PDMA Research On New Product Development Practices: Updating Trends and Benchmarking Best Practices

jjjj
PDMA Research on New Product Development Practices:

Updating Trends and Benchmarking Best Practices
Abbie Griffin
Product development professionals may have the feeling that yet another
buzzword or magic bullet always lurks just around the corner. However, re-
searchers have devoted considerable effort to helping practioners determine
which tools, techniques, and methods really do offer a competitive edge. Starting
30 years ago, research efforts have aimed at understanding NPD practices and
identifying those which are deemed “best practices.” During the past five years,
pursuit of this goal has produced numerous privately available reports and two
research efforts sponsored by the PDMA.
Abbie Griffin summarizes the results of research efforts undertaken during the
past five years and presents findings from the most recent PDMA survey on NPD
best practices. This survey, conducted slightly more than five years after PDMA’s
first best-practices survey, updates trends in processes, organizations, and out-
comes for NPD in the U.S., and determines which practices are more commonly
associated with firms that are more successsful in developing new products. The
survey has the following objectives: determining the current status of NPD
practices and performance; understanding how product development has
changed from five years ago; determining whether NPD practice and perfor-
mance differ across industry segments; and, investigating process and product
development tools that differentiate product development success.
The survey findings indicate that NPD processes continue to evolve and
become more sophisticated. NPD changes continually on multiple fronts, and
firms that fail to keep their NPD practices up to date will suffer an increasingly
marked competitive disadvantage. Interestingly, although more than half of the
respondents use a cross-functional stage-gate process for NPD, more than one-
third of all firms in the study still use no formal process for managing NPD.
The findings suggest that firms are not adequately handling the issue of
team-based rewards. Project-completion dinners are for the most frequently used
NPD reward; they are also the only reward used more by best-practice firms than
by the rest of the respondents. The best-practice firms participating in the study
do not use financial rewards for NPD. Compared to the other firms in the study,
best-practice firms use more multifunctional teams, are more likely to measure
NPD processes and outcomes, and expect more from their NPD programs.
Address correspondence to Prof. Abbie Griffin, University of Illinois,

Urbana-Champaign 350 Commerce West, Champaign, IL 61820, Phone:
(217) 244-8549, E-mail: abbieg@uiuc.edu.
J PROD INNOV MANAG 1997;14:429 – 458

© 1997 Elsevier Science Inc. All rights reserved. 0737-6782/97/$17.00
655 Avenue of the Americas, New York, NY 10010 PII S0737-6782(97)00061-1
430 J PROD INNOV MANAG A. GRIFFIN
1997;14:429 – 458
Introduction: Why Track Product presented to potential users as useful aids without
Development Practices definitive proof as to how they improve development
or even that they improve development at all. One of
I
f the world was stable, there would be no need to the ways of differentiating between more and less
change business operations and methods, nor to useful NPD approaches is by tracking NPD manage-
understand what has changed and what works ment and performance over time and linking changes
well. However, firms operate in dynamic environ- in outcome to practices which have been implemented.
ments, not stable ones. Both the competitive and in- Part of the Product Development & Management
ternal environments in which firms operate evolve Association’s (PDMA’s) mission is to create and dis-
over time. In response, management processes must seminate knowledge about managing and improving
also change over time so that firms can remain effec- new product development. As one means of fulfilling
tive and profitable through the changing situation. this mission, the association developed a research
Some of the changes in business with the potential to agenda in 1989 to proactively aid in the process of
impact the ways in which new product1 development product development knowledge creation [7]. This
(NPD) is practiced and managed over the last five research agenda has resulted in several papers and
years include: many presentations of findings in three streams of
• increased levels of competition (more firms com- research: profiles and compensation of new product
peting for the same markets) [34,40,59] professionals [25,26], measuring product development
• rapidly changing market environments [8,53] success [31,32], and tracking the practices of manag-
• higher rates of technical obsolescence [49] and ing product development [36,50,51]. This article adds
• shorter product life cycles [3,58]. to the last stream of PDMA research.
This article presents results from the second PDMA
A primary impact of these environmental changes is to survey on product development best practices fielded
drive firms to implement changes which help speed in the summer of 1995, just over five years after
products through development, and improve process PDMA’s first best practices survey [50,51]. The ob-
efficiency and overall NPD effectiveness. jectives of the second wave of best practices research
Over previous decades, and the last five years in were to:
particular, many new processes, techniques, and tools
• Determine the current status of product develop-
purporting to improve the practice of product devel-
ment practices and performance
opment have been developed by academics, consult-
• Understand how product development has changed
ants and practitioners and implemented by a few or
from five years ago
many firms. While the effectiveness of some of the
• Determine whether differences exist in NPD prac-
ideas espoused as useful has been rigorously re-
tice or performance across industry segments
searched prior to general diffusion to the product de-
• Investigate process and product development tools
velopment community [11,12,30,57], some have been
which differentiate product development success.
1
Throughout this article, “product” refers to both manufactured goods A summary of the key findings is contained in Exhibit
and services. When physical products are considered separately from
services, they are referred to as manufactured goods. 1.
While parts of the survey instrument were kept
identical to those used in 1989 so that direct longitu-
BIOGRAPHICAL SKETCH
dinal comparisons could be made, the instrument was
Abbie Griffin is a Professor of Marketing at the University of
vastly expanded in terms of topics covered and num-
Illinois, Urbana-Champaign. Her research focuses on measuring ber of questions asked. In addition, the survey was sent
and improving the process of new product development. Although to a much larger community than just PDMA mem-
JPIM is her preferred publication outlet, she also has published
articles on product development in Marketing Science, Manage-
bers. Details about the survey instrument and research
ment Science, Sloan Management Review, and Journal of Market- method are in the methods section.
ing Research. Prior to becoming an academic, she worked as an This article is the first of a series of articles which
engineer at Polaroid Corporation, in product development at Corn- are expected to be published from the PDMA 1995
ing Glass Works, and as a technology strategy consultant at Booz,
Allen and Hamilton. Her personal background includes a passion best practices data. The purpose of this article is to
for quilting. describe how product development practices have
changed, both relative to the results obtained five
PDMA RESEARCH ON NEW PRODUCT DEVELOPMENT PRACTICES J PROD INNOV MANAG 431
1997;14:429 – 458
years ago and relative to what other investigations

Exhibit 1. Summary of Key Findings have said about product development practices. It also
presents analyses of product development processes
1. The Best do not succeed by using just one NPD practice and organizations which are associated with higher
more extensively or better, but by using a number of
them more effectively simultaneously. performance.
2. NPD change is evolutionary, but unceasing. It moves
forward simultaneously on multiple fronts.
3. NPD processes have continued to evolve and become The History of Tracking Product
more sophisticated over time. Development Practices
● While nearly 60% of US firms use a cross-functional
stage gate process for NPD, 38.5% of all firms still There is a long history of studies tracking product
use no formal process for managing NPD. Best prac- development management practices. Across these
tice firms have implemented stage gate processes to a
studies, several consistent themes emerge. Table 1
greater extent than the Rest of the firms.
summarizes the studies presented in this section.
● Best practice firms are more likely to drive product Issues in new product development practices were
development efforts through specific NPD strategies
at both the program and project level. first investigated in the aggregate by Booz, Allen and
Hamilton (BAH) in 1968, with the effort repeated in
● NPD processes used by Best practice firms are more
likely to start with a strategy step and be more com-
1982. Their ground-breaking results were privately
plex because they include more steps. published in managerially-targeted brochures [4,5].
Booz, Allen and Hamilton’s 1968 report, based on
● Processes for service firms are less complex than for
manufactured goods firms. knowledge accrued from over 800 client assignments
4. Firms support NPD efforts in two separate locations of and data obtained from just over 49 firms, reported
their organization, on average. No single or combination of that almost 1⁄3 of all product development projects
structures relates to achieving best practice. commercialized by firms were failures, with this rate
5. Project managers (61%) and champions (43%) are most essentially independent of industry [4]. Most of the
likely to lead NPD projects. Management appoints NPD
commercialization failures occurred because the idea
leaders over 70% of the time.
6. Firms have not grappled adequately with team-based
or its timing was wrong. This report presented the
rewards. Project completion dinners are the most fre- product development mortality curve, which showed
quently-used NPD reward, and the only reward used that, on average, 58 ideas were considered for every
more by Best practice firms than the Rest (72% versus
54%). Best practice firms do not use financial rewards successful new product commercialized. This report
for NPD. also delineated a common six-stage process for prod-
7. Over 84% of the more innovative projects use multi- uct development which they found most typically used
functional teams. On average, however, multi-functional by firms. Although aspects of the process were slightly
teams are used in only 40–50% of the less innovative
projects. Best practice firms use multi-functional teams modified by firm to cater to the needs of the industry,
more extensively in these less innovative projects (50– product type, and corporate culture, the basic process
60% of the time).
consists of exploration, screening, business analysis,
8. Best practice firms are more likely to measure NPD
performance and expect more out of their NPD efforts. development, testing, and commercialization. The au-
Best practice firms expect 45% of their sales to come thors of the report concluded that heavy attention
from products commercialized in the last three years. In should be focused on the first three stages of the
actuality, 49.2% of their sales did come from products
commercialized over the last five years, about twice the process to minimize failure. They also found that new
rate of the Rest of the firms. products departments, product teams, and a new prod-
9. Even with all the NPD improvements implemented, the uct committee were organizational forms which, sin-
average outcomes have improved only slightly across
many measures. gly or in combination, were found in companies where
● The success rate is stable at 59% of those products
more consistent NPD success had been achieved.
which make it to market. Booz, Allen and Hamilton’s 1982 report is based on
● It takes 6.6 ideas to generate one success, down from in-depth interviews with more than 150 NPD execu-
7 in 1982. Firms are more efficient in weeding out tives and survey responses from more than 700 US
less probable projects earlier in the NPD process. manufacturers [5]. From the analysis of these data, and
● The most progress has been made in reducing NPD based on the recommendations from research on the
cycle times, which average 23.8 months for more use of strategic planning techniques to guide NPD by
innovative projects, down approximately 1⁄3 from five
years ago. Crawford [18], they recommended adding a seventh
1997;14:429 – 458
Table 1. Summary of Prior Best Practice Studies
Differentiating
Year Sponsor Sample the Best Findings (Best Practice)
1968 Booz, Allen and 50 firms unspecified ● Delineated a 6-stage process
Hamilton [4] ● 58 ideas/success
● NPD success rate 5 67%
1982 Booz, Allen and 150 interviews unspecified ● Add strategic planning step to
Hamilton [5] 700 survey responses regression of factors against NP NPD process
broad manufacturing cross- sales as a % of total sales ● 7 ideas/success
section ● NPD success rate 5 65%
1990 Page, PDMA 189 survey responses ● Measure NPD performance

