Insurance Corporation W
BRLHardy HBS 9-300-018 Wines and spirits Corporation W
Carrefour INSEAD 195-001-1 and 195-002-1 General retailing Corporation A
Euro RSCG IMD IMD-5-0573 Marketingcommunications Advertising business W
ICIPaints IMD GM-557 Paints Corporation W
Kikkoman HBS 9-504-067 Foods manufacturing Soy Sauce business W
Microsoft HBS 9-588-028 Computer software Microsoft Works program W
Montgras HBS 9-503-044 Wines and spirits Corporation W
Nestl HBS 9-585-013 Foods manufacturing Culinary products W
Polygram Classics HBS 9-598-074 Recorded music Corporation W
Sargan plc CardiffBusiness School 594-047-1 Health care products Consumer brands division W
Tesco HBS 9-503-036 Retailing Tesco chain W
Toyota ICFAI 304-100-1 Automobiles Corporation W
Dimension Cluster1 (archetypeA) Cluster2 (ArchetypeB) Cluster3 (ArchetypeC) 'Pseudo-t'statistic
n=22 n= 16 n= 13
Centroid Std dev Centroid Std dev Centroid Std dev A vs B
Brand name standardization(ST_BRAND)8.59 1.56 8.34 1.83 6.08 2.49 0.407 3.

Productdesign standardization(ST_DESIGN)7.48 2.11 4.25 3.11 3.85 3.00 3.656 3.
Advertisingtheme standardization 6.67 1.87 3.80 1.71 4.65 1.79 4.856 3.
(ST_ADTHEME)
Sales promotion tactics standardization 5.96 2.00 3.97 1.65 5.68 1.64 3.346 0.
(ST_PROMO)
Channel design standardization 5.79 1.90 6.15 1.79 2.92 1.32 0.627 4.
(ST_CHANNEL)
Pricing standardization(ST_PRICING) 6.45 2.1 3 4.21 2.41 4.28 2.80 2.843 2.
Product design and development 8.23 1.22 6.78 2.34 5.27 2.15 2.349 4.
concentration (CONC_PDM)
Logistics and distribution planning 6.51 2.23 5.45 2.31 3.18 1.87 1.496 4.
concentration (CONC_LOGDIST)
Advertising and promotional planning 7.79 2.27 4.43 2.28 4.80 2.37 4.449 3
concentration (CONC A&P)
Competitive decision-making integration 7.79 1.60 4.24 1.70 5.00 1.91 6.301 4.
(INT_DECISION)
"? International marketing strategy archetypes LewisKSLimet al
512
ArchetypeB the fact that the company uses local subsidiaries

In contrast to the Global Marketers,this archetype and agents in different markets, the competitive
pursues a rather mixed standardization policy for decision-making process across countries did not
its market offering. While its brand name and appear to be tightly coordinated (average item
channel design elements are arguably more stan- rating: 3.0). On the other hand, production was
dardized (similar to the Global Marketers), its concentrated in four countries - Italy, Germany,
advertising theme and sales promotion tactics are Argentina, and Uruguay - with apparently centrally
relatively more localized as compared with at least controlled product design and development func-
one other archetype (t> 2.529). Along with Arche- tion (average item rating: 7.0):
type C below, its product design and pricing Focusedon the production of value-addedrice, F&PGruppo
dimensions are also more localized than those of described itself as 'the rice specialist' and was one of only a
the Global Marketers (t> 2.843). Similarly, its few companies in the world involved in the entire process,
advertising and promotional planning function is from growing and milling to the packaging and marketing
more geographically dispersed as compared with of brand rice. The company added value through research
and development of new and improved strains of high-
the Global Marketers (t=4.449). Nonetheless, its
quality rice, proprietary manufacturing processes, and
product design/development and distribution/ packaging... A high percentage of the resulting profits
logistics planning functions are moderately con- were, in turn, reinvested in research and development
centrated (average item ratings of 6.8 and 5.5, (p. 1).
respectively, in between the other two archetypes).
Compared with at least one other archetype, this
archetype is also apparently less integrated in its ArchetypeC
competitive decision-making and communication Like the InfrastructuralMinimalists, this archetype
and mutual consultation (t,>3.143). In view of its adopts a mixed standardization policy, but in a
selective approach of standardizing only the brand different way. Although it is moderately standar-
name and channel design with corresponding dized in the area of sales promotion tactics (average
concentration of product design/development and item rating: 5.7), its product design, advertising
distribution/logistics functions, we label this arche- theme, pricing policy, and especially channel
type the InfrastructuralMinimalists. Companies design are more localized as compared with at least
classified under this archetype appear to emphasize one other archetype (t> 2.584). And, although its
the provision of global infrastructureto their local average brand name standardization rating is not
units with otherwise minimal intervention in the low (at 6.1), this rating is comparatively lower than
respective local operations and decisions. A case those of the other two archetypes (ta 3.166). Its
that is representative of this archetype, again based marketing value chain activities, especially the
on the least sum of squared deviations from the product design/development and distribution/
cluster centroid, is Gallo Rice (Laidler, 1998), logistics functions, are also more dispersed
featuring the Italian company F&PGruppo, market- (t->2.160). Likewise, its advertising and promo-
er of the Gallo brand of rice. According to the case, tional planning function is clearly more dispersed
the Gallo brand name was used by the company than that of the Global Marketers (t=3.716). But
across all country markets, resulting in a brand perhaps how this archetype truly differs from the
name standardization average rating of 9.0: Infrastructural Minimalists is that it is compara-
The Gallo brand name and Gallo rooster logo were used tively more integrated in its competitive process in
consistently across geographic markets... (p. 2).
terms of communication and mutual consultation
(t=3.143), even while its competitive decision-
However, the exact line of rice sold differed from making process does not appear to be particularly
country to country (average product design stan- integrated. Because the emphasis of its interna-
dardization rating: 3.0). Similarly, the communica- tional marketing strategy is in the coordination
tion strategy differed somewhat across country (namely, mutual consultation) of tactics (namely,
markets (average advertising theme and sales sales promotion tactics) rather than in the standar-
promotional tactics standardization ratings: 4.5 dization of tangible elements such as product
and 5.5, respectively). Nonetheless, most sales were designs and channel design, or in the concentra-
made to retailers rather than to institutions (aver- tion of marketing functions, we label this archetype
age channel design standardization rating: 7.0). But the Tactical Coordinators. A case that is representa-
perhaps because of the nature of the product and tive of this pattern is AXA: The Global Insurance

International marketing strategy archetypes LewisKSLimet al ?
513
Company (Goodman and Moreton, 1995). Like markets (average sales promotional tactics standar-
many other multinational insurance firms, AXA dization rating: 7.5).
pursued a rather localized approach to marketing In summary, each of the three derived archetypes
its products, owing to heavy local regulation of exhibits a distinctive configuration of market
insurance sales: offering, structural/organizational, and competitive
Insurancewas one of the most heavily regulated industries process. The configurations are somewhat complex,
in the world, with wide variation in the degree and scope of with no clear-cut correlations among the strategy
regulation. In general, all countries required that insurers dimensions, which means that they would not
obtain a license to sell insurance... In addition, insurance have been well captured using unidimensional
regulations often stipulated accounting methods and scales or an aggregate score alone.
reporting requirements... Insurerswere also limited in the To corroborate the above observation, we per-
types of assets in which they invested premiums. In a few formed a descriptive canonical discriminant analy-
countries, regulatorsalso strictly controlled premium rates
and contract terms (p. 6). sis based on the cluster groupings. Figure 3a shows
the positions of the various cases and their
Moreover, AXA'sinternational expansion in the late respective cluster centroids relative to two canoni-
1980s and early 1990s occurred mainly through cal discriminant functions or axes. The horizontal
acquisitions, resulting in a complex holdings axis (Function 1), which discriminates the Global
structure with multiple registered company names
in different countries. Hence the average standar-
dization item ratings for product design, channel a K-meanscluster
design, and pricing ranged from a low of 2.5 to 3.0. 0 ClusterA
The concentration of marketing value-chain activ- 4- 0 ClusterB
A ClusterC
ities was also apparently rather low (e.g., with an o [E GroupCentroid
average distribution and logistics planning concen- 0 o
tration rating of 3.0), presumably because of the 2-

