Product Development: Past Research, Present Findings, and Future Directions

Author(s): Shona L. Brown and Kathleen M. Eisenhardt

Source: The Academy of Management Review, Vol. 20, No. 2 (Apr., 1995), pp. 343-378
Published by: Academy of Management
Stable URL: http://www.jstor.org/stable/258850
Accessed: 07-03-2015 13:11 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact [email protected].

Academy of Management is collaborating with JSTOR to digitize, preserve and extend access to The Academy of Management
Review.

http://www.jstor.org

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

0 Academy of Management Review

1995, Vol. 20, No. 2, 343-378.

PRODUCTDEVELOPMENT:
PAST RESEARCH,
PRESENTFINDINGS, AND FUTUREDIRECTIONS
SHONAL. BROWN
M. EISENHARDT
KATHLEEN
Stanford University
The literature on product development continues to grow. This research is varied and vibrant, yet large and fragmented. In this article
we first organize the burgeoning product-development literature into
three streams of research: product development as rational plan, communication web, and disciplined problem solving. Second, we synthesize research findings into a model of factors affecting the success
of product development. This model highlights the distinction between process performance and product effectiveness and the importance of agents, including team members, project leaders, senior
management, customers, and suppliers, whose behavior affects these
outcomes. Third, we indicate potential paths for future research
based on the concepts and links that are missing or not well defined
in the model.

Innovation research splits into two broad areas of inquiry (Adler,

1989). The first, an economics-oriented tradition, examines differences in
the patterns of innovation across countries and industrial sectors, the
evolution of particular technologies over time, and intrasector differences
in the propensity of firms to innovate (e.g., David, 1985; Dosi, 1988; Nelson
& Winter, 1977; Urabe, Child, & Kagono, 1988). The second, an organizations-oriented tradition, focuses at a microlevel regarding how specific
new products are developed (e.g., Ancona & Caldwell, 1992b; Clark &
Fujimoto, 1991; Zirger & Maidique, 1990). Here, the interest is in the structures and processes by which individuals create products. In this article,
we focus on this latter area of the broader innovation literature.
Product development is critical because new products are becoming
the nexus of competition for many firms (e.g., Clark & Fujimoto, 1991). In
industries ranging from software to cars, firms whose employees quickly
develop exciting products that people are anxious to buy are likely to win.
The writing of this article was generously supported by funding from the Alfred P. Sloan
Foundation to the Stanford University Computer Industry Project. Kathleen M. Eisenhardt
also was generously supported as the Finmeccanica Faculty Scholar. The authors thank
Deborah Dougherty, D. Charles Galunic, Rebecca Henderson, Neil Kane, Laura Kopczak,
Dorothy Leonard-Barton, and Mark Zbaracki for their helpful comments. In addition, they
thank Susan E. Jackson and the anonymous reviewers at AMR for their insightful suggestions.
343

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

344

Academy of Management Review

April

In contrast, firms introducing "off-the-mark" products are likely to lose.

Product development is thus a potential source of competitive advantage
for many firms (Brown & Eisenhardt, 1995). Product development is also
important because, probably more than acquisition and merger, it is a
critical means by which members of organizations diversify, adapt, and
even reinvent their firms to match evolving market and technical conditions (e.g., Schoonhoven, Eisenhardt, & Lyman, 1990). Thus, product development is among the essential processes for success, survival, and
renewal of organizations, particularly for firms in either fast-paced or
competitive markets.
During the past 10 years, the pace of product-development research
has quickened as numerous academic scholars have probed the secrets of
product-development prowess (e.g., Ancona & Caldwell, 1990; Clark &
Fujimoto, 1991; Dougherty, 1990; Zirger & Maidique, 1990). Interest in product development plus preachings about the importance of proficient product development have reverberated in the popular press (Business Week,
1992; Dumaine, 1991; Schendler, 1992). The underlying rationale appears
to be that, although technical and market changes can never be fully
controlled, proactive product development can influence the competitive
success, adaptation, and renewal of organizations. However, because
this large and fragmented literature has not been tied together to create
cogent understanding, it is difficult to grasp what is actually known.
Thus, the purpose of this article is to improve the understanding of product development.
We begin by organizing the empirical literature on product development into three streams: product development as rational plan, communication web, and disciplined problem solving. By the empirical literature
on product development, we mean articles published in major Englishlanguage organizations-oriented North American and European journals
where this work is likely to appear' and a few, unpublished studies. We
specifically focus on normative empirical studies of product development
in which the development project is the unit of analysis. Even with these
constraints, it is still impossible to cover all studies in one review article.
As we built the literature review, we selected studies based on the rigor
of their empirical methods and the degree to which they were cited by
others. We allowed the network of studies to grow forward and backward
in time with no constraints. As a cross-check against the burgeoning
network of citations, we iteratively returned to the journals. Through this
process, the three distinct networks of research that we describe and the
temporal boundaries between 1969 and the present emerged.

1 We included Administrative Science Quarterly, Academy of Management Review,

Organization Science, Organizational Studies, Journal of Management Studies, Harvard
Business Review, California Management Review, Research Policy, and Journal of Product
Innovation Management.

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

1995

Brown and Eisenhardt

345

Second, we synthesize these research findings into a model of factors

affecting the success of product development (see Figure 1). This model
integrates the common findings of the research streams and blends in
complementary ones. In addition, the model attempts to build a theoretical framework for the findings in these streams.
Third, we indicate potential paths for future research. These paths
are centered on the effects that the development process and product
concept have on product success, patterns of organizing productdevelopment work, strategic management, and customer/supplier involvement. Overall, the intent of the article is to contribute an understanding of the past literature, a model of current thinking, and a sense of
future directions.