[50,51] broad industry cross-section ● Use multi-functional teams
79% goods unspecified ● Best practice is context-specific
● 11 ideas/success
● NPD success rate 5 58%
1991 Arthur D. Little [1] 701 survey responses ● Use multi-functional teams
9 manufacturing industries ● Provide top management attention
● Early supplier involvement
1993 Kuczmarski [37,38] 77 fax survey responses rated NPD as “successful” or “very ● Tangible and visible top
broad industry cross-section successful” (53% of total) management support
● Provide adequate resources
● Spend more time on up-front steps
● Focus on newer products
1994 Mercer Management 193 R&D managers top 1⁄3 in cycle time, ● Use a customer-centered,
Consulting and broad industry cross-section innovativeness, success rate and disciplined NPD process
R&D revenue contribution, combined ● Cultivate a supportive NPD
Management [46] infrastructure
● Manage the NPD portfolio
● Use a planning stage
1994 Product 129 survey responses top 1⁄3 based on financial, sales and ● Distinguish between features and
Development market share performance needs
Consulting [45] ● NPD scope includes entire
augmented product
1995 Group EFO [33] 103 marketers from 83 firms no Best practices specified ● 25 ideas/success
consumer packaged goods ● Lack of NPD commitment
1995 Mitchell Madison 15 service firms self-rated as better than the ● Actively manage the NPD portfolio
Group [47] competition-top 27% ● Provide adequate resources
● Use an empowered team
● Dedicated NPD team members
1995 Pittiglio Rabin Todd over 200 participants top 20% across 6 metrics ● Measure both project performance
& McGrath [52] many high-tech industries and development effectiveness
● Use multi-functional teams
● Structured process, action-oriented
phase reviews
● Manage product strategy and the
NPD pipeline
● 9.5% cycle time reduction
1996 Southwestern Bell 134 respondents, 7 industries no Best practices specified ● 57% use a formal NPD process
[24] 32-item fax survey ● 58% have reduced cycle time
services and goods ● for 80%, team membership is a
part-time responsibility
1997;14:429 – 458
step to the front of the process delineated in their 1968 resulted from eleven new product ideas or concepts. In
publication. This step has the NPD process beginning aggregate, it appeared that trends in managing NPD
with identifying the new product strategy, then mov- were changing, albeit more slowly than might have
ing into exploration. In 1982, BAH claimed that 77% been expected.
of the respondent firms used product strategy devel- Page [51] also identified several best practices for
opment as the first step in their product development high-tech and low-tech product-producing companies
process. They again found, in this round of research, which differentiated between NPD performance. For
that most firms used multiple organizational structures high-tech firms, these practices included having man-
to guide NPD programs, with the structure used tied to ufacturing devote at least 10% of their time to NPD,
product-specific requirements. In addition, they found having “satisfies customer needs” as an explicit suc-
that since the previous study more management atten- cess criterion, including concept testing and market
tion and financial resources were being given to the testing in the NPD process, and using non-financial
early steps in the NPD process (as they had previously rewards and incentives to motivate people. For low-
recommended), firms were becoming more efficient at tech firms, best practices include using product man-
product development (spending a larger percentage of agers as part of the organization’s new product devel-
their NPD expenditures on successes rather than fail- opment structure, and having “satisfies customer
ures), and needed only seven new products ideas to needs,” “strategic fit or synergy,” and “uniqueness of
generate one success in the marketplace. However, the the new product” as explicit success criteria. In other
average success rate from NPD had not improved, words, Page found that although average performance
even with these operational improvements which had and trends could be identified across the sample, at
been made. least some best practices were context-specific. One of
Over the years since 1982, the results in these the ways in which corporate context could be defined
publications have been widely referenced in academic was by looking at the technology position of the firm.
research on new product development Since the first PDMA survey was completed, a
[10,12,15,32,50]. Although BAH was the first to ana- number of other organizations have investigated and
lyze changes in product development management reported on different aspects of product development
over time, time lags between projects were long. This best practices. The investigations have been published
opened a research opportunity for more routine track- privately, but are publicly available with the results
ing of product development changes over time. presented in open managerial forums.
PDMA has accepted the challenge of periodically In 1991, Arthur D. Little surveyed 701 companies in
providing information about the process and manage- 9 manufacturing industries about their product inno-
ment changes in product development through spon- vation processes, focusing primarily on top manage-
soring regularly-scheduled research on this topic ment’s concerns and improvement efforts [1]. They
which is both managerially interesting and academi- found major differences in concerns and approaches to
cally rigorous. They sponsored their first study of overcome those concerns depending upon the geo-
product development practices and norms in the fall of graphic location of the responding firm: Japanese
1990 with a survey distributed only to PDMA mem- firms were developing new products in markedly dif-
bers. This first PDMA research effort established ferent ways from US firms. The three top US product
norms describing product development changes since development concerns they found were improving
1982 [50] and determined several best practices found product appeal, getting products out on time, and
in high impact new product programs [51]. decreasing product development time. The approach
This initial PDMA research found that over 76% of which most frequently was found to lead to increased
the firms responding to the 1990 survey used multi- US success was “multifunctional teams with signifi-
disciplinary product development teams [50]. Formal cant autonomy dedicated to a project.” Other research
measurement of NPD programs had increased to 76% has shown that top managers formally control the
of the sample, from 66% in 1982. However, even with budgets and plans of these groups only loosely, but
these improvements, the average efficiency (spending exert considerably tighter control over them infor-
on successes rather than failures) and success rate mally by providing them with significant amount of
were unchanged over the previous decade. The mor- top management attention and contact [2]. The next
tality rate of products proceeding through develop- most successful approach to creating successful NPD
ment had increased slightly— one successful product was “early supplier selection and involvement,” as
1997;14:429 – 458
also recommended by the research of Clark and Fuji- practices across the entire process. Responses in this
moto [9]. survey indicate that marketers and engineers in best-
Kuczmarski & Associates published a study on in-class firms carefully distinguish between customer
product development best practices based on the re- requirements (needs) and product features (solutions
sults from 77 respondents from a broad cross-section to customer needs) and have explicit documents which
of industries to a survey in 1993 [37,38]. They found describe each separately. In-depth interviews with cus-
that more successful companies (the 53% of their tomers used to gain a deep understanding of the needs
respondents who rated their NPD performance as of a moderate number of customers (11) are market
“very successful” or “successful”) were more likely to research best practices uncovered in this survey. Fi-
pursue NPD to gain or maintain a competitive advan- nally, the scope of product definitions is wider to
tage, fill a growth or profit gap, arrest margin erosion include the entire augmented product—the sales pro-
or utilize a new technology than less successful firms, cess, packaging, installation, complementary products,
who were more likely to use NPD to retain current and support and service—at best practice firms.
attract new customers. More successful firms had Group EFO, in conjunction with Brandweek, con-
more tangible and visible signs of top management ducts an annual poll on new manufactured goods in the
commitment to NPD, especially in terms of providing consumer packaged goods area [33]. In the 1995 In-
adequate funding and resources. They also focused novation Survey of 103 marketers from 83 firms, they
more effort on new-to-the-world and new-to-the-com- found that only 4 ideas out of 100 considered became
pany products, devoted a larger percentage of the commercial successes, or expressed in mortality curve
product development process to concept screening and language, 25 ideas were required to produce one mar-
testing and rated themselves as being more effective in ketplace success. Consumer packaged goods firms
terminating projects during development. seem to be less effective in generating ideas, in turning
Mercer Management Consulting, Inc., in conjunc- ideas into products, or both, than the predominantly
tion with R&D Magazine, gathered survey responses business-to-business samples in other surveys. These
from 193 R&D managers from a variety of industries firms also indicated that their product development
and linked product development practices to NPD processes are only effective about half the time—and
performance, which was defined as combined self- in no case is the process used effective more than 75%
assessments of cycle time, innovativeness, success rate of the time. Only 60% of the managements were
and revenue contribution [46]. They found that high reported to have a clear point-of-view of the role of
performers (top third) were differentiated from lower new products in their firms. Fully two-thirds of the
performers (bottom third) in their execution of a com- respondents felt that their new product programs were
monly agreed to, customer-centered and disciplined under-resourced. EFO interprets that the survey re-
NPD process, their cultivation of a supportive organi- sponses indicate that there is a lack of commitment to
zation and infrastructure for NPD, and in setting the new products in these firms, or that management does
NPD agenda and managing the portfolio of projects in not understand what commitment to NPD entails.
aggregate. Some of the practices contributing to suc- Also in 1995, the Mitchell Madison Group released
cess the most while differentiating between lower and a study of NPD in a small sample (15) of service firms
higher performers included involving potential cus- [47]. They found that while service firms differ from
tomers directly through numerous stages of product goods-producing firms, notably by the effect of ‘in-
development, formulating product strategy early in the tangibles’ in the satisfaction equation and by the tight
project, consistently following the project execution linkage between service delivery and service manu-
process from project to project, having top manage- facturing—the production of the service is the delivery
ment visibly and tangibly committing to NPD and of the product [21]—many of the key success factors
explicitly formulating and communicating the firm’s for service NPD are identical to those identified for
NPD strategy. manufacturing firms. Using formal processes for ser-
Product Development Consulting, Inc., developed a vices NPD is a fairly recent circumstance—none of the
description of best practices in product definition from firms had been using a process for more than five
129 responses to a survey they administered in 1994 years, over 25% had been using a process for less than
[45]. This research focused only on practices associ- one year, and half had been using a process for be-
ated with the initial few stages of the product devel- tween 1 and 4 years. The “best companies” (self-rated
opment process, rather than trying to understand best as better than the competition) more actively manage
1997;14:429 – 458
the portfolio of projects under development, limit the Summary of Findings from Tracking NPD
number of projects underway at any one time to match Practices
the available resources and involve multi-functional
teams early in the process. These firms also use dif- Over these 11 studies, several themes consistently
ferent organizational structures for NPD depending emerge. First, although NPD processes are a relatively
upon the needs of the project, but allow the project recent phenomena [10,11,15,16], they consistently are
team to fully “own” the process, with the ability to seen as being necessary to effective NPD. Over the
make virtually all the decisions within the team. One years, the focus on the NPD process has moved from
large difference between service firm teams and man- defining an appropriate process, to assuring imple-
ufacturing NPD teams seems to be that virtually all the mentation, to better managing the up-front portion, to
service core team members are dedicated to the team. measuring the process better, to continue improving
Pittiglio Rabin Todd & McGrath (PRTM) also used the process. Interestingly, mortality curves and overall
responses from over 200 organizations from 6 industry success ratios seem to be stable, even with operational
groups in 1995 to determine NPD best practices [52]. improvements being made in the process. The latest
They defined the “best-in-class” as the top 20% aspects of process which are receiving renewed atten-
against a set of six NPD metrics: time-to-market, time- tion are obtaining customer needs for product devel-
to-profitability, project goal attainment, NPD revenue opment and managing the resource allocation process
contribution, R&D Effectiveness Index [44] and across projects (portfolio and pipeline management).
Organizationally, there are two consistent themes.
wasted development project spending. The study iden-
First, virtually every study claims that effectively im-
tified several “best practices,” although the methodol-
plementing multi-functional teams is crucial to NPD
ogy for arriving at these is not defined in this publi-
success. Second, the studies consistently relate that
cation. At the project level best practices include using
NPD is melded into the firm’s organization through
cross-functional teams, a structured development pro-
multiple structures within each firm, and that no one
cess with action-oriented stage reviews and an inte-
structure seems to be associated with consistently
grated set of development tools (such as QFD, rapid
higher performance. We have not yet been able to
prototyping and simulation). Best practice firms also
define the organization and infrastructure which best
manage across projects with product strategy, pipeline
supports effective multi-functional teams over time
management, and technology management processes. and across projects [35].
In the summer of 1996, Southwestern Bell commis- Another consistent finding for producing successful
sioned Fact Finders Inc. to perform a bench-marking NPD is the need for tangible and visible top manage-
study of NPD [24]. A 32-item fax-questionnaire was ment support of NPD, especially in terms of providing
returned by 134 respondents from 7 industries span- adequate funding and resources and explicit, consis-
ning services and manufactured goods, as well as both tent strategies. This includes having a rational process
consumer and business-to-business products. About at firms for allocating resources across projects and a
57% of the respondents used a formal NPD process, well-thought-out strategy for both NPD at the firm,
with 52% including test marketing as part of that and the particular project under way.
process prior to launch. Just over half (56%) of the Finally, recent studies suggest that best practices
firms measure their product development processes, may be somewhat context-specific, and some efforts
with 58% of the respondents reporting reductions in are taking place to better define best practices within
NPD cycle times in the last year. Processes in these contexts rather than in aggregate. For example, Page
respondent firms have been in place longer than in the [51] found different success differentiators depending
service firms from the Mitchell Group study. Over upon the technology base of the firm (high- versus
32% of the firms have had formal processes for five low-tech). The Mitchell Madison services study found
years or more, 54% have used a formal process for that all teams were dedicated NPD teams, while the
1– 4 years, and only 14% have just within the last year Southwestern Bell study, encompassing a number of
implemented a process. Another difference between industries, found only 20% of the respondents used
NPD at these firms and those of the Mitchell study is dedicated teams. The processes used in the consumer
that for 80% of the respondents, team membership is packaged goods firms in the EFO Group seem much
on a part-time basis rather than a full-time basis. Best less effective than those found in more business-to-
practices were not determined in this study. business-oriented surveys because of the differences
1997;14:429 – 458
found in the numbers of ideas required to produce a Variable Operationalizations