bC pster B
lack of scale economies in sales and marketing o 0 0
0o
activities: 0 g
U0 0 0 A o
Clus-Wr
Typically,sales and marketing, along with claims adjusting,
represented the two largest cost factors for an insurance a 0
o
company, followed by underwritingand asset management. a Clustg•r o
-2 - a o
Of these four activities, only claims adjusting and asset
management demonstrated any scale economies. Under- B A
writing and sales were labor intense and generally varied in

direct proportion to the volume of premiums written (p. 3). -4 -2 0 2 4
Function 1
However, what stood out in the AXA case was a b

moderate-to-high level of competitive process inte- 0.8
gration, with average competitive decision-making

and communication/mutual consultation ratings 0.6 ST CHANNEL
of 5.0 and 7.5, respectively. The following quote ST BRAND PDM
0.4 -CONC
illustrates this pattern: CONC
LO•_(T
Still, Bebear'sobjective was to make AXAnot just a French 0.2
RI
company with foreign subsidiaries, but a truly global
organization that would draw on the strengths of every 0
.2 ~
0.3
CdNC A&P
0.4 0.5 0.6 0.7 0.8
country in which it operated. To implement this objective,
he created the Strategy Committee, which would set -0.2
- ST DESIGN
corporate objectives and oversee their implementation.
ADTHEME
Apart from Bebear,the group included five senior French -0.4INT CONSULT
04ST
INTDECISION
executives.., and five representativesof AXA'smajor foreign
operations... The company also began to put into place a -0.6
common MIS system.., to improve the quality and
comparabilityof information in the company (p. 13). Figure 3 Plots from canonical discriminant analysis of cluster
groupings:(a) positionsof cases and centroidsrelativeto two
As a result, some of AXA's sales promotional tactics discriminant functions; (b) unstandardized coefficients for
could have been well coordinated across country Function 1 (horizontal) and Function 2 (vertical).

) International marketing strategy archetypes LewisKSLimet al
514
Marketers from the other two archetypes, is posi- approach to managing their marketing operations,
tively related to all of the 11 strategy dimensions such as minimal or diminishing cultural differences
used in the cluster analysis (see Figure 3b). This in purchase and consumption behaviors, the lack of
appears to be a broad-spectrum function that or disappearing peculiar local regulations, scale
behaves like Zou and Cavusgil's (2002) aggregate economies in marketing value-chain functions,
GMS factor. Cases rated collectivelyhigh on the 11 and the presence of 'global' competitors that
dimensions would tend to be classified under the required the shared attention of all subsidiary
Global Marketersarchetype. More interesting, how- managers. Similar to the construct of 'external
ever, is the vertical axis (Function 2), which, unlike globalizing conditions' defined by Zou and Cavus-
the GMS factor, serves to further discriminate gil (2002), such conditions encourage, or even
between the Infrastructural Minimalists (more necessitate, high levels of market offering standar-
positive) and the Tactical Coordinators (more dization, value-chain concentration, and competi-
negative). This function is positively related to tive process integration. For example, the Sony
brand name standardization, channel design stan- Europa case (Kashani and Kassarjian, 1998)
dardization, product design and development con- describes how Sony's European unit was grappling
centration, and distribution and logistics planning with a set of market forces, including the increasing
concentration, all of which characterize the Infra- consolidation of competitors and buyers. Along
structural Minimalists archetype when highly with the potential for greater cost savings from
rated. This function is also negatively related to streamlining operations across markets and the
sales promotional tactics standardization and inte- desire to unify the representation of regional
gration of communication and mutual consulta- interests, this led to the creation of a pan-European
tion, both of which are characteristic of the marketing organization (treated as a global market-
Tactical Coordinators archetype when highly ing strategy when the geographic scope of analysis
rated. Thus, when a firm's strategy does not is restricted to the region).
clearly belong to the Global Marketer archetype Second, an examination of the cases classified as
(with high values on all dimensions), it will the InfrastructuralMinimalists suggests that many
fall under either the Infrastructural Minimalists of these companies were involved in product
or Tactical Coordinator based upon the configura- businesses where consumer tastes and preferences
tion of values it has on the determinants of differed significantly across countries. Several food
Function 2. companies fall into this cluster, including Gallo
Rice, Ben & Jerry's (an ice-cream producer), and
Evolutionary drivers of archetypes Nando's (a fast-food chain), as do a couple of
Having found evidence of the existence of three beauty/grooming products companies, namely
distinct archetypes of international marketing Mary Kay and L'Oreal. In these businesses, there
strategy, it is appropriate at this point to ask how might also be certain local regulations concerning
the archetypes came about: that is, whether there the production and sale of the product. Thus some
might have been any common exogenous forces of the advantages of standardizing the market
imposed on firms belonging to each archetype, offering might not be relevant to these firms.
forces that have driven their similarity through one Furthermore, perhaps because of the need to deal
or more of the evolutionary mechanisms described with local competitors, a worldwide coordination
by Miller (1981, 1986) - namely, environmental of competitive processes might not be necessary.
selection, mimetic action, and firm strategic Nevertheless, there might still be substantial brand
choice. We tried to answer this question by equity effects associated with these businesses, as
returning to the cases to look for commonalities demonstrated in the Ben & Jerry's case (Hagen,
within clusters. In keeping with the configurational 1999), where the company frequently used licen-
thinking, we sought to identify holistic sets of sing mode to enter foreign markets. Also, there
exogenous factors as drivers. Because of the quali- could be established channel and distribution
tative nature of this step, we were limited in our strategies that have worked very well in these
analytical precision. Yet we noted three interesting industries, as demonstrated by Mary Kay's (Laidler,
patterns. 1996) frequent use of its direct selling method.
First, an inspection of the list of cases classified as These companies therefore could have found it
the Global Marketers reveals that many of these advantageous to provide a minimal brand and
businesses faced conditions favoring a global channel infrastructure for their international mar-

)
International marketing etrategy archetype LewisKSLimet a
515
keting operations, but otherwise leave the major Archetype performance potential: subsidiary
marketing decisions to local managers. network contingent fit considerations
Third, a look at the cases typifying the Tactical After considering the possible drivers of the differ-
Coordinators points to the possibility that many of ent international marketing strategy archetypes,
these companiesnot only did not experienceany the next question one might pose is: 'Which
market forces that encouraged standardizing their archetype(s) is/are liable to lead to stronger multi-
market offering in a major way, but also did not see national firm performance?'7This is an important
substantial scale economies in consolidating any question because, arguably, the ultimate concern
part of their marketing value chains. The AXA case for the study of any area of strategy is to explain
mentioned earlier illustrates this possibility. None- firm performance variations (Rumelt et al., 1994).
theless, the similarity of competitive environments, (Here, of course, we take the position that perfor-
including the presence of shared competitors across mance variations among the archetypes are still
country markets, could have motivated these likely, even though it has been stated earlier that
companies to coordinate their competitive deci- each of the archetypes is potentially high perform-
sions and tactics on a worldwide scale. Forexample, ing.) However, this is also a difficult question
several companies, such as Nestle, Euro RSCG,and because the relationship between strategy and
ICI Paints, faced global competitors even as they performance is itself a complex issue. In particular,
operated in local country environments. It might the implementation context within which a firm
therefore have been critical for these companies to operates may influence the effectiveness of any
have high degrees of communication and mutual given strategy (Walker and Ruekert, 1987; Noble
consultation as well as harmonization of competi- and Mokwa, 1999). The implementation context
tive tactics across country units. imposes peculiar constraints upon the firm, so firm
In summary, a combination of environmental performance is partly a function of how unencum-
factors and market forces appears to drive the type bered the strategy is by those constraints. An
of international marketing strategy adopted by any important aspect of a multinational firm's imple-
given multinational firm. mentation context is its network of international
subsidiaries,"given that a significant proportion of
Pla: A configurationof environmentaland marketfactors a multinational's marketing activities is executed
that createsincentivesfor a globalapproachto marketing
(whether or not planned) at the local subsidiary
management,including a high degree of similarityin level (Birkinshaw and Morrison, 1995; Solberg,
customer tastes and preferencesacross countries, the
absenceof local regulations,the presenceof scale econo- 2000). Research suggests that subsidiary-level fac-
mies in operatingmarketingvalue-chainactivities,and the tors have the potential to either help or hurt the
presenceof global competitors,is likely to engendera effective implementation of a multinational's
strategy resembling the GlobalMarketersarchetype. intended strategy (e.g., Wiechmann and Pringle,
1979; Hulbert et al., 1980; Hewett and Bearden,
Plb: A configuration of environmental and market factors 2001). Therefore, in discussing the performance
that does not particularly encourage a global marketing
potential of any international marketing strategy
management approach, including different customer tastes
and preferences across countries, existing local regulations, archetype, the characteristics of the subsidiary
and the need to deal with local competitors, but that at the network need to be considered. In other words,
same time encourages sharing a common brand and rather than simply order the relative superiority
channel infrastructure,such as when there are global brand of the different archetypes, it is necessary to
recognition and established distribution strategies, is likely examine the contingent fit (Doty et al., 1993)
to engender a strategy resembling the Infrastructural Minim-
alists archetype.
of each archetype with the subsidiary network
characteristics.
There are potentially several different subsidiary
Plc: A configuration of environmental and market factors
network characteristics to be considered. These
that does not particularly encourage a global marketing
management approach, including different customer tastes include the scope of subsidiary responsibilities,
and preferencesacross countries, existing local regulations, the level of autonomy held by each subsidiary,
and the absence of scale economies in marketing opera- the degree of subsidiary dependence on the head
tions, but that at the same time encourages the coordina- office, and the degree of interdependence among
tion of competitive decisions and tactics across markets,
such as the commonality of competitive environments and subsidiaries. One way to incorporate the effect of
shared competitors, is likely to engender a strategy these characteristics into the prediction of arche-
resembling the TacticalCoordinatorsarchetype. type performance is to model each characteristic as

516
a moderator of the archetype-performance rela- has a wide geographic scope but a narrow
tionship. Statistically, this involves analyzing the functional scope of responsibilities. The degree of
interaction between each characteristic and the strategic autonomy held by each subsidiary is
various strategy variables that make up an arche- moderate, and subsidiaries are highly dependent
type. This is referredto as the interaction approach on the head office as well as on one another for
to studying contingent fit (Drazin and Van de Ven, materials and resources. Because each subsidiary
1985). However, as Miller (1981) and Meyer et al. contributes to the overall functioning of the multi-
(1993) point out, such a 'reductionist' approach is national in a specialized manner, this subsidiary
generally aimed at isolating the effects of single network type is labeled the 'Specialized Contribu-
contingency variables, and often unrealistically tor'. In relation to previous literature, this type of
assumes unidirectional linear relationships among network resembles Bartlett and Ghoshal's (1986)
the variables. From the perspective of configura- 'Contributor', Jarillo and Martinez's (1990) 'Recep-
tional theory, this approach is not ideal for under- tive', and White and Poynter's (1984) 'Rationalized
standing holistic combinatorial patterns of fit Manufacturer'. Such a subsidiary network is likely
between multiple strategy variables and multiple to impose constraints on the implementation of
contingency factors. An alternative approach that either an InfrastructuralMinimalist or a Tactical
is more consistent with the configurational think- Coordinator strategy archetype, because each sub-
ing underlying this research is to identify config- sidiary does not possess a full range of functional
urations of contextual variables that fit well with (marketing) expertise to effectively determine a
the respective strategy archetypes. Like the arche- local mixed adaptation strategy, even though it
types, each of these contextual configurations is a possesses moderate strategic autonomy. Moreover,
multidimensional combination of distinct charac- the rationalized global design does not seem to be
teristics. This way, multiple subsidiary network compatible with the relatively dispersed marketing
characteristics serving as contextual contingencies value-chain structure of either of the two men-
can be simultaneously considered. Strategy arche- tioned strategy archetypes. On the other hand,
types and contextual contingencies that fit well this subsidiary network type seems to support a
with each other can be viewed as extended config- Global Marketer strategy, because (a) the wide
urations of mutually supportive elements, similar geographic and narrow functional scope of sub-
in spirit to Miller's (1986) matching of compatible sidiary responsibilities facilitate the creation of a
configurations of strategy and structure. At the standardized market offering, (b) the activity
same time, when there is a lack of unique one-to- specialization design accommodates a concentrated
one fit between strategy and context, equifinality marketing value chain, and (c) the moderate
can be said to exist. autonomy and high interdependence among sub-
While the lack of data prevents a fresh taxonomic sidiaries are conducive for achieving integration in
discovery of common configurations of subsidiary the competitive decision and marketing planning
network characteristics along with an integrated process.
testing of fit with the strategy archetypes earlier
uncovered, research on multinational subsidiary The local implementer
behavior has previously identified three configura- A second type of subsidiary network is character-
tions of subsidiary roles and structural contexts ized by both limited geographic scope and low
(Birkinshaw and Morrison, 1995). Because strategic autonomy assigned to each local subsidi-
these 'subsidiary roles' and 'structural context' ary (Birkinshaw and Morrison, 1995). This type of
variables appear to broadly describe the nature of network also exhibits a high degree of subsidiary
a multinational subsidiary network, the three dependence on the head office, as well as high
configurations can be used here to analytically interdependence among subsidiaries for materials
examine the contingent fit of the three uncovered and resources. Because each subsidiary is essentially
archetypes. tasked to execute the strategy formulated by the
head office, such a subsidiary network is labeled the
The specialized contributor 'Local Implementer'. This type of network closely
As described by Birkinshaw and Morrison (1995), corresponds to Bartlett and Ghoshal's (1986)
one type of subsidiary network is characterized by 'Implementer', D'Cruz's (1986) 'Branch Plant', and
high levels of subsidiary expertise in specific White and Poynter's (1984) 'Miniature Replica'
functions or activities. Specifically, each subsidiary subsidiary types. Such a subsidiary network is

International marketing strategy archehypes LewisKSLimet al "?
517
unlikely to facilitate a Global Marketer strategy, archetype may fit well with this subsidiary network
given the lack of activity specialization on a type, creating an equifinal situation. In the case of
worldwide scale and the lack of strategic integra- an Infrastructural Minimalist strategy, country
tion across subsidiaries. Also, because the subsidi- subsidiaries can share certain infrastructural mar-
aries' marketing strategy formulation capabilities keting strategy elements (e.g., brand name and
are likely to be limited (because of the historical channel structure) that do not interfere with the
lack of strategic autonomy), this subsidiarynetwork development of differential product mandates
type seems incompatible with the Tactical Coordi- across subsidiaries. In the case of a Tactical Coor-
nator strategy,which requires a high degree of local dinator strategy, subsidiaries merely consult regu-
responsiveness. In contrast, the implementation of larly with one another to coordinate their compe-
an Infrastructural Minimalist strategy seems less titive tactics (e.g., sales promotions) without under-
encumbered. Even though the Infrastructural mining each subsidiary'sproduct-market authority.
Minimalist strategy also requires some degree of In summary, the three archetypes may be differ-
local initiative in the design of market offering entially effective given different implementation
and competitive decision-making, the level of contexts, depending upon their contingent fit
corporate support in terms of branding, channel within each context (see Figure 4):
structure, and product development is greater.
Individual subsidiaries could take advantage of P2a: When implemented within a subsidiary network
such infrastructure to build their local marketing context that is characterized by a wide geographic but
expertise. narrow functional scope of subsidiary responsibilities,
moderate subsidiary autonomy, high dependence on the
head office, and high interdependence among subsidiaries,
The world mandate the contingent fit of the GlobalMarketersarchetype will be
A third type of subsidiary network is characterized the strongest, and hence its performance potential will be
by the delegation of worldwide/regional responsi- the greatest among the three known archetypes.
bilities (or 'mandate') for entire product businesses
to individual subsidiaries (Birkinshaw and Morri- P2b: When implemented within a subsidiary network
son, 1995; Birkinshaw,1996; Birkinshawand Hood, context that is characterizedby a narrow geographic scope
1998). Each subsidiary has almost complete strate- of subsidiary responsibilities, low subsidiary autonomy,
gic autonomy and functional control over its high dependence on the head office, and high interdepen-
mandated business, and is not dependent on the dence among subsidiaries,the contingent fit of the Infra-
structuralMinimalistsarchetype will be the strongest, and
head office or other subsidiaries for materials and hence its performancepotential will be the greatest among
resources associated with that business. This type of the three known archetypes.
subsidiary network is referred to as the 'World
Mandate' (Birkinshaw and Morrison, 1995), and
P2c: When implemented within a subsidiary network
closely corresponds to Bartlettand Ghoshal's (1986) context that is characterized by a wide geographic and
'StrategicLeader',Gupta and Govindarajan's(1991) functional scope of subsidiary responsibilities, high sub-
'Integrated Player', and Roth and Morrison's (1992) sidiary autonomy, low dependence on the head office, and
'Global Subsidiary Mandate'. Because subsidiaries low interdependence among subsidiaries,the contingent fit
in this type of network have extensive management of both the InfrastructuralMinimalistsarchetype and the
TacticalCoordinatorsarchetype will be equally strong, and
capabilities and knowledge in at least one product their performancepotentials will be greaterthan that of the
business, their managers are unlikely to be recep- GlobalMarketersarchetype.
tive to a Global Marketer strategy for the entire
range of the multinational firm's products. Each Note that strong contingent fit within each context
subsidiary's managers may see unique linkages does not mean a complete lack of encumbrances or
among the various products, and hence be moti- constraints that may be posed to the strategy.
vated to customize the market offerings for their Therefore, archetype performance may not always
respective country market. At the same time, be optimized (as is likely with the Local Implemen-
because individual subsidiaries carry a full range ter network situation described above). In Gresov
of functional expertise but not necessarily product and Drazin's (1997) terms, this is possible because a
category expertise in all categories, some degree of given context can place multiple, conflicting
resource-sharing or cross-market consultation may demands on the organization, rendering all orga-
be desirable. Consequently, either an Infrastructur- nizational strategies suboptimal in some ways.
al Minimalist or a Tactical Coordinator strategy However, when the subsidiary network context