commercial success in the marketplace. Future re-
search in best practices, to be most useful, then, should Success Outcome Variables. A total of 7 success
also investigate best practices for specific contexts. criteria were included in the survey measuring overall,
relative, market and financial success at the firm level
[31,32]. Overall success is obtained from the single-
item categoric measure, “position in your industry”
Research Method (most successful, top 1/3, middle 1/3, bottom 1/3)
[51]. Degree to which the NPD program met its ob-
Survey Development
jectives and degree to which the NPD program is a
In developing the PDMA research, a number of “best success allow the firm to assess how well they are
practice” surveys developed since 1982 were gathered. doing relative to their program’s objectives. These
The PDMA 1995 Best Practices survey was developed criteria, from [10], have been used extensively in NPD
by combining questions from several different sour- research. They are measured as the degree to which
ces: the 1990 PDMA [50,51], 1982 BAH [5], 1994 the respondent agrees with the relative success of the
Mercer Management/R&D Magazine [45] and the program on a 1 (completely disagree) to 9 (completely
Arthur D. Little surveys [1]. A multiplicity of sources agree) scale (5 is neutral, neither agree nor disagree).
was used because this research was to cover a broader These two items are highly correlated in the data ( r 5
set of issues than any one of these previous surveys .76, p , .01) and a scale constructed by averaging the
covered. two items has an a of .86. The 4 criteria measuring
Potential questions to be included in the survey market and financial success (% of products catego-
were identified and formatted. This set of questions rized as successes in the last five years, $ sales of
was then circulated to three academics and three prac- products commercialized in the last five years as a %
titioners who had previously been involved in best of total sales, $ profits of products commercialized in
practices research. Changes were made in a number of the last five years as a % of total profits, and % of
questions to increase their clarity and several issues products categorized as financial successes in the last
and questions were eliminated based on their sugges- five years) are linear in nature, measured as %’s, with
tions. Another draft of the survey was developed and a maximum range from 0% to 100%. These measures
pretested with twelve knowledgeable product devel- are from prior NPD best practices and success mea-
opment practitioners for clarity and usefulness of is- surement research [31,32,50]. These four items are all
sues addressed. Additional rewordings of the survey statistically significantly correlated in the data. The
were made based on their suggestions to increase the scale constructed from the average of these four items
clarity of a few questions. has an a of .79. As Table 2 illustrates, overall industry,
The final survey consisted of 9 pages of questions relative and market-financial success are significantly
and a 1-page cover letter. Questions covered issues correlated in the data. However, the correlations ac-
surrounding the product development process, orga- count for only about 1/4 of the variability in the data
nizing for product development, tools supporting (the square of the correlation coefficient). Although
product development, measuring product develop- correlated, each multi-item measure represents a
ment, product development outcomes, and background slightly different aspect of success.
information on the respondents. The questions inves- Process and Organization Variables. The ques-
tigated in the analyses presented here come from the tions investigating strategy, product development pro-
product development process, organization, measure- cess, and organization may be found in the Appendix.
ment, outcomes, and respondent background sections. They covered whether strategies and processes had
The survey was a pre-addressed self-mailer. Re- been used, and how long, of what steps the process
spondents folded the finished survey in half, taped or
stapled it shut, placed a stamp in the corner and Table 2. Correlation Matrix for Success Variables (r)
dropped it in the mail. No incentives were provided to
participants for filling out this long survey. However, Overall Success Relative Success
if they stapled a business card to the form, we prom- Relative Success .50** —
ised to send working papers derived from the research Market-Finance .45** .42**
as they became available. ** p , .01.
1997;14:429 – 458
consists, product idea mortality, how NPD is orga- sample diversity. Previous studies had shown that
nized and led, multi-functional team usage, and team PDMA members are primarily manufactured goods
and leader rewards for NPD. producers in the business-to-business market, from
Demographic Variables. The survey also gath- somewhat larger firms [25,50]. Table 4 shows that the
ered information on several pieces of demographic 1995 PDMA sample is similarly characterized. The
information. Respondents indicated whether their firm percentage of manufactured goods has increased from
primarily provided manufactured goods or services to 78.8% in the 1990 sample to 91.0% in the 1995
consumer or business purchasers, or to a mix of both PDMA sample. The 1990 PDMA sample was 34.8%
types of purchasers. They indicated whether their firm high-tech, 34.8% mixed, and 30.4% low-tech. Thus,
is high-tech, low-tech, or a mix of both, and checked the 1995 sample is slightly less high-tech than previ-
boxes according to industry participation and firm size. ously. The 1995 sample also consists of a somewhat
smaller set of firms. The median sales for the sample
Sample is under $500 million, whereas in 1990 median sales
The survey was sent out to over 14,500 potential U.S. were $1 billion. Because summary statistics are not
respondents obtained from three separate sources as available for the PDMA membership, it is not known
outlined in Table 3. The cover letter stressed the need if the respondent sample for this survey differs from
for a respondent knowledgeable with the scope of the PDMA population in any meaningful way.
NPD activities and requested that the recipient pass the Because the goal was to get a broader understanding
survey to a more qualified person, should they not feel of NPD practices, lists were sought which would pro-
equipped to provide the breadth of information re- vide more consumer, service-providing, and smaller-
quested. Reminder postcards were mailed to all poten- firm respondents. Table 4 shows that the diversity of
tial respondents two weeks after the initial survey was the sample was increased statistically significantly
sent out. The overall response rates obtained are low across four demographic variables by using the AMA
for survey research. However, this is not unexpected and CorpTech lists in addition to surveying PDMA
due to the length of the questionnaire, complexity of members. The AMA sample provided an increased
some of the questions, and lack of incentives provided number of consumer and service respondents. The
for taking the time to respond. CorpTech sample increased the number of high-tech
The PDMA members to whom the survey were sent and small firms in the sample. The sample is still more
are a sub-sample of the PDMA population. The sur- manufactured goods-oriented than service-oriented,
veys were sent out only to non-academics or service with 80.6% of the respondents indicating that their
providers (consultants) to obtain data from people firms primarily produce manufactured goods, and pre-
associated with firms actually involved in product dominantly sell into the business-to-business market
development on an ongoing basis. In addition, random rather than directly to consumers. The CorpTech list
samples of two mailing lists were purchased from the helped increase the numbers of smaller firms in the
American Marketing Association (AMA) and Corp- sample.
Tech to obtain a significant sample size of responses. Average success rates differed statistically across
The AMA sample came from those in the database the three samples for one of the measures used, as
who had checked off the “new product development” Table 5 illustrates. Respondents from the CorpTech
interest category. The CorpTech sample consisted of sample indicate that they achieve higher percentages
those with “business development,” “product develop- of market and financial success in the products they
ment,” or “development” in their titles. commercialize than respondents from the other two
The mailing lists were chosen to maximize the groups. Importantly, even with this difference, all
Table 3. Sample and Response Rates
PDMA Members American Marketing Association CorpTech List Total

# Mailed Out 1,601 6,650 6,500 14,751
# Undelivered 0 225 57 252
# Usable Returns 159 86 138 383
Response Rate 9.9% 1.3% 2.1% 2.7%
1997;14:429 – 458
Table 4. Demographics by Source of the Sample
PDMA Members AMA Sample CorpTech Total Sample

# % PDMA # % AMA # % CorpTech # % Total
Technology Base1 High-Tech 43 27.7% 18 22.2% 67 49.3% 128 34.4%

Mixed 64 41.3% 41 50.6% 48 35.3% 153 41.1%
Low-Tech 48 31.0% 22 27.2% 21 15.4% 91 24.5%
Market2 Consumer 39 25.0% 20 24.7% 7 5.2% 66 17.8%
Mixed 28 17.9% 14 17.3% 22 16.4% 64 17.3%
Business 89 57.1% 47 58.0% 105 78.4% 241 65.0%
Product Type3 Goods 142 91.0% 39 48.8% 119 87.5% 300 80.6%
Services 14 9.0% 41 51.3% 17 12.5% 72 19.4%
Sales4 ,$10 M 6 4.0% 12 15.3% 44 33.6% 62 17.0%
$10–$24 M 6 4.0% 7 9.0% 28 20.7% 41 11.2%
$25–$99 M 18 11.8% 19 24.4% 43 31.9% 80 21.9%
$100–$499 M 63 41.4% 22 28.2% 16 11.9% 101 27.7%
$500–$999 M 18 11.8% 6 7.7% 2 1.5% 26 7.1%
$$1,000 M 41 27.0% 12 15.4% 2 1.5% 55 15.1%
1
Anova test: F 5 11.1, p , .01, df 5 371.
2
Anova test: F 5 12.4, p , .01, df 5 370.
3
Anova test: F 5 40.5, p , .01, df 5 371.
4
Anova test: F 5 79.1, p , .01, df 5 364.
three groups of respondents are undifferentiated in tions and financial services firms frequently serve both
their average industry success level and with overall markets rather than just one or the other. However,
NPD program performance, relative to their programs’ NPD performance is independent of product type, as
objectives. This means that while the CorpTech re- the last three lines of Table 6 show. This simplifies the
spondents have higher performance on one dimension, analysis in identifying practices which are found to
they also expect more in that dimension. differ in achieving success across these two sub-sam-
The statistically significant demographic differ- ples.
ences of the different sample populations are predom-
inantly captured in the market and manufactured good/ Analysis
service demographic variables, as shown in Table 6.
The only demographic difference across manufactured Best practice publications in the recent past have pre-
goods and service producers is the percentage of con- sented aggregate results for broad populations of
sumer versus business-to-business firms in the sample. firms, or for firms within one particular context. While
Over 2⁄3 of the manufactured goods producers serve analyzing aggregate results may suggest NPD prac-
only business-to-business customers, while less than tices associated with success independent of context,
half of the service firms serve only them. Service firms firms in different industries or managerial contexts
are more than twice as likely to serve both consumer may require additional or different practices to maxi-
and business-to-business customers than manufactured mize the probability of success. For example, manu-
goods producers. For example, both telecommunica- factured good producers differ from service providers,
Table 5. Average Success by Source of the Sample
PMDA Members AMA CorpTech Total Sample

Industry Success* 2.6 2.6 2.7 2.6
Relative Success† 5.5 5.3 5.7 5.5
Market-Financial Success1 40.1% 42.2% 50.5% 44.6%
* 4 5 most successful in industry, 3 5 top third, 2 5 middle third, 1 5 bottom third.
† 9 5 completely agree, met criteria, 1 5 completely disagree, met criteria (higher is better).
1
Higher percentages are better. Anova test: F 5 7.1, p , .01, df 5 267.
1997;14:429 – 458
Table 6. Descriptive Statistics of the Sample
Manufactured Goods
Service Firms Firms All Firms
# % of Service # % of Product # % of Total
Technology Base High-Tech 24 33.8% 103 35.3% 127 35.0%

Mixed 33 46.5% 119 40.8% 152 41.9%
Low-Tech 14 19.7% 77 26.4% 91 25.1%
Market1 Consumer 16 22.5% 50 17.1% 66 18.2%
Mixed 22 31.0% 42 14.4% 64 17.6%
Business 33 46.5% 206 70.6% 239 65.8%
Sales ,$10 M 16 22.5% 45 15.4% 61 16.8%
$10–$24 M 4 5.6% 37 12.7% 41 11.3%
$25–$99 M 14 19.7% 65 22.3% 79 21.8%
$100–$499 M 15 21.1% 86 29.5% 101 27.8%
$500–$999 M 6 8.4% 20 6.9% 26 7.2%
$$1,000 M 12 16.9% 42 14.4% 54 14.9%
Sample Total 71 100% 292 100% 363 100%
2
Success Industry 2.5 — 2.7 — 2.6 —
Relative 5.5 — 5.5 — 5.5 —
Market-Financial 39.6% — 45.8% — 44.8% —
1
Statistically different percentages across cells by Anova. F 5 7.77, p , .01, df 5 388.
2
A higher number is higher success.
at least in terms of the linkage between the manufac- the top third of their industry for NPD success, and
ture and delivery of the service or good [47]. One also were above the mean on relative success and
might thus expect that service firms could be helped market-financial success for the entire sample. A total
by slightly different practices than goods-producing of 85 respondents, 22.2% of the sample, met all three
firms. of these criteria. “The Rest” are the other 298 re-
This research purposely sought responses from sponses in the sample, who fell short of the mean on at
firms across a broad set of industrial contexts so that least one multi-item criteria or who were below the top
contextual differences in NPD practices could be bet- third in their industry.
ter understood. This article thus analyzes results, not Table 7 shows how “The Best” responses are dis-
just in the aggregate, but also across the manufactured tributed across the product and market demographics.
good/service segments, when appropriate. Both de- Although more of the goods-producing firms than
scriptive trends and associations with higher and lower service-delivering firms (both absolute and as a per-
performance are included in the analysis. centage of the total available) meet the three Best
For assessing “best practice,” the sample was split Practice, criteria, Anova tests of the distribution across
into two groups based on success performance across these demographic variables show no statistical differ-
multiple criteria. “The Best” are those firms who in- ences for the percentage of the total by product, mar-
dicated that they were either the most successful or in ket, or product and market together.
Table 7. Distribution of “The Best” by Market and Product Type
Consumer Market Mixed Business-to-Business Full Sample

# % of Total # % of Total # % of Total # % of Total
Services 2 12.5% 3 13.6% 5 15.2% 10 14.1%

Goods 13 26.0% 9 21.4% 53 25.7% 75 25.2%
Full Sample 15 22.7% 12 18.8% 58 24.3% 85 22.2%
1997;14:429 – 458
The data are analyzed using several methods.