518
S-- --------------i-.
5\ I, \I I
GlobalMarketers '' InfrastructuralMinimalists TacticalCoordinators
II II I
II II I
V-- -- -- -- -- -
I - - - -
,
"---"- ---
-"-
SpecializedContributor Local Implementer I World Mandate

* Wide geographicbut narrow * Narrowgeographicscope of * Wide geographicandfunctional
functionalscope of subsidiary I subsidiaryresponsibilities scope of subsidiary
responsibilities I Low subsidiaryautonomy responsibilities
* Moderatesubsidiaryautonomy I I High dependenceon headoffice High subsidiaryautonomy
* High dependenceon head office High interdependenceamong I Low dependenceon headoffice
* High interdependenceamong I subsidiaries * Low interdependenceamong
subsidiaries I subsidiaries
j II \ I I/
t- *I- --_________________
- - - ---
Dual-directional arrow denotes a contingent fit between an archetype and a configuration of subsidiary network characteristics.
Contingent fit implies strong performance potential. Multiple instances contingent fitfor a given configuration of subsidiary
network characteristics indicate an equifinal situation (Gresov and Drazin, 1997).
Pattern of characteristics within dotted area may be viewed as an extended configuration of mutually-supportive elements. This can be
treated as a contingent ideal profile (Doty et al., 1993) and tested using a profile deviation methodology.
Figure4 Contingentfit betweenarchetypesand configurations

of subsidiarynetworkcharacteristics.
facilitates the implementation of more than one alike a useful vocabulary with which to describe the
strategy archetype, equifinality is present (as is very concept of 'international marketing strategy'.
likely with the World Mandate network situation However, this paper also puts forth the argument
described above). This is because alternative strate- that, in order to advance our body of knowledge,
gies are capable of meeting different subsets of the the study of international marketing strategy
contextual demands (Gresov and Drazin, 1997). should also take on a holistic, configurational view
and not be confined to a single characterization -
Conclusion and future directions be it standardization-adaptation, concentration-
A fundamental element in the development of a scientific dispersion, or integration-independence - or be
reliant on a single index of 'globalness' (e.g., Zou
body of knowledge is the availability of a widely accepted
and usable classification scheme... (McKelvey,1975, 509). and Cavusgil, 2002). A configurational approach
provides tremendous utility not only in terms of
This paper began with the premise that the mean- the richness of description, but also in its capacity
ingful characterization and classification of inter- to guide scholarly and managerial thinking about
national marketing strategies are essential to the multifaceted nature of international marketing
understanding the various strategic options that a strategy. In adopting this approach, scholars and
multinational firm has in structuring its marketing managers should consider how the various strategy
approaches across different country markets. dimensions combine or align themselves into
Indeed, the field has, over the years, attained a gestalt patterns or modal archetypes.
good appreciation of the nature of international To substantiate our argument, we have conducted
marketing strategies simply by knowing how they a novel case coding/clustering study to uncover
can differ. Beginning with Buzzell (1968), scholars distinct combinatorial patterns along 11 theory-
have tried to describe the international marketing based dimensions of standardization-adaptation,
strategies of multinational firms using such cate- concentration-dispersion, and integration-inde-
gory labels as 'standardized' vs 'adapted' (e.g., Jain, pendence. We found three archetypes within a
1989). Later scholars have also invoked the cate- sample of exemplars of international marketing
gories of 'concentrated' vs 'dispersed' (e.g., Craig strategy. While one of these archetypes (the Global
and Douglas, 2000) and 'integrated' vs 'indepen- Marketers) could well have been captured using a
dent' (e.g., Hamel and Prahalad, 1985). These labels single aggregate globalness score (e.g., the GMS,
have given researchers, students, and practitioners Zou and Cavusgil, 2002), the other two archetypes

519
were detected mainly because we have used a national marketing strategies. Pedagogically, they
configurational procedure, which discriminated can facilitate a more in-depth discussion of the
between different mixed patterns of globalness. different types of international marketing strategy,
The detection of these archetypes shows that especially when illustrated with representative
strategic marketing options available to multina- teaching cases. Furthermore, by linking them to a
tional firms can be more richly described and relevant set of evolutionary forces, and raising the
meaningfully compared when a holistic, multi- issue of their contingent fit with certain subsidiary
dimensional perspective that unifies the three network characteristics, a more insightful dialogue
separate characterizations is taken. about causal processes in the international market-
After uncovering the archetypes, a further step to ing arena can be encouraged.
synthesize our knowledge of international business With the above conclusions in mind, an agenda
and international strategy necessarily includes the for the further study of international marketing
identification of evolutionary drivers of the respec- strategy archetypes and examination of their
tive archetypes and an assessment of potential drivers and performance variations is hereby pro-
performance variations among the archetypes. In posed. A crucial first step in this agenda would be to
our study, several archetype drivers were induced systematically validate the archetypes by gathering
from common environmental factors and market additional evidence other than from published
forces within case clusters. Archetype performance cases. This would involve the measurement of
potential, on the other hand, remains a more international marketing strategies using a variety
complex issue involving the consideration of con- of alternative sources (e.g., archival records, inter-
tingent fit between each archetype and the imple- views, surveys). In this regard, the diverse data
mentation context. As we are focusing on collection approaches of Miller and Friesen (1980),
marketing strategies of multinational firms, we Doty et al. (1993), and Ketchen et al. (1993), among
propose that configurations of subsidiary network others, could be borrowed. Alternative procedures
characteristics represent important implementa- for quantifying the dimensions of international
tion contexts. Indeed, subsidiary-level issues are marketing strategy should also be explored. Cluster
increasingly gaining attention within the interna- analysis, or other statistical techniques, could then
tional business literature (Ghoshal and Bartlett, be performed on the data to derive modal patterns
1990; Nohria and Ghoshal, 1997; Rugman and of international marketing strategy. To ensure
Verbeke, 2001). More significantly, as international reliability, the derived archetypes should be cross-
marketing activities rely on the efforts of local validated with results from multiple data sets as
subsidiaries to occur, it is reasonable to expect the well as from different statistical procedures.
type of subsidiary network to either facilitate or After a robust set of archetypes has been estab-
hinder the effective implementation of any given lished, the next step in the research agenda would
international marketing strategy. In the final be to systematically examine the observations
analysis, whether strategy archetypes result in grouped under each archetype to analyze their
superior or equifinal performance would depend common evolutionary drivers. Perhaps it is the
on the extent to which their implementation is environmental forces that dictate the adoption of a
unencumbered by the existing subsidiary network particular type of strategy. Perhaps the historical
characteristics. development or conventional practices of their
The new holistic characterization of international respective industries have somehow favored the
marketing strategy as archetypes and the added same kind of strategy. Or perhaps the competitive
understanding of archetype drivers and perfor- and social interactions among different players
mance potential serve not only to advance the within a network of firms have encouraged the
theory of international marketing, but also to sharing of knowledge and preferences with respect
provide a multifaceted perspective for teaching to strategy design. In Miller's(1987) terms, different
the subject. To that end, our study has contributed 'imperatives' or causal forces can give rise to
three archetypes of international marketing strat- configurational archetypes. These possible under-
egy that would not have been understood using lying causes should be thoroughly investigated
any of the traditional textbook descriptors (e.g., from multiple theoretical perspectives. In addition,
'standardized', 'concentrated', 'integrated', etc.) methods such as comparative case studies and
alone. Yet these archetypes represent modal combi- longitudinal tracking of strategy profiles for a panel
natorial patterns found among exemplars of inter- of firms may be utilized for empirical evaluations.

; LewisKSLimet al
International marketing strategy archetypes
520
Finally, after archetypes have been identified and Notes

their antecedents understood, the performance
1In jain
potential of different archetypes would be the main (1989), a brief section was devoted to
subject of interest. Direct comparisons of perfor- pointing out that the variousmarketingmix elements
mance indicators across archetypes are unlikely to (product design, brandname, packaging,etc.) can be
result in any meaningful findings, because, as differentiallystandardized. Nevertheless, the author
discussed, the relationship between international did not discuss the potential implications of such
marketing strategy and multinational firm perfor- differential standardization levels. Moreover, the
mance is likely to be contingent upon the imple- author states, 'Conceptually, standardizationof one
mentation context. Traditional statistical testing or more partsof the marketingprogramis a function
of strategy-performance moderators involving of five [antecedents]...' (Jain, 1989, 71), implying
interaction terms may also not be appropriate either that standardization of one part of the
because of the need to consider multiple contextual marketingprogramsufficientlyqualifiesas a 'standar-
contingencies. As illustrated above, the matching dization' strategy, or that standardizationsof different
of compatible configurations of strategy and parts of the marketing program are conceptually
implementation contexts from a contingent fit equivalent.
perspective (Miller, 1986; Doty et al., 1993) may 2The eight first-order dimensions were product
prove to be a more efficacious approach to under- standardization,promotion standardization,standar-
standing archetype performance variations. To dized channel structure,standardizedprice, concen-
seriously pursue the contingent fit approach, richer tration of marketing activities, coordination of
configurations of implementation contexts marketingactivities,global market participation,and
based on a comprehensive set of subsidiary integrationof competitive moves.
network characteristicsand other relevant variables 3Allscale items were developed through extensive
should be explored from the same data set(s) consultations among the authors based on existing
used to generate the strategy archetypes. Compa- definitions of strategy dimensions in the literature
tible matched pairs of these configurations (e.g., Zou and Cavusgil,2002) and with the inputs of
with the respective archetypes can then be treated colleagues who had knowledge of international
as extended configurations and tested for contin- marketing. It is important to note that our coding
gent fit using a profile deviation methodology scheme scale is not comparablewith a typical survey
(Drazin and Van de Ven, 1985; Gresov, 1989), scale. Even though each strategy dimension was
which analyzes the effect of deviation (measured captured by only a single scale item, the process of
in Euclidean distance) from each extended completing each scale item on our coding scheme was
configurational profile. As part of this assessment, a painstakingone involving a thorough reading of a
different forms of equifinality among the case by the coder concerned, a search for relevant
archetypes (Gresov and Drazin, 1997) can also be passages in the case to support the rating given, and
tested. (if necessary) a discussion with the other coder to
resolve any discrepanciesin each other's ratings.We
therefore believe a single item scale was sufficiently
Acknowledgements reliablefor each dimension.
We gratefullythank the IndianaUniversityCenter for 4Admittedly,by removing the five items with high
International Business Education and Research (IU incidence of missing values, the amount of informa-
CIBER)for funding the empirical component of this tion availablefor cluster-analyzinginternationalmar-
study. We also thank ProfessorsAlan Rugman, Hans keting strategies was reduced, and the weight
Thorelli,KeithBlois,Shaoming Zou, RockneyWalters, distributed to each broad strategy dimension was
Rebecca Slotegraaf, Krishna Erramilli, and Changsu affected. Further, by replacing the remaining missing
Kim for their valuable comments on earlier drafts of values with item means, the total variance in the
this paper. Special thanks go to JIBS Departmental dataset was reduced in unpredictable ways (we thank
Editor Professor G. Tomas M. Hult and the two one of the anonymous reviewers for pointing that
anonymous reviewers for their helpful criticisms and out). Unfortunately, missing data was an inherent
guidance throughout the review process. Finally, we limitation of our data collection method, and we had
thank the two groups of Kelley School student to deal with it in the most pragmaticway. We did,
assistants whom we engaged in Spring 2004 and Fall however, examine the correlations between removed
2004 to perform the data coding. and retained items, and found them to be moderately

521
high and significant within each broad strategy p-values or significance levels for the pseudo-t statis-
dimension (e.g., integration-independence),meaning tics. The pseudo-t's are used here for interpretive
the retained items should have captured some purposes only; no statisticalinferencesabout popula-
informationcontained in the removed items. We did tion parameters are made. As a rule of thumb,
also try to performthe clusteranalysisusing an equal however, we pay attention to pseudo-t's of greater
number of items (specifically,two) per broad dimen- than 2.0 and treat these as sizeable differences.
sions, selecting the ones that had greatervarianceand 7Multinationalfirm performanceis broadly defined
were less correlated with other items. The cluster here as organizationaleffectivenessat the corporation
solution turned out to be similarto the one we report, level that translatesinto superiormarketand financial
meaning our results were reasonably robust despite performance.It is acknowledged that distinctionsmay
the reductionof total varianceand the redistributionof be made across multiple dimensions of firm perfor-
weight among the three broad strategy dimensions. mance (e.g., market share, profitability, growth).
5Subtlerdistinctionswere illuminatedwhen the top When disaggregated, performance may also vary
branchin the two-clustersolution (see dendrogramin considerably across country subsidiaries. However,
Figure1) was treated as two sub-branchesinstead. As we focus here on a general corporation-levelperfor-
we found later, those two sub-branches were asso- mance construct and leave the finer distinctions to
ciated with two (separate) mixed configurations of future research.
standardization,concentration,and integrationstrate- 8The term 'subsidiary'is broadlyused here to refer
gies, labeled as the InfrastructuralMinimalistsand the to any unit of a multinational corporation that is
TacticalCoordinators,respectively.Withoutseparating located in a country outside its home base. For the
these sub-branches,the two clusterswould have been present purpose, no distinction is made across
identified as a single cluster of (moderately) low differentlegal/equity holding statusesper se of various
ratingsacross all dimensions. subsidiaries, except when these statuses affect the
6Because statistical testing of cluster differences substantivecharacteristicsof the subsidiarynetwork,in
along dimensions used to derive the clustersis mean- which case the configurationof subsidiarycharacter-
ingless (Lattin et al., 2003), we do not report the istics should capture the differences.
References
Bartlett,C.A. and Ghoshal, S. (1986) 'Tap your subsidiariesfor Doty, D.H., Glick,W.H. and Huber, G.P. (1993) 'Fit, equifinality,
global reach', HarvardBusinessReview64(6): 87-94. and organizational effectiveness: a test of two configurational
Birkinshaw,J. (1996) 'How multinational subsidiary mandates theories', Academyof ManagementJournal36(6): 1196-1250.
are gained and lost', Journalof InternationalBusinessStudies Douglas, S.P. and Wind, Y. (1987) 'The myth of globalization',
27(3): 467-496. ColumbiaJournalof WorldBusiness22(4): 47-58.
Birkinshaw,J. and Hood, N. (1998) 'Multinational subsidiary Drazin,R.and Van de Ven, A.H.(1985) 'Alternativeforms of fit in
evolution: capability and charter change in foreign-owned contingency theory', AdministrativeScience Quarterly30(4):
subsidiarycompanies', Academyof ManagementReview23(4): 514--539.
773-795. Elinder,E. (1961) 'How international can European advertising
Birkinshaw, J.M. and Morrison, A.J. (1995) 'Configurations be?' The InternationalAdvertiser(December): 12-16.
of strategy and structure in subsidiaries of multinational Fatt, A.C. (1964) 'A multinational approach to international
corporations', Journalof InternationalBusinessStudies 26(4): advertising', The InternationalAdvertiser(September): 17-20.
729-753. Ghoshal, S. and Bartlett, C.A. (1990) 'The multinational
Boddewyn, J.J.,Soehl, R. and Picard,J. (1986) 'Standardization corporation as an interorganizational network', Academy of
in internationalmarketing:is Ted Levittin fact right?' Business Management Review15(4): 603-625.
Horizons29(6): 69-75. Goodman, J.B. and Moreton, P. (1995) 'AXA: the global
Buzzell, R.D. (1968) 'Can you standardize multinationalmarket- insurance company', Harvard Business School Case No. 9-
ing?' HarvardBusinessReview46(6): 102-11 3. 793-094 (rev. 15 November 1995), HarvardBusiness School
Child, J. (1972) 'Organizational structure, environment Publishing:Boston, MA.
and performance: the role of strategic choice', Sociology Gresov, C. (1989) 'Exploringfit and misfitwith multiple contin-
6(1): 1-22. gencies', AdministrativeScience Quarterly34(3): 431-453.
Corey, E.R.(1998) 'Writingcases and teaching notes', Harvard Gresov, C. and Drazin, R. (1997) 'Equifinality: functional
BusinessSchool Note No. 9-399-077, HarvardBusinessSchool equivalence in organization design', Academyof Management
Publishing:Boston, MA. Review22(2): 403-428.
Craig, C.S. and Douglas, S.P. (2000) 'Configural advantage in Gupta, A.K.and Govindarajan,V. (1991) 'Knowledge flows and
global markets',Journalof InternationalMarketing8(1): 6-25. the structure of control within multinational corporations',
D'Cruz,J.R.(1986) 'Strategic Management of Subsidiaries',in H. Academyof Management Review16(4): 768-792.
Etemad and L.S. Dulude (eds.) Managing the Multinational Hagen, J.M. (1999) 'Ben & Jerry's- Japan', RichardIvey School
Subsidiary,Croom Helm: London, pp: 75-80. of Business Case No. 9A99A037, Ivey Management Services:
DiMaggio, P.J.and Powell, W.W. (1983) 'The iron cage revisited: London, Ontario, Canada.
institutional isomorphism and collective rationality in Hair,J.F.,Anderson, R.E.,Tatham, R.L.and Black,W.C. (1998)
organizational fields', American Sociological Review 48(2): Multivariate Data Analysis, 5th edn, Prentice-Hall: Upper
147-1 60. Saddle River,NJ.