Means, means across sub-groups and Anova tests are
used for continuous variables. Cross-tabulations and
Chi-squared tests are used to analyze differences for
discreet variables. These analyses investigate the rela-
tionship between individual variables and success. In-
teraction effects will be investigated in later analyses
[41,42].
Results
Figure 2. Product Development Processes: The Best versus
Product Development Process Impact on NPD The Rest
Using a formal NPD process and not skipping steps in

the process has long been a differentiating factor be- and third generation (those with fuzzy and flexible
tween successes and failures at the project level of gates [12]) stage gate processes. Mercer [46] found
analysis [10,15]. However, even though the effective- that 81% of the high performers in their sample used
ness of product development processes has been well- stage gate processes, but only 68% followed them
proven, many firms still do not use a formal NPD consistently, a percentage very close to that found in
process. Across the entire sample, 38.5% of the re- this sample. Only 56% of the lower performers in
spondents (140) still either use no process at all, or use Mercer’s study used stage gate processes for NPD, and
only an informal process. As Figure 1 shows, how- only 38% followed them consistently. Of the PDMA
ever, manufactured goods producers are statistically best practice firms, 22 of the 85 were able to achieve
more likely than service providers to have and use a high NPD performance in spite of not following a
formal NPD process (x2 5 22.3, p , .01, df 5 6). formal NPD process. However, of those 22, only 4 had
Nearly 60% of the service firms responding to this no process at all—the other 18 informally follow some
survey do not use a formal process for NPD. In addi- procedure.
tion, very few firms now use a functional, sequential The firms in this sample using processes have been
process. Those who use processes have migrated to using them for longer periods of time than other stud-
some form of multi-functional stage gate approach. ies have found, as shown in Figure 3. While the vast
Figure 2 shows that the extent of process use also majority of firms have used their NPD processes for
differs statistically between best practice firms and the between 1 and 4 years, over 31% of the total sample
rest of the population (x2 5 14.4, p , .05, df 5 6). has used their process for at least five years. Over 27%
Multi-functional stage gate approaches are used by of the service firms in this sample have used a process
69% of the best NPD firms, but only 52% of the rest. for at least five years, which is a contrast to the
Differences in use across these two groups are espe- Mitchell finding that none had used a process that long
cially apparent when looking at the use of facilitated [47]. Of the best practice firms, over 40% have used
their process for at least five years. This is consistent
with the 1982 BAH findings for Best practice firms.
Figure 1. Product Development Processes: Services versus

Manufactured Goods Figure 3. Years Process Has Been in Use
1997;14:429 – 458
A constant theme which has run throughout NPD processes include each of the 9 steps described in the
best practices reports is the need for a NPD strategy, Appendix (Question 5). The development and test and
both at the overall program level and for each project validation steps are the most-included NPD process
undertaken [1,4,5,33,37,46]. Although Booz, Allen steps. The processes for higher-performing respon-
and Hamilton espoused the need for a strategy in 1982, dents are always more likely to include any particular
and encouraged firms to set one prior to undertaking step than the processes for lower-performing respon-
NPD, Page reported that only 56.4% of his sample had dents. For each step, with the exception of the Busi-
a specific NPD strategy in 1990 [50]. Strategy use has ness Analysis step, these differences are statistically
increased since then. Of the 383 responses in this significant, as analyzed using Anova.
research, 62.7% (240), “have a specific strategy for Overall, the processes for higher performing re-
their NPD activities which directs and integrates the spondents average 6.2 of the 9 possible steps, while
entire new product program.” Mercer [46] found that lower-performing respondents only average 4.7, again,
having a strategy for the NPD program as a whole was a statistically smaller number (Anova: F 5 10.0, p ,
an important differentiator between higher and lower .01, df 5 363). Previous research has demonstrated
performance (68% in higher performers and 43% in that not skipping steps increases the probability of
lower performers). That finding is replicated here: a success for any project [10,15]. The PDMA 1995
statistically higher percentage of the best-performing research provides additional support that processes
respondents, 75.9%, have specific strategies, while which consist of more of the total set of activities
only 58.8% of the lower performers have one (Anova: required for product development are associated with
F 5 12.1, p , .01, df 5 362). There are no differences firms with higher overall performance for their prod-
in having an overall NPD strategy between service and uct development programs.
manufactured goods producers. NPD processes for services differ from the pro-
In 1982, Booz, Allen and Hamilton reported that cesses used for developing manufactured goods, as
77% of firms they surveyed started their product de- Figure 5 illustrates. NPD processes for services are
velopment process with a strategy developing step for always less likely to contain any particular step. As
each project [5]. On average, the NPD processes for with manufactured goods, “development” is the most-
55.6% of the 1995 PDMA sample includes a specific included process step. However, the step next most
NPD strategy. However, as Figure 4 illustrates, 70% likely to be included is concept generation, not test and
of the higher performing respondents have a process validation. Service processes are as likely as goods-
which includes a strategy step, versus only 51% for the developing processes to include a concept generation
lower-performing respondents. These differences in phase. Service processes also consist of a statistically
strategy as a step in the NPD process are statistically significantly fewer number of steps than manufactured
significant (Anova: F 5 9.6, p , .01, df 5 363). goods processes. Service processes, on average, con-
Figure 4 also illustrates what percentage of the NPD sist of 3.8 steps, versus 5.4 for the average manufac-
Figure 4. NPD Process Components Figure 5. NPD Process Components

1997;14:429 – 458
tured goods-developing NPD process (Anova: F 5 manently-staffed new products department is charged
9.8, p , .01, df 5 360). with the responsibility of recommending new product
While, on average, service processes are less likely objectives, planning the programs, making screening
to include any step than are manufactured goods pro- decisions, and directing the progress of projects
cesses, these differences almost completely go away through all stages of development. The full-time re-
when service and goods differences are analyzed for sponsibility of people in this department is new prod-
just “the Best” firms. With one exception, processes uct development, and nothing else. Relegating NPD
for the best service firms are as likely to include a responsibility to a new product committee is a pre-
particular step as they are for manufactured goods cursor structure to the new product department. This
firms. Service firm processes statistically still are less committee is charged with evaluating and coordinating
likely to contain a manufacturing development step. new products at a firm, however, the personnel on this
Only 18% of “the Best” service processes include this committee fulfill this task only part time. They have
step versus 73% for manufactured goods developers other primary tasks and responsibilities.
(Anova: F 5 14.4, p , .01, df 5 83). The total number Souder [54] identified several other organizational
of steps, 4.7, in a NPD process for “the Best” service structures for NPD. Although he found that a func-
firms, while still less than the 6.5 for “the Best” tional structure was superior for handling routine prob-
manufactured goods firms, is not statistically lower lems, well-known technologies, stable environments,
due to the smaller sample sizes in the two subgroups. low product evolution rates, and well-defined markets,
Even if a best practice service NPD process need not others have found that projects developed using func-
include a manufacturing development step because tional structures were less successful than those devel-
there is no manufacturing development, service pro- oped in matrix structure or by a multi-functional
cesses still are nearly one full step less than manufac- project team [39]. In Souder’s research, a divisional or
tured goods processes. strategic business unit (SBU) structure in which each
Summary. NPD processes and process use have division is responsible for commercializing new prod-
continued to evolve over the last five years. Change ucts was able to respond to diverse market needs and
seems to occur more evolutionarily, however, rather evolutionary product improvements. A third structure
than revolutionarily. Furthermore, the ways in which again organized current products into SBU’s, but pro-
processes are implemented continue to differentiate vided for fostering the development of radical innova-
between best practice and the rest of the firms and tions using a new enterprise division, or venture group
between service and manufactured goods producers. whose task was to nurture the risky projects into the
In the best firms, processes are more likely to be used, market place.
and when used, are likely to be more complex and A more recently-developed way for the responsibil-
complete. Additional effort needs to be expended to ity of NPD to be assigned in the organization is by
develop processes which better meet the needs of creating a process-oriented NPD structure, with one or
providers of services. more NPD process owners [27,44]. NPD process own-
ers are responsible for developing a formal process for
Organizing for Product Development NPD, documenting and improving the process, and
facilitating process deployment and use across the
How best to organize NPD, where to locate responsi- firm. This structure is designed to help ensure that
bility for NPD in the firm, and who should lead NPD NPD procedures and methods are implemented in
projects have long been questions to which firms standard ways across projects and divisions.
would like answers. Both in 1968 and 1982, BAH The survey listed these six organizational structures
found that more than half of the firms they surveyed and requested that respondents indicate all of the or-
used multiple structures to organize for NPD [4,5], as ganizational structures in which responsibility for their
did Page in 1990 [50]. Souder [54] recommended that firm’s more innovative NPD projects are found. Fig-
the most appropriate structure depended upon the level ure 6 displays the overall results, starting from the
of innovation desired and the stability of the market most frequently-used structure on the left side of the
and technical environments. graph, as well as how responsibilities differ depending
In 1968, BAH identified two basic organizational upon whether the firm produces manufactured goods
forms which had evolved specifically to meet the or services. This graph repeats the long-standing find-
needs of new product development programs. A per- ing that multiple structures are used for organizing the
1997;14:429 – 458
marketing function, and least likely to report to an

engineering function, when NPD takes place in a func-
tional structure. It is unclear whether this difference is
due to cultural differences between the two types of
firms or because service firms do not have engineering
functions.
NPD reporting structures for best practice firms do
not differ statistically from reporting structures for the
rest of the firms. While the Best are less likely for
NPD to report into a function or to use process owners,
these differences are not significant (Anova: p . .1).
Reporting structure does not contribute materially to
Figure 6. Organizations for NPD
differentiating the Best from the Rest.
Another important question addressed in this survey
NPD effort within each firm. Page [50] reported that was who leads NPD and does who leads NPD differ
53% of the firms surveyed in 1990 used more than one depending upon whether the firm produces manufac-
structure, and on average, each firm used 1.9 structured goods or services, or is a best practice firm or
tures for organizing more innovative NPD projects. In not? Mercer [46] found that nearly 80% of the higher-
1995, 61.9% of the firms surveyed organize more performing firms in their survey pinpointed specific
innovative NPD projects in multiple ways, with an individuals responsible for NPD and product introduc-
average of 2.0 structures per firm. However, manufac- tion, while only 55% of lower-performing named
tured goods producers use more structures on average, those responsible. PDMA’s 1995 survey asked respon-
at 2.0 per firm, than service providers, who only use dents to indicate how the more innovative projects are
1.7 (Anova: F 5 6.2, p , .05, df 5 371). Service firms led at their firms. The survey listed five options.
statistically are less likely to have NPD organized at Product development can be led in a number of
the SBU level. Permanent NPD staffs and NPD pro- different ways. Project managers coordinate develop-
cess owners are the most frequent reporting structures ment activities through liaison representatives from
for service organizations, although the differences in each function, in the case of lightweight project man-
use are not statistically higher than for manufactured agers, or have full responsibility and “clout” to man-
goods-producing firms. age development activities across functional areas, in
The most-used structure for more innovative prod- the case of heavyweight project managers [6,9]. The
uct development projects is to have NPD report to a survey did not differentiate between whether the
functional area. As Figure 7 illustrates, the functional project manager was a lightweight or heavyweight
area in which NPD resides is almost equally likely to manager. A second leadership mechanism is the
be marketing, R&D or engineering. NPD reports into project champion, which Crawford [19] defines as: “A
planning only seldomly. Figure 7 also shows how person who takes an inordinate interest in seeing that
NPD structures differ between services and goods. a particular process or product is fully developed and
Service development is most likely to report to the marketed. The role varies from situations calling for
little more than stimulating awareness of the opportu-
nity to extreme cases where the champion tries to force
a project past the strongly entrenched internal resis-
tance of company policy or that of objecting parties.”
The champion’s role in new product development has
been a topic of discussion for over 30 years, with
much anecdotal evidence suggesting they materially
contribute to product development success [41]. An-
other source of NPD leadership may come from the
NPD process owners, in a firm that uses a process-
oriented reporting structure [43]. Process owners are
responsible for implementing and maintaining a suc-
Figure 7. The Function in Which NPD Resides cessful NPD process in an SBU or company [27].
1997;14:429 – 458
However, in some firms, these owners may also be