522
Hambrick, D.C. (1984) 'Taxonomic approaches to studying Comments, 10th edn, Pearson Prentice-Hall:Upper Saddle
strategy: some conceptual and methodological issues', Journal River,NJ, pp: 324-340.
of Management 10(1): 27-41. Larsson, R. (1993) 'Case survey methodology: quantitative
Hambrick,D.C. and Schecter, S.M. (1983) 'Turnaroundstrate- analysis of patterns across case studies', Academyof Manage-
gies for mature industrial-productbusiness units', Academyof ment Journal36(6): 1515-1546.
ManagementJournal26(2): 231-248. Lattin, J.M., Carroll, J.D. and Green, P.E. (2003) Analyzing
Hamel, G. and Prahalad, C.K. (1985) 'Do you really have a MultivariateData, Thomson Brooks/Cole:Pacific Grove, CA.
global strategy?' HarvardBusinessReview63(4): 1 39-148. Levitt,T. (1983) 'The globalization of markets', HarvardBusiness
Hannan, M.T. and Freeman, J. (1977) 'The population ecology Review61(3): 92-102.
of organizations', American journal of Sociology 82(5): McKelvey, B. (1975) 'Guidelines for the empirical classification
929-964. of organizations', Administrative Science Quarterly 20(4):
Henzler, H. and Rall,W. (1986) 'Facing up to the globalization 509-525.
challenge', McKinseyQuarterly86(1): 52-68. Meyer, A.D., Tsui,A.S. and Hinings,C.R.(1993) 'Configurational
Hewett, K. and Bearden, W.O. (2001) 'Dependence, trust, and approaches to organizational analysis', Academy of Manage-
relationalbehavior on the part of foreign subsidiarymarketing ment Journal36(6): 1175-1195.
operations: implications for managing global marketing Miles, R.E. and Snow, C.C. (1978) Organizational Strategy,
operations', Journalof Marketing65(4): 51-66. Structure,and Process,McGraw-Hill:New York.
Hout, T., Porter, M.E. and Rudden, E. (1982) 'How global Miller, D. (1981) 'Toward a new contingency approach: the
companies win out', HarvardBusinessReview60(5): 98-105. search for organizational gestalts', Journal of Management
Hovell, P.J.and Walters, P.G.P. (1972) 'Internationalmarketing Studies 18(1): 1-26.
presentations: some options', EuropeanJournalof Marketing Miller, D. (1986) 'Configurations of strategy and structure:
6(2): 69-79. towards a synthesis', Strategic Management Journal 7(3):
Hulbert, J.M., Brandt, W.K. and Richers, R. (1980) 'Marketing 233-249.
planning in the multinational subsidiary: practices and Miller, D. (1987) 'The genesis of configuration', Academy of
problems', Journalof Marketing44(3): 7-15. Management Review12(4): 686-701.
Jain, S.C. (1989) 'Standardization of international marketing Miller, D. (1996) 'Configurations revisited', Strategic Manage-
strategy: some research hypotheses', Journal of Marketing ment Journal17(7): 505-512.
53(1): 70-79. Miller,D. and Friesen, P.H. (1977) 'Strategy making in context:
Jain, S.C. (2001) Marketing Planning and Strategy, 6th edn. ten empirical archetypes', Journal of Management Studies
Thomson-Southwestern Publishing:Cincinnati,OH. 14(3): 251-280.
Jarillo, J.C. and Martinez, J.I. (1990) 'Different roles for Miller, D. and Friesen, P.H. (1978) 'Archetypes of strategy
subsidiaries:the case of multinationalcorporations in Spain', formulation', Management Science24(9): 921-933.
StrategicManagementJournal11(7): 501-512. Miller,D. and Friesen,P.H. (1980) 'Archetypesof organizational
Jauch, L.R.,Osborn, R.N. and Martin, T.N. (1980) 'Structured transition', AdministrativeScience Quarterly25(2): 268-299.
content analysis of cases: a complementary method for Miller, D. and Friesen, P.H. (1984) Organizations:A Quantum
organizational research', Academy of Management Review View,Prentice-Hall:Englewood Cliffs, NJ.
5(4): 517-525. Miller, D. and Mintzberg, H. (1983) 'The Case for Configura-
Kashani,K.and Kassarjian,J.B.M.(1998) 'Sony Europa(A)', IMD tion', in G. Morgan (ed.) BeyondMethod: Strategiesfor Social
Case No. IMD-5-0488, Lausanne, Switzerland: International Research,Sage: BeverlyHills,CA, pp: 57-73.
Institutefor Management Development. Milligan, G.W. and Cooper, M.C. (1985) 'An examination of
Keegan, W.J. (1969) 'Multinationalproduct planning: strategic procedures for determining the number of clusters in a data
alternatives',Journalof Marketing33(1): 58-62. set', Psychometrika50(2): 159-1 79.
Kelly,G.A. (1955) ThePsychologyof PersonalConstructs,Norton: Noble, C.H. and Mokwa, M.P. (1999) 'Implementing marketing
New York. strategies: developing and testing a managerial theory',
Kerin, R.A. and Peterson, R.A. (2004) Strategic Marketing journal of Marketing63(4): 57-73.
Problems:Cases and Comments, 10th edn, Pearson Prentice- Nohria, N. and Ghoshal, S. (1997) The DifferentiatedNetwork:
Hall:Upper Saddle River,NJ, pp: 324-340. Organizing Multinational Corporations for Value Creation,
Ketchen, D.J., Combs, J.G., Russell,C.J.,Shook, C., Dean, M.A., Jossey-Bass:San Francisco,CA.
Runge, J., Lohrke, F.T., Naumann, S.E., Haptonstahl, D.E., Ohmae, K. (1989) 'Managing in a borderless world', Harvard
Baker, R., Beckstein, B.A., Handler, C., Honig, H. and BusinessReview67(3): 152-161.
Lamoureux, S. (1997) 'Organizational configurations and Olusoga, S.A. (1993) 'Market concentration versus market
performance: a meta-analysis', Academy of Management diversificationand internationalization:implications for MNE
Journal40(1): 223-240. performance', InternationalMarketingReview10(2): 40-59.
Ketchen, D.J.and Shook, C.L. (1996) 'The application of cluster Ozsomer, A. and Prussia,G.E. (2000) 'Competing perspectives
analysis in strategic management research: an analysis and in international marketing strategy: contingency and process
critique', StrategicManagementJournal17(6): 441-458. models', Journalof InternationalMarketing8(1): 27-50.
Ketchen, D.J., Thomas, J.B. and Snow, C.C. (1993) 'Organiza- Porter, M.E. (1986) 'Changing patterns of internationalcompe-
tional configurations and performance: a comparison of tition', CaliforniaManagement Review28(2): 9-40.
theoretical approaches', Academy of Management Journal Punj,G. and Stewart, D.W. (1983) 'Clusteranalysisin marketing
36(6): 1278-1313. research: review and suggestions for application', journal of
Kotler, P. (1986) 'Global standardization: courting danger', MarketingResearch20(2): 134-148.
Journalof ConsumerMarketing3(2): 13-15. Quelch, J.A. and Hoff, E.J.(1986) 'Customizing global market-
Laidler, N. (1996) 'Mary Kay Cosmetics: Asian Market ing', HarvardBusinessReview64(3): 59-68.
Entry', Harvard Business School Case No. 9-594-023 (rev. Roberts, M.J. (2001) 'Developing a Teaching Case (Abridged)',
26 July 1996), HarvardBusiness School Publishing: Boston, Harvard Business School Note No. 9-901-055, Harvard
MA. Business School Publishing:Boston, MA.
Laidler,N. (1998) 'Gallo Rice', HarvardBusinessSchool Case No. Roostal, I. (1963) 'Standardization of advertising for Western
9-593-018 (rev. January6, 1998), HarvardBusiness School Europe',Journalof Marketing27(4): 15-20.
Publishing:Boston, MA. Roth, K. (1992) 'International configuration and coordination
Lambin, J. (2004) 'Godiva Europe', in R.A. Kerin and R.A. archetypes for medium-sized firms in global industries',Journal
Peterson (eds.) Strategic Marketing Problems: Cases and of InternationalBusinessStudies23(3): 533-549.

International marketing strategy archetypes LewisKSLimet al "-
523
Roth, K. and Morrison, A.J. (1992) 'Implementing global logical University,Singapore.His researchfocuses on
strategy: characteristicsof global subsidiarymandates', Journal
of InternationalBusinessStudies23(4): 715-735. strategy configurations and capability-buildingprac-
Roth, K., Schweiger, D.M. and Morrison, A.J. (1991) 'Global tices that explain firm performancevariations.
strategy implementation at the business unit level: operational
capabilities and administrative mechanisms', journal of Inter- Frank Acito (PhD, SUNY Buffalo) is Professor of
national BusinessStudies22(3): 369-402.
Rugman, A.M. and Verbeke, A. (2001) 'Subsidiary-specific Marketing and Associate Dean for Academic Pro-
advantages in multinational enterprises', Strategic Manage- grams at the Kelley School of Business, Indiana
ment Journal22(3): 237-250.
Rumelt, R.P., Schendel, D.M. and Teece, D.J. (1994) Funda- University. His research focuses on the assessment
mental Issues in Strategy, Harvard Business School Press: and implementation of marketing research tools
Boston, MA. and techniques and on marketing strategy. His
Rutenberg,D.P. (1982) MultinationalManagement,Little,Brown
& Company: Boston, MA. research has appeared in the Journalof Marketing
Samiee, S. and Roth, K. (1992) 'The influence of global Research, the Journal of Consumer Research, the
marketingstandardizationon performance',Journalof Market- Journalof Marketing,and other publications.
ing 56(2): 1-17.
Sarle, W.S. (1983) SAS Technical Report A-108: The Cubic
ClusteringCriterion,SAS Institute:Cary, NC. Alexander Rusetski is a doctoral candidate in
Sheth, J.N. (1986) 'Global markets or global competition?'
Journalof ConsumerMarketing3(2): 9-11. Marketing at the Kelley School of Business, Indiana
Solberg, C.A. (2000) 'Standardization or adaptation of the University and a Lecturerof Marketingat the School
international marketing mix: the role of the local subsi- of Administrative Studies, YorkUniversity, Toronto.
diary/representative',Journalof InternationalMarketing8(1): His research focuses on managerial decision-mak-
78-98.
Sorenson, R.Z. and Wiechmann, U.E. (1975) 'How multi- ing, brand management, and marketing strategy,
nationals view marketing standardization', HarvardBusiness
Review53(3): 38-167. including international marketing strategy.
Szymanski,D.M., Bharadwaj,S.G. and Varadarajan,P.R.(1993)
'Standardizationversusadaptation of international marketing
strategy: an empirical investigation', Journal of Marketing Appendix A: Case qualification checklist
57(4): 1-17.
Theodosiou, M. and Leonidou, L.C. (2003) 'Standardization * Is this case concerned with the marketing
versus adaptation of international marketing strategy: an
strategyof a company operatingin an interna-
integrative assessment of the empirical research', International tionalenvironment?
BusinessReview12(2): 141-171.
Vorhies, D.W. and Morgan, N.A. (2003) 'A configuration theory * Does the case feature at least one identifiable
assessment of marketing organization fit with business business unit/level for which a single, distinct set
strategy and its relationship with marketing performance', of international marketing strategy is formulated
Journalof Marketing67(1): 100-115.
Walker,O.C. and Ruekert,R.W.(1987) 'Marketing'srole in the and executed?
implementation of business strategies: a critical review and * Is this case more about the featured business
conceptual framework',Journalof Marketing51(3): 15-33.
White, R.E. and Poynter, T.A. (1984) 'Strategies for foreign- unit's overall international marketing strategy
owned subsidiaries in Canada', Business Quarterly 49(2): than about its marketing strategy for specific
59-69.
Wiechmann, U.E. and Pringle, L.G. (1979) 'Problems that countrymarkets?
plague multinational marketers', Harvard Business Review * Is the case devoted more to describing the
57(4): 118-124. featured business unit's international marketing
Wind, Y. (1986) 'The myth of globalization', Journalof Consumer
Marketing3(2): 23-26. strategy than to analyzing a particularmarketing
Woo, C.Y.Y. and Cooper, A.C. (1981) 'Strategies of effective event or decision?
low share businesses', Strategic Management Journal 2(3): * Does the case describe, in a comprehensive and
301-318.
Yip, G.S. (1989) 'Global strategy... in a world of nations?' Sloan
detailed manner, the various aspectsof the featured
Management Review31(1): 29-41. company's international marketing strategy?
Zou, S. and Cavusgil, S.T. (2002) 'The GMS: a broad o Standardization-adaptation dimensions
conceptualization of global marketing strategy and its effect
on firm performance', Journalof Marketing66(4): 40-56. o Concentration-dispersion dimensions
Zou, S., Andrus, D.M. and Norvell, D.W. (1997) 'Standardiza- o Integration-independence dimensions
tion of marketing strategy by firms from a developing * Does the case writer describe the details of the
country', InternationalMarketingReview14(2): 107-123.
case vividly enough to enable quantification of
the featured company's international marketing
About the authors strategy?
Lewis Lim is a doctoral candidate in Marketing at the * Has the strategy described in the case been
Kelley School of Business, Indiana University, and actually implemented for a reasonable period of
SeniorTutorof Marketingand InternationalBusiness time (as opposed to being in the planning stage
at the Nanyang Business School, Nanyang Techno- or being a transition strategy)?