responsible for leading NPD projects. According to
the practitioners pre-testing the survey, some firms are
also experimenting with “leaderless” structures. These
structures involve work processes which do not re-
quire formal leadership. Finally, respondents were
given the option of indicating that no one leads NPD
at their firm.
Figure 8 shows that project managers are by far the
most widely-used leaders for NPD, followed by
project champions. Project managers have become the Figure 9. Who Appoints NPD Project Leaders
most-used mechanism for leading NPD projects.
Championing projects has been remarkably stable over
the last 10 years. BAH reported that nearly half of the selection method is that the function in which NPD
firms they surveyed encouraged champions [5]. In resides dictates who the team leader shall be. Teams
Page’s respondents, 43.4% of the firms encouraged and peers choose their own leaders less than 10% of
champions [5]. This is close to the 42.4% leadership the time. Leader selection probabilities do not differ
by champion number obtained here. between manufactured goods and service sub-groups
While in over 30% of these firms, product develop- or best practice and the rest of the firms.
ment resides in a process owner group (Figure 6), only The survey instrument also gathered data on re-
slightly more than 10% of the projects are led by NPD warding NPD personnel, another area on which sev-
process owners. In other words, process owner NPD eral best practices investigations have focused. In
structures generally operate more like staff groups 1982, BAH found that only 5% of the firms in their
than line organizations responsible for implementing survey tied compensation directly to new product per-
projects. Leaderless projects still are very experimen- formance [5]. In 1990, Page [50] found a slight in-
tal, with fewer than 5% of the sample indicating that crease in the proportion of firms, to 7.4%, who tied
they use them for their more innovative NPD projects. compensation plans (through the bonus system) di-
There are no statistically significant differences in rectly to successful performance of the new product.
the proportions of who leads NPD between service or Another 15.9% tied part of a person’s bonus to
manufactured goods firms, or best practice firms and whether the NPD project was completed. Only 20.6%
the rest of the sample. This finding differs from the of the 1990 sample used incentives and awards as
Mercer study, where champions were associated with NPD personnel motivational tools, with 15.8% provid-
63% of the higher-performing firms, but only 41% of ing financial incentives or awards, and 10.0% provid-
the lower-performing firms [46]. ing non-financial awards and recognition. In 1993,
Leadership selection is clearly not an egalitarian Kuczmarski [37] found that recognition (pats on the
process (see Figure 9). Project leaders, whether project back) was used by nearly 90% of the respondents in
managers or champions, are overwhelmingly ap- their survey, while awards and plaques were used by
pointed by management. The second most frequent 43%, bonuses based on new product performance
were used by about 15% of the firms, and project-
based stock was used by less than 5%. In an extensive
survey of the 1994 compensation levels for new prod-
uct professionals, Feldman found that performance-
based financial incentives played a minor role overall
[26]. At the lower ranks of the organization, less than
10% of total compensation resulted from NPD perfor-
mance-based financial incentives. At the vice presi-
dential level, 20% of total compensation could be
derived from NPD performance. Feldman found that
about 76% of NPD professionals received non-finan-
cial incentives. Of these, 38.4% received a plaque or
Figure 8. Who Leads NPD Projects certificate, 36.7% received completion dinners or
1997;14:429 – 458
award dinners, and 12.3% received some sort of mer- projects, than do other firms, which only use them
chandise commemorating project completion. 62.2% of the time (Anova: F 5 5.6, p , .05, df 5
The results presented in Figure 10 uphold the gen- 260). When considered in aggregate, best practice
eral trends found by Page [50], Kuczmarski [37], and firms do not use multi-functional teams as extensively
Feldman [26]. Project-based financial rewards still are as previously reported.
seldom used in NPD. Most interesting is that none of Larson and Gobeli [39] found that more complex
the best practice firms use either project-based profits projects were more successful when project teams
or stock as rewards. The most frequent source of were used in development. When multi-functional
reward for NPD is the completion dinner, where the team use is investigated by project type, the story
NPD team goes out for a final meal with the firm changes somewhat, as illustrated in Figure 11. Multi-
picking up the cost. This reward mechanism is the functional team use is much higher for more innova-
only one which best practice firms use to a statistically tive projects than for less innovative projects. Over
significantly higher extent than the rest of the firms. 84% of all new-to-the-world, new-to-the-firm and ma-
Overall, rewards for NPD are much more likely to be jor revisions use multi-functional teams for NPD. For
what is referred to as “off W-2 compensation,” or these projects there is no statistical difference in use
compensation which does not increase the employee’s between best practice firms and the rest of the firms. In
reported earnings. Best practice firms do more of all of Page’s 1990 results 76.2% of the companies surveyed
the forms of off W-2 compensation, but with the used multi-functional teams for their more innovative
exception of completion dinners, the differences are NPD projects [50]. Thus, multi-functional team use for
not statistically significant. There is no difference in more innovative projects has increased over the last
the reward and recognition patterns between goods five years.
producers and service providers. Figure 11 also shows that there is less overall multi-
The final organizational question addressed in this functional team use in less innovative projects such as
article is the extent to which multi-functional teams incremental improvement, repositioning, and cost re-
are used in NPD. Mercer [46] found that well over duction projects. Overall, only about 45% of these
two-thirds of their respondents used multi-functional projects are completed using multi-functional teams.
teams for product development, with little difference However, multi-functional team use is statistically
in use between higher-performing (78%) and lower- higher for best practice firms than for the rest of the
performing (66%) firms. On average, the 1995 sample firms for each of these project types. Best practice
shows slightly lower overall use of multi-functional firms use multi-functional teams in the majority of all
teams. Only 64.0% of all projects use multi-functional NPD projects, regardless of level of innovativeness.
teams. However, best practice firms use multi-func- The rest of the firms are far less likely to use multi-
tional teams more extensively, at 70.7% of the total functional teams for less innovative projects.
Figure 10. Team Leader and

Member Rewards
1997;14:429 – 458
while only 28% of the lower-performing firms track

NPD performance.
In this 1995 sample of firms, 75.6% develop formal
financial objectives against which actual performance
will be evaluated. While a slightly larger proportion of
best practice firms (83.9%) develop formal financial
objectives, the difference in proportions across the two
groups is not statistically different. However, even
though objectives are set, firms do not always go back
and evaluate actual performance. Best practice firms
only assess performance against objectives in 63.2%
Figure 11. Percentage of Projects Using Multi-Functional of the projects, while the rest of the firms assess
Teams performance less than 50% for the time (48.0%), with
the difference being statistically significant (Anova:
F 5 4.0, p , .05, df 5 149). On average, when they
Summary. Organizationally, changes in NPD also evaluate projects against objectives, the projects are
are occurring more by evolution than by revolution. reviewed about 16 months after initial introduction.
However, organizational issues result in far fewer dif- There is no difference in this time frame across prod-
ferences in success outcome than do process issues. uct type or whether or not the firm is a best practice
Best practice firms use multi-functional teams more firm.
extensively, especially for less-innovative projects, Best practice firms are more likely than the rest of
and provide more completion dinners as rewards for the firms to set a target for the portion of revenue
NPD. In addition, there are almost no organizational growth to come from new product development.
NPD practice differences between across service and Nearly 2⁄3, 64.7%, of the best practice firms set reve-
product producing firms. NPD in service firms is less nue growth targets, with the average goals of these
likely to report to SBU leadership or to the engineering firms being 45% of sales to come from products com-
function, and is more likely to reside in the marketing mercialized in the last three years. Only 46.5% of the
function. This research demonstrates that one “best rest of the firms set these targets, with the average goal
way” to organize NPD has not been found. being to derive 25% of sales from products commer-
cialized in the last four years. Each of these differ-
ences is statistically significant (p , .05). There are no
Product Development Measurement and differences across manufactured goods and services.
Expectations Best practice firms also have higher expectations for
future NPD performance than the rest of the firms. In
In some senses, “you are what you measure.” Unless a the next five years, best practice firms expect 53.5% of
firm measures NPD performance, it cannot determine their sales and 56% of their profits to come from new
either how well they are doing, or whether they are products. The rest of the firms only expect 37.6% of
improving or declining in NPD performance. One their sales and 37.0% of their profits to come from
interesting finding from the 1990 PDMA best prac- new products over the same time period. These differ-
tices study was that several of the criteria used to ences are statistically significant (p , .01).
measure NPD performance were associated with The vast majority of firms now measure NPD per-
higher NPD performance [51]. However, not all firms formance routinely, however, even the best practice
measure NPD performance. BAH claimed that 2⁄3 of firms do not consistently measure it across all projects.
their sample measured product development in 1982 Best practices seem to be associated with both higher
[5]. In 1990, Page found that 76.2% of his respondents levels of measurement and higher expectations for
developed formal financial objectives against which NPD performance.
each NPD project is evaluated. Since then, two other
groups have reported slightly lower findings. South- Outcomes from Product Development
western Bell found that only 56% of their respondents
measured NPD [24]. Mercer [46] reported that 50% of As the previous sections have shown, firms have made
the higher-performing firms track NPD performance, a number of changes in their product development
1997;14:429 – 458
programs over the last 15 years. But, have organiza-

tions gotten any better at developing new products?
This section looks at several aspects of NPD out-
comes, compares current results to past results, and
compares best performers to the rest of the sample.
The first comparison looks at how NPD success
rates have changed. In a summary of research from the
1960s and early 1970s, Crawford found that success
rates for consumer goods varied from 58% to 73%,
averaging 64% across five studies [17]. Success rates
in two studies of industrial goods were slightly higher,
ranging from 74% to 80% [17]. Earlier best practices
surveys found that around 2⁄3 of all new goods and Figure 12. Product Introductions by Project Type
services met their original criteria when commercial-
ized. Specifically, in BAH’s predominantly industrial
product sample, 67% of the projects met them in 1968 portfolio commercialized 13 years ago, with relatively
[4], and 65% met the original criteria in 1982 [5]. By comparable success rates. There is no difference in the
1990, the first PDMA study found that only 55% of portfolio structure between the Best and the Rest of the
new products were successful in the US. In a separate firms.
study, 54.3% of European products were successful Table 8 presents average values for the four specific
when commercialized, as were 59.8% of Japanese items which make up market-financial success for the
products [23]. Two of the studies published between whole sample and for the Best and the Rest. Also
1990 and 1995 recorded success rates. In a broad presented are averages by product type, technology
sample of industries, the success rate was found to be base, market served, and annual sales. Average suc-
56.5% [37]. However, only 38% of consumer pack- cess rates and profit success rates do not vary across
aged goods were successful [33]. In 1995, the average any of the partitions (except between the Best and the
success rate across the sample is 59%. The average for Rest). Firms experience market success rates ranging
the Best practice group, however, is 82.7%. Overall, from 53% to 61% and profit success rates from 49% to
these results suggest that success rates for a broad US 58%. However, NP sales and profits as a percentage of
industry sample, while they have declined a bit since total do differ by product type, technology base and
the early 1980s, have remained in a range between annual sales. Manufactured goods, high-tech, and
55% and 60%. Success rates in Europe and Japan are smaller firms all have higher NP sales and profits as a
probably comparable, but success rates within the con- percentage of the total. Firms may use these figures to
sumer packaged goods industry may be lower. There benchmark their own NPD performance on these suc-
is no difference in success rate across product type cess measures, given their own competitive situation.
(service or manufactured good). The second NPD outcome comparison presents
These success rates have been achieved through changes in new product idea mortality curves. The
commercializing a balance of project types which is mortality curve represents the progressive rejection of
only slightly different from the portfolio reported in ideas or projects through stages of the new product
1982 [5]. Figure 12 shows that the percentages of both development process. Figure 13 presents mortality
new-to-the-world and new-to-the-firm projects are un- curves by product development project phase for
changed over a 13-year period. In recent years, firms 1968, 1982, 1990, and 1995. BAH started publicizing
undertake fewer repositionings, cost reductions and mortality curves in their 1968 report [4]. At that point
line extensions, preferring to focus more on improving in time, 100 ideas lead to only 1.7 commercially
current product performance. The proportion of successful products. Turned around, it took 58 new
projects which are improvements of previously-com- product ideas to generate just one successful commer-
mercialized goods and services is 35% higher than it cial product. By 1982, it took only seven ideas to
was 13 years ago. The cost reduction and repositioning generate one successful product, or 100 ideas led to
projects are generally considered the least innovative 14.3 commercially successful products [5]. Page [50]
NPD projects. On balance, then, the portfolio of found that the mortality curve had eroded a bit in
projects seems to be slightly more innovative than the 1990. In his sample, 11 new product ideas were re-
1997;14:429 – 458
Table 8. Average Success by Demographic Category
Success Rate1 % Profit Success2 NP Sales %3 NP Profit %4

Full Sample 59.0% 54.6% 32.4% 30.6%
The Best 79.8%** 78.0%** 49.2%** 49.2%**
The Rest 52.5%** 47.1%** 25.2%** 22.0%**
Product Type
Manufactured Goods 59.6% 55.3% 34.0%* 32.4%*
Services 58.2% 52.7% 24.1%* 21.7%*
Technology Base
High-Tech 60.5% 56.5% 42.3%** 38.8%**
Mixed 60.0% 55.3% 28.7%** 26.9%**
Low-Tech 55.2% 50.3% 23.7%** 24.5%**
Market Served
Consumer Products 58.1% 53.2% 36.2% 32.9%
Mix of Both 60.8% 55.2% 24.9% 23.1%
Business-to-Business 58.6% 54.5% 33.4% 32.0%
Annual Sales
#$24 million 62.3% 56.8% 40.7%** 37.2%**
$25 to $99 million 60.7% 57.4% 33.2%** 35.2%**
$100 to $499 million 60.9% 56.0% 28.6%** 27.6%**
$$500 million 53.0% 48.9% 23.7%** 18.8%**
1
Success Rate: % of products commercialized in the last 5 years categorized as successes.
2
% Profit Success: % of products commercialized in the last 5 years categorized as financial successes.
3
NP Sales %: $ sales of products commercialized in the last 5 years as a % of total sales.
4
NP Profit %: $ profits from products commercialized in the last 5 years as a % of total profits.
** Anova test: p , .01.
* Anova test: p , .05.
quired to generate one success (100 ideas lead to 9.4 words, while the same number of ideas are creating
successes). The 1995 sample shows a slight improve- one success, projects are eliminated much earlier in
ment over BAH’s 1982 results—100 ideas leads to the NPD process, where less time and money has been
15.2 successes (one success for every 6.6 ideas). How- spent on any particular idea [4,50,54]. Today’s port-
ever, the important difference between 1982 and 1995 folios of NPD projects are wasting less money on
is in the shape of the curve. While the 1982 mortality unsuccessful projects. This is a clear improvement in
curve declines slowly in the beginning, with the larg- outcomes over previous reports.
est number of projects weeded out in the development Best practice firms are more efficient, according to
stage, current practices weed out the largest number of mortality curve analysis, than the rest of the sample, as
projects in idea screening, with the second largest Figure 14 illustrates. At each stage of the NPD pro-
number eliminated in business analysis. In other cess, it takes statistically fewer projects to create one
commercial success than for the rest of the firms. For
best practice firms, the vast majority of projects are
eliminated from the process prior to entering the de-
velopment stage. For every three projects which enter
development in best practice firms, two go on to
become commercial successes. Development is the
stage where large NPD expenses first accrue [54].
While best practice mortality curves differ from the
rest of the sample, there are no differences in mortality
curves between services and manufactured goods.
Faster product development has become a goal of
many firms [1,24], with many firms reporting de-
creases in the NPD cycle time of specific products or
Figure 13. New Product Idea Mortality Curve projects. Research in expected development times in
1997;14:429 – 458
Figure 15. NPD Cycle Time by Project Type