524
Evaluation * This case may be suitable, although a handful of

items cannot be coded; second opinion to be sought.
* This case is definitely suitable for our project; * This case is definitely not suitable; drop it from
very few items cannot be coded. the sample.
Appendix B: International marketing strategy case coding scheme

Pleaseratethe internationalmarketingstrategyof the companyfeatured ANDLOGISTICSPLANNING
9.DISTRIBUTION
in the case along each of the followingdimensions.If you cannotprovide Each countryoffice is Distributionand logistics
a rating for an item, check the box "No information on this item." foritsown - - - - - - - - -
responsible planningactivities are
distributionandlogistics 0 1 2 3 4 5 6 7 8 9 10 consolidateglobally at a
planning single location
A. Standardization-Adaptation of Marketing Mix Elements Noinformationonthisitem
Referencepage(s)aragraphs(:
1. BRAND NAME
10. AFTER-SALE SERVICE AND TECHNICAL SUPPORT
A differentbrandname, The exact same brand Each countryoffice is After-sale service and
includinglogoisadoptedin 0 1 2 3 4 5 6 7 8 9 10 name, includingthe logo, is responsiblefor its own technical supportactivities
each market used in all markets after-sale service and 0 1 2 3 4 5 6 7 8 9 10 are consolidatedglobally at
technical support a single location
Referencepoge(sa)paragraphs
():_ Ljo information
onthisiten
Referencepage(s)/paragraphs(s): [ Noinformation on thisitem
2. PRODUCT DESIGN 11. ADVERTISING AND PROMOTION PLANNING

For each market,the The exact same product Each countryoffice is Advertisingand
productis designedto meet specifications,including responsiblefor its own promotionalplanning
thelocalneedsand 0 1 2 3 4 5 6 7 8 9 10 physicaldimensions,are advertisingand 0 1 2 3 4 5 6 7 8 9 10 acti-ities are consolidated
preferences used in all markets promotionalplanning globallyata singlelocation
Referencepage(s)/paragraphss): LNo information
tonhisitm Referencepage(s)/paragraphs(s): onthisitem
Noinformation
3. PRODUCT PACKAGING
12. MARKETING RESEARCH
The
Theproductpackagingand exactsameproduct
labelng are customizedfor packaging,includingthe Each countryoffice is Marketingresearch
1 2 3 4 5 6 7 8 9 10 labeling,isusedinall activities are consolidated
each market responsibleforitsown 0 1 2 3 4 5 6 7 8 9 10
markets marketingresearch globally at a single location
Referencepage(s)/paragraphs(s): LNe information
onthisitem
Referencepage(s)/paragraphs(s): Noinformation
on thisitem
4. ADVERTISING
THEME
C. Integration-Independence of Competitive Actions
A localizedadvertising The exact same advertising
templateandtheme,is 2 3 4 5 6-7 8 9 to stylemessae, and format
foreachmarket
created isused obaly formt 13. MARKETINGPLANNINGANDCOMPETITIVE
DECISION
MAKING
Referencepage(s)/paragraphs(s) Noinfomationonthisite Each countryoffice Marketingplanningand
formulatesits own competitivedecision
marketingplan and makes 0 1 2 3 4 5 6 7 8 9 10 makingare tightly
5. SALES PROMOTION TACTICS its own competitive coordinatedacross all
decisions markets
A differentmix of sales The exact same sales
Referencepage(s)/paragraphs(s): U Noinformation onthisitem
promotiontools is used in 0 1 2 3 45 6 7 8 9 10 promotiontools and tactics
eachmarket amusedglobally
o information
onthisite 14. CROSS-SUBSIDIZATION
OFCOMPETITIVE
CAMPAIGNS
Reference Each countryoffice is Each countrymarketis
page(s)/paragraphs(s):
treatedas a standalone assigned a specific
6. CHANNEL DESIGN profit center, i.e., as apart 0 1 2 3 4 5 6 7 8 9 10 competitiverole and
of a collection of profits fromone market
t
The channelstructureand The exact same channel independentsubsidiaries used to financethe
for each strucure and systemare whose performancesare competitivecampaignsin
e 0niqu
1 2 3 4 5 6 7 8 9 10 ctandsystemsepartey eluated othermarkets
market
ymarkem used in all markets
Referencepage(s)/paragraphs(s): L No
information
onthisitem
Referencepage(s)/paragraphs(s): onthisitem
UNo information
15. COMPETITIVE RESPONSE INTEGRATION
7. PRICING Competitivebattlesare on a
Pricelevelsaredetermined The same price position, fought locally in each attles on a
Thcompetitive
battles
competitiveb
0 1 2 3 4 5 6 7 8 9 10
ertirelyaccordingto local relativeto key global countrymarketand
competitivesituationand 0 1 2 3 4 5 6 7 8 9 10 competitors,is maintained countrymanagersdo not to copetitors
tobalscalepond
attacksin one marketby
subsidiarygoals acrossall markets meddle with the affairs of counterattackingin another
othercountriescoua cket
Referencepage(s)/paragraphs(s): on
L3No information market
thisiitem Referencepage(s)/paragraphs(s):. Li Noinformation onthisitem
B. Concentration-Dispersion
of MarketingValue ChainActivities
16. COMMUNICATION AND MUTUAL CONSULTATION
country
Different managers Country
managers
8. PRODUCT DESIGN AND DEVELOPMENT do not communicateand frequentlycommunicate
Each countryoffice is Productdesignand share informationwith one 0 1 2 3 4 5 6 7 8 9 10 andshare informationwith
responsiblefor its own anotherabouthow they one anotherso as to
developmentactivities are
productdesign and new 0 1 2 3 4 5 6 7 8 9 10 consolidatedglobally at a deal with competition coordinatetheir
productdevelopment single location competitiveactions
Reference Noinformation
onthisitem Referencepage(s)/paragraphs(s): i onthisitem
Noinformation
page(s)/paragraphs(s):
Acceptedby G. Tomas M. Hult, DepartmentalEditor, 11 February2005. This paper has been with the authors for two revisions.

Lim and Acito

Uploaded by

Copyright:

Available Formats

Lim and Acito

Uploaded by

Document Information

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

Lim and Acito

Uploaded by

Copyright:

Available Formats

Development of Archetypes of International Marketing Strategy

Author(s): Lewis K. S. Lim, Frank Acito, Alexander Rusetski

Development of archetypesof international

Lewis KS Liml'2, FrankAcito' Abstract

For the past four decades, business scholars have sought to

journal of International Business Studies

Journal of International Business Studies

Journal of International Business Studies

Journal of International Business Studies

Table 1 Exampleof two combinations of strategy ratingssharing the same GMSscore

Strategydimension Firmla Firm2a

GMSI: product standardization 3.86 -0.35

Journal of International Business Studies

Journal of International Business Studies

Rescaled Distance Cluster Combine

regarding the source of the difference (which was Daewoo 4

instances were the coders unable to successfully DHL

recoding the items ourselves. Haier

resulted in a coded case sample of about 80 cases. Toyota

that did not have at least 10 out of the 16 items Montgras

tion and analysis of our data set. To prepare our Kikkoman 30

item means. We believed these two steps collec- PSA Peugeot

tively helped us maintain the integrity of our data Jurassic Park

utilized cluster analysis because of the technique's

Journal of International Business Studies

Table 2 Pseudo-F, cubic clustering criterion (CCC), and agglom-

journal of International Business Studies

Localizedad Same advertising

Differentmix of Same sales

Each country . Productdesign

Each country Distributionand

Each country \ Advertisingand promotion

Each country I Marketingplanning

Journal of International Business Studies

Centroid Std dev Centroid Std dev Centroid Std dev A vs B

Brand name standardization(ST_BRAND)8.59 1.56 8.34 1.83 6.08 2.49 0.407 3.

ArchetypeB the fact that the company uses local subsidiaries

Journal of International Business Studies

tration rating of 3.0), presumably because of the 2-

writing and sales were labor intense and generally varied in

However, what stood out in the AXA case was a b

gration, with average competitive decision-making

Journal of International Business Studies

journal of International Business Studies

journal of International Business Studies

journal of International Business Studies

Journal of International Business Studies

SpecializedContributor Local Implementer I World Mandate

Figure4 Contingentfit betweenarchetypesand configurations

Journal of International Business Studies

Journal of International Business Studies

Finally, after archetypes have been identified and Notes

journal of International Business Studies

Journal of International Business Studies

Journal of International Business Studies

Journal of International Business Studies

Evaluation * This case may be suitable, although a handful of

Appendix B: International marketing strategy case coding scheme

2. PRODUCT DESIGN 11. ADVERTISING AND PROMOTION PLANNING

Journal of International Business Studies