Figure 14. Mortality Curves: The Best versus The Rest
services requires somewhere on the order of 31⁄2 years,

the German mechanical engineering industry has esti- new-to-the-firm projects require between 2 and 21⁄2
mated that managers could expect to eliminate some- years, major revisions to existing products take 1 to
where between 30% and 40% of current development 11⁄2 years, line extensions require about a year, and
times through a concerted reduction effort [48]. In the incremental improvements are generally completed in
PRTM survey, firms had actually reduced time to 2⁄3 of a year. It is unclear as to whether the “more
market by 9.5%, on average, between 1992 and 1994 innovative projects” are considered new-to-the-firm,
[52]. Across 21 projects reporting specific time reduc- new-to-the-world, or somewhere in between, in terms
tions in the business press between 1988 and 1992, the of newness. In terms of the 1995 sample, the reported
average time reduction was 48% [28]. However, this is overall cycle times are most comparable to new-to-
not a random sample of projects. Because these the-firm projects which are not new-to-the-world.
projects were specifically chosen for reporting due to Although there are no differences in cycle time in
management’s ability to reduce NPD cycle time, one the PDMA sample between the Best and the Rest, as
might expect that these projects could have larger Figure 16 shows, manufactured goods always take
reductions than the average project. Page [50] found longer to develop than services across different types
that firms’ “more innovative projects” took 35.5 of projects. Service development requires about half
months to complete in 1990 (3 years). On average, the the development time of goods, but only for major
1995 sample is developing their “more innovative revisions is this difference statistically significant.
projects” in 23.8 months (2 years), nearly a one-third Nearly 41% of 1990 respondents claimed that their
decrease in reported development time. This amount firms were developing products more quickly than 5
of cycle time reduction, while lower than the sample years ago [50]. A higher percentage of firms indicated
reported in [28], falls in the expected range from [48]. that they had taken time out of the NPD development
NPD cycle time does seem to be getting shorter for
“more innovative projects.”
Previous research has not found that NPD cycle
time and success are correlated [29]. Interestingly, the
Best firms in this sample do not differ from the Rest,
in terms of cycle time, supporting the earlier finding.
They are neither faster nor slower. Service firms are
four months faster than goods-developing firms, on
average, but this difference is not statistically signifi-
cant.
NPD cycle times correlate with the newness of the
project [28,29]. Two separate studies, in addition to
this research, have investigated recent average re-
ported cycle times by project type [37,46]. Figure 15 Figure 16. NPD Cycle Time: Manufactured Goods and Ser-
suggests that developing new-to-the-world goods and vices
1997;14:429 – 458
cycle over the last 5 years than was reported in 1990. NPD process use has moved from functional and
Over 60% of all firms report that they have shortened sequential approaches to multi-functional approaches,
cycle times for incremental improvements, major re- with formal stages and gates for moving from one
visions, and new-to-the-firm projects. Over 50% of the stage to the next. Even so, the percentage of firms
firms reported decreasing cycle time for new-to-the- following no process, or just using an informal process
world projects. While this focus on shortening cycle is still astoundingly high, at 38.5% of the total sample.
time is independent of whether the firm produces However, this figure is down from Page’s findings that
manufactured goods or services, statistically higher only 45.5% of firms had and followed a formal NPD
percentages of the Best firms report shortening cycle process. Each year, another 3.2% of firms implement
times for new-to-the-firm and new-to-the-world a formal NPD process. Using a formal process for
projects over the last five years. Over 3⁄4 of the Best controlling NPD is only slowly moving into firms,
firms report shortening cycle times for these newer even though previous research has demonstrated that
types of projects. The Best also report taking statisti- formal NPD processes improve the probability of
cally higher amounts of time out of the cycle for product development success [10].
new-to-the-world projects. They claim to have short- In the early 1980s, BAH recommended that NPD
ened this development cycle by nearly 29%, nearly programs be strategy-driven, and that each NPD
three times the 10% decrease claimed by the Rest of project start with a strategy-setting step. Although no
the firms (x2: F 5 8.0, p , .01, df 5 125). On average, one since 1982 has found strategy use to be as high as
development cycles have declined by between 15% BAH reported in their sample, strategy is becoming a
and 20% across the population, with the cycles of the more integral aspect of NPD. More NPD programs are
Best firms declining around 25%. Even though the specifically linked to business strategy, and more
Best firms claim to have eliminated more of the de- projects begin with a strategy-setting step. However,
velopment cycle, their development times are no lower because nearly 1⁄3 of the firms in this sample still
than those found in the rest of the population. Either develop products unlinked to strategy, this is still an
area with further improvement opportunities.
these two sets of findings conflict, or the Best firms
The number of organizations in a firm responsible
started the period with slightly longer development
for NPD are increasing, with a larger number of op-
cycles than the rest of the population.
tions available for where NPD reports within the firm.
Summary. Although overall NPD success rates
In general, the conclusion is that organizing for prod-
have not increased, firms appear to be more efficient
uct development is becoming more complex. The
in weeding out less probable projects earlier in the
BAH studies [4,5] have always reported that NPD
process and developing products in less time. The
reported to multiple parts of the organization. Page
portfolios of projects commercialized consist of a
[50] found that firms organized for NPD using 1.9
slightly higher proportion of moderately-innovative
structures. On average, this has moved to 2.0 struc-
projects, and fewer less-innovative projects.
tures in the last five years, although service producers
use fewer structures (1.7). No one organizational form
for NPD has been the trend toward which the bulk of
Discussion firms are moving. How “best” to organize NPD is not
clear.
NPD Trends Each of five possible organizational structures is
used by at least 30% of the sample. Venture groups are
Overall, the trends in NPD over the last 15–20 years the only structure investigated which are seldom used
have been more evolutionary than revolutionary. Per- by firms to manage more innovative NPD projects.
haps because NPD is so complex, involving so many The most-used reporting structure for more innovative
people across so many functions in the corporation, NPD projects is within the functional organization.
change comes, but comes slowly. A firm with static The extensive use of this structure for these projects is
NPD processes, policies and methods will find them- at odds with the capabilities of the structure, which has
selves falling behind. However, not even the Best been suggested as being most appropriate only for
firms have radically changed the face of product de- projects where little or no innovation is required [54].
velopment, or its outcomes. This suggests that functional silos are still strong in US
1997;14:429 – 458
firms. While the largest number of firms use the func- Differences Between the Best and the Rest
tional structure, the most frequently-preferred of the
structures is to have NPD report into the SBU. This Many different means could be used to differentiate
reporting structure is most appropriate for evolution- higher versus lower NPD performance. Some investi-
ary product improvement, not for radical innovation. gations have split their sample at an overall perfor-
However, perhaps this structure matches organiza- mance mean [37,38,47]. Others have compared the top
tional needs the majority of the time, since 71.8% of third to the bottom third, eliminating the middle third
the products commercialized by these firms in the last completely from the analysis [46]. Still others have
five years consisted of changes and improvements to used regression to determine performance differentia-
current products (48.7% of the total) or additions to tors [51]. While powerful, regressions only allow the
existing products lines (23.1%). analysis of continuous variables.
Project managers were recommended in the mid- The technique used to sort the Best NPD firms from
the Rest in this research is rather stringent and allows
1980s as the most appropriate leadership structure for
the analysis of discontinuous variables. It requires that
NPD [9]. The vast majority of firms have taken this
the Best firms be better than other firms along each of
advice and implemented their use extensively. New
three separate dimensions of NPD performance, not
leadership modes are moving only slowly into exper-
just one. This dividing process created groups of un-
imental use. Empowerment may have been espoused
equal sizes. The Best is only 22% of the total popu-
by many firms over the last few years, but its appli-
lation. Many of the firms included in the Rest perform
cation is not particularly evident yet in NPD. Team
well in one or two of the dividing criteria. A number
leaders are not selected by the team or by peers. They
of them are quite good at NPD. However, each of the
come from management, either upper management, or
Rest falls below the mean on at least one performance
functional management.
criteria, or is not in the top third of their industry for
Although multi-functional NPD teams are becom-
NPD performance.
ing endemic, firms have not grappled adequately with Many, but not all, of the practices the Best use differ
team-based rewards. Multi-functional teams have now statistically from the practices of the Rest. It seems
been instituted broadly for developing new products. that the Best do not succeed by using just one NPD
How best to reward joint efforts is a primarily unan- practice more extensively, or better, but by using a
swered question. Current practice has trended to “off number of them more effectively simultaneously. This
W-2” rewards. More sophisticated rewards are not seems key to differentiating between Best practice
used to any material extent. There are no major trends firms and the Rest. Further research on these data will
moving toward more sophisticated reward structures. try to identify the important simultaneous sets of prac-
As a result of these evolutionary process and orga- tices which lead to increased success.
nizational changes, NPD is slightly more efficient and A consistent finding in other studies is that process
somewhat faster, even though overall success rates are differentiates across NPD success [4,5,46]. The Best
unchanged, or even slightly lower than 13 years ago. are more likely to have NPD processes and strategies,
Product portfolios consist of a higher proportion of start the NPD process with a strategy, and include any
moderately-innovative projects, and fewer less-inno- particular step in their NPD process. As previously
vative projects, but the same proportion of highly- found [5], they are more likely to have used their
innovative projects. Perhaps increased success will process longer than lower performing firms. Their
come only by focusing on projects which are newer NPD process also has more steps than the process used
and more innovative? by the Rest. However, process use can be thought of as
In summary, the trends in NPD are evolutionary. a necessary, but not sufficient, condition to produce
Firms cannot allow their NPD practices to stagnate high NPD performance. A significant number of firms
because their competitors do not. If they do not put still do not consistently use a formal process, even
into place a mentality of and process for continual though they have been demonstrated to lead to higher
change, they could, slowly but unceasingly, be left NPD success [10]. Interestingly, the difference in use
behind competitively. Additional understanding of the extent is higher across manufactured good and service
impact of organizational structures, project leadership differentiations than across higher versus lower per-
and rewards on NPD might help firms manage these formance differentiations.
areas better. Far fewer organizational differences are found be-
1997;14:429 – 458
tween Best practice firms and the Rest. The way in performance not in just one aspect of NPD, but on
which NPD reports into the management structure of simultaneously improving several managerial aspects
the firm is independent of whether the firm is a best of NPD. No one managerial dimension is enough to
practice firm or not. While best practice firms are produce success in NPD.
more likely to have NPD report into the SBU and less
likely to have them report to a functional group, these
differences are not statistically significant. There does Differences Between Service and Goods Producers
not seem to be one best way to organize NPD in the
firm. Leadership and leader selection also do not differ In some very important ways, NPD is the same inde-
between Best practice and the Rest. pendent of whether services or manufactured goods
Only two organizational differences were found be- are being developed. Strategies always matter, both
tween Best practice firms and the Rest. The Best firms overall for the program and for each project as it is
use more rewards which are unrelated to compensation initiated. Leaders are still generally project managers
for teams (called “off W-2” rewards in the US), espe- appointed by management or the functional area.
cially providing a much higher incidence of project Multi-functional teams are still used, and important in
completion dinners or lunches. Best practice firms also differentiating between NPD success. Finally, deter-
use multi-functional teams more extensively for less- mining appropriate rewards is no less difficult for
innovative projects. services than for goods.
Best practice firms measure more and expect more However, services differ from manufactured goods
out of their NPD programs than their less successful in several unique ways [21,55,56]. Services cannot
counterparts. This may be one of the reasons they inventory their output. Service is a highly perishable
good. If it is not used as available in time, it disap-
achieve higher performance. Interestingly, the Best do
pears. In services, manufacturing takes place in real
not achieve higher performance in every dimension of
time (as the service is consumed) and, most frequently,
NPD outcomes. Best practice firms do not develop
the service user is a participant in the manufacturing
new products faster than other firms, supporting ear-
process. Finally, some aspects of services are intangi-
lier evidence that success and NPD speed were uncor-
ble, where the benefits to the consumer are not always
related [29]. Their focus appears to be on achieving
obvious. These differences affect three aspects of NPD
success over achieving speed. You do not necessarily
for services as compared to NPD for physical goods.
have to be fastest to be highly successful in NPD. As NPD processes, organizational structures, and cycle
others have suggested, there are hidden costs associ- time differ across developing physical goods and ser-
ated with highly accelerated NPD, which can have vices.
serious deleterious effects [20]. On the other hand, Research on differentiators between success and
because these firms have claimed to eliminate a higher failures in service NPD has concluded that having a
percentage of their NPD cycle over the last five years, market-driven new product process was the dominant
Best practice firms feel they cannot be slow and still ingredient for achieving new service success [14]. It
successful. Apparently, firms walk a knife edge be- has also demonstrated that service processes which
tween taking the time to do NPD right and being too more completely defined the strategy or charter [18]
slow because they are trying to take the time to do for projects and which more completely tested services
things right [20]. prior to commercialization were correlated with higher
Kuczmarski found that more successful firms at success [21]. However, this research also found that
NPD in their survey had product development portfo- few companies used a formal new service develop-
lios which focused more on more innovative projects. ment process. This result is replicated here. A statis-
This research does not support that finding. Best prac- tically lower percentage of service producers use pro-
tice firms are not likely to have any higher proportions cesses for NPD. The majority still either have no
of new-to-the-world or new-to-the-firm projects than process at all, or use an informal process, reproducing
the Rest of the firms. Best practice firms have a higher earlier results [21].
proportion of moderately-innovative projects, and In general, service processes are simpler than those
fewer less-innovative project, but no difference in the used to develop manufactured goods, even for the Best
highly-innovative projects. service firms. When a process is used in new service
The Best NPD firms appear to focus on improving development, it consists of fewer steps than those for
1997;14:429 – 458
manufactured goods. “Manufacturing development” is Limitations and Future Research

almost universally not included in the service NPD
process, probably due to the inherent nature of service Several limitations to this research should be kept in
delivery. However, service processes are nearly two mind when applying the results, either to what a firm
steps less than the average goods-developing process. might want to change to improve NPD performance,
A step in addition to manufacturing development is or to future research in NPD best practices. First, the
not used. Which specific step is eliminated depends response rate was very low, even for survey research.
upon the firm, however, the relative focus in the ser- While low response rates were expected with the non-
vice process seems to be more on steps in the front end PDMA sample, we hoped that a much higher percent-
of the process than the later stages. Both service and age of PDMA members would respond. Page [50,51]
manufactured goods processes are most likely to in- obtained a 27.6% response rate from PDMA member
clude a development step. However, concept genera- firms in the 1990 survey, which, however, was shorter
tion is the next most frequently included step for and traditional in style.
service development processes, whereas it is the next This low rate probably is attributable to several
to least frequently used step in manufactured goods specific actions which were purposefully taken in the
processes. The relative rank orders of the upfront research. The survey was long, and parts were com-
plex, requiring more effort and thought than just cir-
(pre-development) stages in service processes are
cling pre-typed numbers or responses. In addition,
lower than the rank orders for the same stages in
some types of questions were non-traditional in style.
manufactured goods processes.
Although the managers in the pre-test did not find the
Organizationally, NPD in service firms is less likely
survey difficult to do, it required more effort than the
to reside at the SBU level. Alternatively, they are more
typical survey a manager receives. Respondents did
likely to report to the marketing function. It is unclear
not receive any incentive, other than the opportunity to
whether these reporting structures result because ser-
receive a working paper upon completion, for filling
vice firms are less likely to be organized into SBU
out the survey. Respondent addresses or business
structures, or whether NPD reporting is different and
cards were received from well over half of the respon-
the overall proportion of firms organized into SBU’s is dents, suggesting that there was a high level of interest
similar to what is found in goods. in this research by those who responded. Finally, re-
The final area of difference is in NPD cycle time. In spondents paid the postage to return the surveys.
general, services take about half the time of manufac- One lesson from this survey seems to be that long
tured goods to develop, for any given level of product and complex surveys will lower response rates, even if
innovation. In other aspects of NPD outcomes, there the sample is interested in the results. In the future,
are no differences between service and manufactured then, NPD research may have to be broken into
good development, but cycle time for services is smaller increments. While this survey tried to under-
shorter. stand everything about NPD practices in one large
NPD is different in the service world. Unfortu- instrument, it may be more appropriate to split the
nately, most NPD research has been focused more on research into different parts. Each year a smaller sur-
understanding the needs of and establishing methods vey, focused on just one or a few aspects of NPD
for manufactured goods producers [55]. Most of the practices, could be mailed out. Across the series,
NPD processes and methods developed have been which is sent out over a five-year period, the entirety
targeted to goods-producing firms [11,16]. Perhaps of NPD aspects could be covered.
because NPD processes have been targeted more at In applying these results, then, managers must real-
manufactured goods producers than at service firms, ize that the respondent firms are most likely different
fewer service firms use them. Service firms, even from non-respondent firms, but in an unknown way.
highly successful ones at NPD, are able to use less Demographic data of the sort gathered and reported on
complex processes with fewer steps and develop ser- here are not available for any of the three mailing lists
vices faster. More research on the NPD needs of used. It is likely, however, that the respondents to this
service firms should be done, and practices specific to survey are more interested in NPD than the non-
meeting those needs should be developed, starting respondents. The firms who are least interested in
with better delineations of best practice processes and NPD are those who are least likely to have responded.
organizational structures for service development. Thus, even the practices of ”the Rest” in this sample
1997;14:429 – 458
may be more sophisticated than the average firm in the References

overall population. 1. Little, Arthur D. The Arthur D. Little Survey on the Product Innovation
A second limitation is that the sample is a US Process, Arthur D. Little: Cambridge, MA (December 1991).
2. Bart, Christopher K. Controlling New Products in Large Diversified
sample. PDMA firms are predominantly US-based. Firms: A Presidential Perspective. Journal of Product Innovation
Because this research was performed to serve the Management 8(1):4–17 (March 1991).
needs of our members, and due to funding limitations, 3. Bayus, Barry. Are Product Life Cycles Really Getting Shorter? Jour-
nal of Product Innovation Management 11(4):300–308 (September,
mailings were made only to US firms. Because man- 1994).
agement practices, cultures and norms differ around 4. Booz, Allen and Hamilton. Management of New Products. New York:
the world, these findings are likely to be less applica- Booz, Allen and Hamilton, 1968.
ble to firms managing NPD outside the US [1]. Dif- 5. Booz, Allen and Hamilton. New Products Management for the 1980s.
New York: Booz, Allen and Hamilton, 1982.
ferences in NPD Best practices around the world, 6. Bowen, H. Kent, Clark, Kim B., Holloway, Charles A. and Wheel-
although expensive to determine, would be a fascinat- wright, Steven C. The Perpetual Enterprise Machine. New York:
Oxford University Press, 1994.
ing piece of future research.
7. Burger, P.C. A Report on the Development of a Research Agenda for
This study found fewer differences between NPD the Product Development & Management Association. Journal of
practices for manufactured goods and services than it Product Innovation Management 6(2):51–60 (March 1989).
did across the Best and the Rest. One possible reason 8. Carlson, T. The Race is On. Brandweek 24–27 (May 9, 1994).
9. Clark, Kim B. and Fujimoto, Takahiro. Product Development Perfor-
is that research in improving NPD has predominantly mance, Boston, MA: Harvard Business School Press, 1991.
focused on goods-producing firms rather than on the 10. Cooper, Robert G. New Products: The Key Factors in Success. Chi-
special needs of service firms, which are different cago, IL: American Marketing Association, 1990.
from those of goods manufacturers [21]. For example, 11. Cooper, Robert G. Stage-Gate System: A New Tool for Managing
New Products. Business Horizons 44–54 (May–June 1990).
test markets in service firms are used primarily to
12. Cooper, Robert G. Third Generation New Product Processes. Journal
ensure the proper functioning of the service rather than of Product Innovation Management 11(1):3–14 (January, 1994).
to provide a base for a national sales projection [22]. 13. Cooper, Robert G. and de Brentani, Ulricke. New Industrial Financial
Services: What Distinguishes the Winners. Journal of Product Inno-
Several factors have been found which statistically vation Management 8(2):75–90 (June 1991).
significantly impact new service success, in addition 14. Cooper, Robert G., Easingwood, Christopher J., Edgett, Scott, Klein-
to those factors which impact manufactured goods schmidt, Elko J. and Storey, Chris. What Distinguishes the Top Per-
forming New Products in Financial Services. Journal of Product
success [13]. Thus, the questions in this survey may Innovation Management 11(4):281–299 (September 1994).
not have captured the unique characteristics and as- 15. Cooper, Robert G. and Kleinschmidt, Elko. An Investigation into the
pects of NPD for services which are truly different New Product Process: Steps, Deficiencies, and Impact, Journal of
Product Innovation Management 3(2):71–85 (June, 1986).
from those for manufactured goods. A separate study
16. Cooper, Robert G. and Kleinschmidt, Elko. New Product Processes at
of Best practices for services is another area of future Leading Industrial Firms. Industrial Marketing Management 20:137–
research which could hold great potential. 147 (1991).
Additional papers from these data have been pub- 17. Crawford, C. Merle. New Product Failure Rates—Facts and Fallacies.
Research Management 9–13 (September, 1979).
lished and are in progress. A larger analysis of data 18. Crawford, C. Merle. Defining the Charter for Product Innovation.
from the PDMA sample was published as Chapter 33 Sloan Management Review 3–12 (Fall, 1980).
of the PDMA Handbook of New Product Development 19. Crawford, C. Merle. New Products Management, Homewood, IL:
Irwin (1991).
[36]. Papers on product development champions [41]
20. Crawford, C. Merle. The Hidden Costs of Accelerated Product Devel-
and organizing for NPD [42] are underway. Additional opment. Journal of Product Innovation Management 9(3):188–199
papers looking at NPD tools and determinants of NPD (September, 1992).
success across processes, organizations, and tools are 21. de Brentani, Ulrike. Success and Failure in New Industrial Services.
Journal of Product Innovation Management 6(4):239–258 (December,
planned. These will be available from the PDMA 1989).
office as they are produced. 22. Easingwood, Christopher J. New Product Development for Service
Companies. Journal of Product Innovation Management 3(4):264–275
(December, 1986).
23. Edgett, Scott, Shipley, D. and Forbes, G. Japanese and British Com-
panies Compared: Contributing Factors to Success and Failure in
NPD. Journal of Product Innovation Management 9:3–11 (March
1992).
Funding for this research was provided by the Product Develop- 24. Fact Finders, Inc. Southwestern Bell Telephone Product Development
ment & Management Association. Additional support was pro- Benchmarking Study, Fact Finders, Inc. (September, 1996).
vided by the University of Chicago’s Graduate School of Business. 25. Feldman, Laurence. A Profile of the New Product Professional. Jour-
nal of Product Innovation Management 8:252–266 (1991).
1997;14:429 – 458
26. Feldman, Laurence. The Role of Salary and Incentives in the New nizational Tools on New Product Development Efficiency and Effec-
Product Function. Journal of Product Innovation Management 13(3): tiveness. University of Chicago working paper (January 1997).
216–228 (May, 1996). 43. McGrath, Michael E., Anthony, Michael T. and Shapiro, Amram R.
27. Graziano, Karen. Process Ownership. In: The PDMA Handbook of Product Development: Success Through Product and Cycle-Time Ex-
New Product Development, Milton D. Rosenau, Abbie Griffin, George cellence, Stoneham, MA: Butterworth-Heinemann, 1992.
Castellion and Ned Anschuetz, (Eds.), John Wiley & Sons (1996). 44. McGrath, Michael E. and Romeri, Michael N. From Experience: The
28. Griffin, Abbie. The Effect of Project and Process Characteristics on R&D Effectiveness Index: A Metric for Product Development Perfor-
Product Development Cycle Time. Journal of Marketing Research, mance. Journal of Product Innovation Management 11(3):213–220
XXXIV, 24–35 (February, 1997a). (June 1994).
29. Griffin, Abbie. Measuring Product Development to Improve the Qual- 45. Mello, Sheila and Vermette, David. Developing Breakthrough Prod-
ity of the Process. In: The Practice of Quality Management, Phillip J. ucts: How the Best in Class Do It. Management Roundtable Confer-
Lederer and Uday S. Karmarkar, (Eds.), Boston, MA: Kluwer Aca- ence on Product Development Best Practices for Defining Customer
demic Publishers, 117–145 (1997b). Needs. New Orleans, LA. January, 1995.
30. Griffin, Abbie and Hauser, John R. Voice of the Customer. Marketing 46. Mercer Management Consulting, Inc. High Performance New Product
Science 12(1):1–27 (1993). Development: Practices That Set Leaders Apart, Mercer Management
31. Griffin, Abbie and Page, Albert L. An Interim Report on Measuring Consulting, Inc.: Boston, MA. November, 1994.
Product Development Success and Failure. Journal of Product Inno- 47. Mitchell Madison Group. New Product Development: Organization
vation Management 10(3):291–308 (September, 1993). and Process in Service Companies, Mitchell Madison Group: New
32. Griffin, Abbie and Page, Albert L. The PDMA Success Measurement York, NY (December 14, 1995).
Project: Recommended Measures for Product Development Success 48. Murmann, Philipp A. Expected Development Time Reduction in the
and Failure. Journal of Product Innovation Management 13(4):478– German Mechanical Engineering Industry. Journal of Product Inno-
496 (November, 1996). vation Management 11(3):236–252 (June, 1994).
33. Group EFO. 1995 Innovation Survey: Report on New Products, Group 49. Norton, John and Bass, Frank M. Evolution of Technological Gener-
EFO Limited: Weston, CT (1995). ations: The Law of Capture. Sloan Management Review, 66–77 (Win-
34. Hayes, Robert H., Wheelwright, Steven C. and Clark, Kim B. Dynamic ter, 1992).
Manufacturing. New York, NY: the Free Press, 1988. 50. Page, Albert L. Assessing New Product Development Practices and
35. Henke, John W., Krachenberg, A. Richard and Lyons, Thomas F. Performance: Establishing Crucial Norms. Journal of Product Inno-
PERSPECTIVE: Cross-Functional Teams: Good Concept, Poor Im- vation Management 10(4):273–290 (September 1993).
plementation. Journal of Product Innovation Management 10:216–229 51. Page, Albert L. Results from PDMA’s Best Practices Study: The Best
(1993). Practices of High Impact New Product Programs. The EEI/PDMA
36. Hustad, Thomas P. Reviewing Current Practices in Innovation Man- Conference on New Product Innovation (June 1994).
agement and a Summary of Selected Best Practices. In: The PDMA 52. Pittiglio, Rabin, Todd & McGrath. Product Development Leadership
Handbook of New Product Development, Milton D. Rosenau, Abbie for Technology-Based Companies: Measurement and Management—A
Griffin, George Castellion and Ned Anschuetz, (eds.), John Wiley & Prelude to Action, Pittiglio, Rabin, Todd & McGrath: Weston, MA
Sons (1996). (1995).
37. Kuczmarski & Associates, Inc. Winning New Product and Service 53. Sellers, Patricia. Pepsi Keeps Going After No. 1. Fortune, 62–70
Practices for the 1990’s, Kuczmarski & Associates: Chicago, IL (March 11, 1991).
(1994).
54. Souder, William. Managing New Product Innovations, Lexington,
38. Kuczmarski & Associates, Inc. Winning New Product and Service
MA: Lexington Books (1987).
Practices for the 1990’s: Presentation Handouts, Chicago PDMA
Chapter meeting, Kuczmarski & Associates: Chicago, IL (June 29, 55. Terrill, Craig A. The Ten Commandments of New Service Develop-
1994). ment. Management Review (February 1992).
39. Larson, Erik W. and Gobeli, David H. Organizing for Product Devel- 56. Terrill, Craig A. and Middlebrooks, Arthur G. Service Development.
opment Projects. Journal of Product Innovation Management 5(3): In: The PDMA Handbook of New Product Development, Milton D.
180–190 (September, 1988). Rosenau, Abbie Griffin, George Castellion and Ned Anschuetz (Eds.),
John Wiley & Sons (1996).
40. Lawrence, R. Inside New-Product Statistics: More, or Less. . .New or
Not? Journal of Advertising Research 33, RC3–RC6 (March–April, 57. Urban, Glen L. and von Hippel, Eric. Leader User Analysis for the
1993). Development of New Industrial Products. Management Science, 34:
41. Markham, Stephen and Griffin, Abbie. The Breakfast of Champions: 569–582 (1988).
Associations Between Champions and Product Development Environ- 58. von Braun, Christoph-Friedrich. The Acceleration Trap. Sloan Man-
ments, Practices and Performance. North Carolina State University agement Review, 49–58 (Fall, 1990).
working paper (June, 1996). 59. Womak, James P., Jones, Daniel T. and Roos, Daniel. The Machine
42. McDonough, Edward F. III and Griffin, Abbie. The Impact of Orga- that Changed the World. New York, NY: Rawson Associates, 1990.
1997;14:429 – 458
Appendix
Strategy, Process, and Organization Questions
I. The Process by Which Products Are Developed Internally
Please answer questions as they apply to the more innovative types of new products or services your
organization develops. These include such things as new applications, major product changes, new products and
product lines, and new-to-the-world products. Please do not include product repositionings, brand or product line
extensions and incremental product improvements.
1. Does your organization have a specific strategy for its new product activities which directs and
integrates the entire new product program?
Yes ❑ No ❑
2. a. Does your organization tend to follow a well-defined, structured process for the development of most
or all of your more innovative new products?
Yes ❑ No ❑ (If no, skip to question 4)
b. Approximately how many years has your organization been following a formal product development
process?
wyears
3. Please check the box which most closely describes your firm’s product development process:
❑ No standard approach to new product development.

❑ While no formally documented process is followed, we have a clearly understood path of the tasks to be
completed in product development.
❑ We have a formally-documented process where one function completes a set of tasks, then passes the
results on to the next function which completes another set of tasks.
❑ We have a formally-documented process where a cross-functional team completes a set of tasks,
management reviews the result and gives the go-ahead for the team to complete the next set of
cross-functional tasks.
❑ We have a formally-documented process where a facilitating “process owner” helps cross-functional
teams move through stages and management reviews.
❑ We have a formally-documented process where cross-functional team uses a staged process with
overlapping, fluid stages and “fuzzy” or conditional stage decisions.
❑ Other (Please describe):
4. Many new product ideas never reach commercialization. What percentage of your organization’s total
new product ideas/concepts (100%) reach each stage of the development process below, whether you
use this process formally or not?
Idea/Concept Generation Idea Screening Business Analysis Development Test and Validation Commercialization
100% w% w% w% w% w%
5. The development of a new product is often described as a series of interdependent and possibly
overlapping stages. Below are descriptions of several development activities.
a. Please place an “X” in the first column if your organization’s formal product development process
includes this activity for the more innovative projects.
b. Please estimate the % of the more innovative projects commercialized by your organization in which
each of the listed tasks was completed prior to commercialization.
c. Please indicate the typical length of time (in weeks) spent on each activity.
1997;14:429 – 458
Process Includes % of projects Completing # of Weeks Spent

Product Line Planning: Analyze the firm’s current
portfolio vis-a-vis the competitive arena. ❑ w% w weeks
Project Strategy Development: Delineate the target
market, determine market need, attractiveness. ❑ w% w weeks
Idea/Concept Generation: Identify opportunities and
initial generation of possible solutions. ❑ w% w weeks
Idea Screening: Sort and rank solutions, eliminate
unsuitable and unattractive options. ❑ w% w weeks
Business Analysis: Evaluate the concept financially,
write business case, prepare protocol/development
contract. ❑ w% w weeks
Development: Convert concept into a working
product. ❑ w% w weeks
Test and Validation: Product use, field, market and
regulatory testing with customers. ❑ w% w weeks
Manufacturing Development: Developing and
piloting the manufacturing processes. ❑ w% w weeks
Commercialization: Launching the new product or
service into full scale production and sales. ❑ w% w weeks
Other: Any other activity included in your firm’s
formal product development process. (Please
specify) ❑ w% w weeks
II. Organizing for Product Development

1. Which of the following best describes the way the new product effort is structured in your organiza-
tion? (Please ✓ all structures used for your firm’s more innovative projects.)
❑ New product department with permanent staff members.

❑ Distinct division or venture group.
❑ A new product committee oversees all development efforts.
❑ Each business unit’s general manager directs their own NPD efforts.
❑ A single function is responsible for NPD: (Which function?)
❑ R&D ❑ engineering
❑ planning ❑ marketing
❑ A product development process owner helps deploy our process across the firm.
❑ Other (Please specify.)
2. Which of the following best describes how the more innovative projects are led at your firm and
how those leaders are obtained? (Please check all that apply.)
Who leads How the leader is obtained
❑ Project champions ❑ Management appoints the leader
❑ Process owners ❑ A person volunteers
❑ Project managers ❑ Team members choose the leader
❑ No one ❑ Peers choose the leader
❑ Other w ❑ Other
1997;14:429 – 458
3. What % of each of the following types of projects use multifunctional teams?

w% New-to-the-world products w% New-to-your-organization products
w% Major revision of current product w% Incremental improvement
w% Product repositioning w% Cost Reduction
4. What product development-based incentives and rewards are provided for team leaders and team
members? Please check all that are used, and estimate the degree to which each type of reward is
used for team leaders and members. (Please check and fill out all that apply.)
Team Leader Team Member
1⁄4 1 ⁄2 3⁄4 Virtually 1⁄4 1⁄2 3⁄4 Virtually
Used Never the Time the Time the Time Always Never the Time the Time the Time Always
❑ Project-based profit-sharing 1 2 3 4 5 1 2 3 4 5
❑ Project-based stock or stock 1 2 3 4 5 1 2 3 4 5
options
❑ Compensation time 1 2 3 4 5 1 2 3 4 5
❑ Recognition in organization 1 2 3 4 5 1 2 3 4 5
newsletters
❑ Recognition at award dinners 1 2 3 4 5 1 2 3 4 5
❑ Plaques, pins, project 1 2 3 4 5 1 2 3 4 5
photographs
❑ Project completion celebration 1 2 3 4 5 1 2 3 4 5
lunches, dinners
❑ Non-financial rewards chosen by 1 2 3 4 5 1 2 3 4 5
the team (eg, trips, family
dinners)
❑ Other financial rewards (specify) 1 2 3 4 5 1 2 3 4 5
w
❑ Other non-financial rewards 1 2 3 4 5 1 2 3 4 5
(specify) w

PDMA Research On New Product Development Practices: Updating Trends and Benchmarking Best Practices

Uploaded by

Copyright:

Available Formats

PDMA Research On New Product Development Practices: Updating Trends and Benchmarking Best Practices

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

PDMA Research On New Product Development Practices: Updating Trends and Benchmarking Best Practices

Uploaded by

Copyright:

Available Formats

jjjj

PDMA Research on New Product Development Practices:

Address correspondence to Prof. Abbie Griffin, University of Illinois,

J PROD INNOV MANAG 1997;14:429 – 458

years ago and relative to what other investigations

Table 1. Summary of Prior Best Practice Studies

1990 Page, PDMA 189 survey responses ● Measure NPD performance

found in the numbers of ideas required to produce a Variable Operationalizations

Table 3. Sample and Response Rates

PDMA Members American Marketing Association CorpTech List Total

Table 4. Demographics by Source of the Sample

PDMA Members AMA Sample CorpTech Total Sample

Technology Base1 High-Tech 43 27.7% 18 22.2% 67 49.3% 128 34.4%

Table 5. Average Success by Source of the Sample

PMDA Members AMA CorpTech Total Sample

Table 6. Descriptive Statistics of the Sample

Technology Base High-Tech 24 33.8% 103 35.3% 127 35.0%

Table 7. Distribution of “The Best” by Market and Product Type

Consumer Market Mixed Business-to-Business Full Sample

Services 2 12.5% 3 13.6% 5 15.2% 10 14.1%

The data are analyzed using several methods.

Using a formal NPD process and not skipping steps in

Figure 1. Product Development Processes: Services versus

Figure 4. NPD Process Components Figure 5. NPD Process Components

marketing function, and least likely to report to an

However, in some firms, these owners may also be

Figure 10. Team Leader and

while only 28% of the lower-performing firms track

programs over the last 15 years. But, have organiza-

Table 8. Average Success by Demographic Category

Success Rate1 % Profit Success2 NP Sales %3 NP Profit %4

Figure 15. NPD Cycle Time by Project Type

services requires somewhere on the order of 31⁄2 years,

manufactured goods. “Manufacturing development” is Limitations and Future Research

may be more sophisticated than the average firm in the References

❑ No standard approach to new product development.

Process Includes % of projects Completing # of Weeks Spent

II. Organizing for Product Development

❑ New product department with permanent staff members.

3. What % of each of the following types of projects use multifunctional teams?

You might also like