LITERATURE
REVIEW
The product-development
literature is vast, ranging from broadbrush explorations to in-depth case studies and across many types of
products, firms, and industries. In this section, we create an organizing
template for this work. Although several templates are possible, we have
organized this one around three emergent research streams: rational
plan, communication web, and disciplined problem solving. The three
streams are summarized along key dimensions in Table 1.
We shaped our review around these streams because each involves
a pattern of cumulative citations evolving from one or two pioneering
studies. The rational stream builds on the Myers and Marquis (1969) and
SAPPHO studies (Rothwell, 1972; Rothwell et al., 1974);the communication
stream, on the early work of Allen at the Massachusetts Institute of Technology (MIT) (1971, 1977); and the problem-solving stream, on Imai and
colleagues' (1985) study of successful Japanese products. Each stream's
focus reflects an evolution based on the constructs highlighted in the
pioneering works. The result has been three relatively coherent and distinct bodies of research.
Further, although there are overlaps in focus across the streams (e.g.,
all streams investigate how different players, processes, and structures
affect performance), research within each stream centers on particular
aspects of product development. The rational plan research focuses on a
very broad range of determinants of financial performance of the product,
whereas the communication web work concerns the narrow effects of
communication on project performance. Disciplined problem solving centers on the effects of product-a
development team, its suppliers, and
leaders on the actual product-development process.
Moreover, the research within each stream is theoretically and methodologically similar. The rational plan perspective is primarily exploratory and atheoretical and, thus, helps to broadly define the relevant
factors for product-development
research. The communication
web

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

346

Academy of Management Review

April

4)~~~~~~~~~~~~~"
4)~~~~~~~~~~~~~~~~~~~i

4)~~~~~~~~~~~~~~~~~~~~nC
U~~~~~~~~~~~~~~~~~~~~t
0~~~~~~~~~~~~
(/2~~~~~~~~~~~~~~~~~~~~~~~(
O)

U)

)4

LA~~~~~~~~~~~~~~~~~~
4)~~~~~~~~~~~4

0~~~~~~~~~~~~~~~~~~~~

C14~~~~~

4)~~~~~~~C

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

347

Brown and Eisenhardt

1995

TABLE 1
Comparison of Three Research Streams
Concepts

Rational Plan

Communication
Web

Disciplined Problem
Solving

Key idea

Success via superior

product, attractive
market, rational
organization

Success via internal

and external
communication

Success via
problem solving
with discipline

Theory

Mostly atheoretical

Information and
resource
dependence

Information
including
problem solving

Methods

Bivariate analysis;
single informant;
many independent
variables

Deductive and
inductive;
multivariate;
multiple
informants

Progression from
inductive to
deductive; multiple
informants; single
industry, global
studies

Product

Product advantagecost, quality,

uniqueness, fit
with core
competence

Market

Size, growth,
competition

Product integrityproduct vision

that fits with
customers and
firm

Senior management

Support

Subtle control

Project team

X-functional, skilled

X-functional

Communication

High crossfunctional

Organization of work

Planning and
"effective"
execution

High internal, high

external-various
types and means

Overlapped phases,
testing, iterations,
and planning
Politician and small
group manager

Project leaders
Customers

Early involvement

Suppliers

Early involvement

Performance
(dependent variable)

Financial success
(profits, sales,
market share)

High internal

Heavyweight

leader

High involvement
Perceptual success
(team and
management
ratings)

Operational
success (speed,
productivity)

stream complements this atheoretical view by relying on informationprocessing and resource dependence theoretical perspectives in the context of traditional research studies. The disciplined problem-solving
stream takes the theoretical perspective of information processing one

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

348

Academy of Management Review

April

step further to problem-solving strategies, using a progression from inductive to deductive research and an emphasis on global industry studies.
Overall, these three streams seem, to us, to capture best the cumulative patterns of product-development research. In this section, we outline these streams, including their key concepts, critical findings, underlying theory, methods, strengths, and weaknesses. However, as noted,
although these streams are coherent bodies of work, they also complement and somewhat overlap one another. So, in the subsequent section,
we emphasize these overlaps and complementarities by blending them
into an integrative model of product development.
Product Development

as Rational Plan

This rational plan perspective emphasizes that successful product

development is the result of (a) careful planning of a superior product for
an attractive market and (b) the execution of that plan by a competent and
well-coordinated cross-functional team that operates with (c) the blessings of senior management. Simply put, a product that is well planned,
implemented, and appropriately supported will be a success.
The focus in this stream is on discovering which of many independent
variables
are correlated with the financial success of a productdevelopment project. The studies often are exploratory, and their perspective is broad. Typically, researchers gather questionnaires or possibly
interview responses from well-placed, single informants. Informants usually are asked to explain why a product succeeded or failed, using a wide
spectrum of internal and external factors. However, because results are
often empirically observed correlations with success, the theoretical understanding of relationships usually is quite limited, and nonsignificant
findings often are not reported. Selected studies in this stream are summarized in Table 2, and a model is presented in Figure 2.
The earliest work in this stream emphasized the importance of market
issues over purely technical ones for successful product development. For
example, Myers and Marquis (1969) studied the development of 567 successful products and processes in over 100 firms and 5 industries. Their
principal result was that market pull (i.e., identifying and understanding
users' needs) was substantially more important to the success of the products than technology push, and thus a cross-functional view was a key
component of product success.
Later studies added failures to the mix (e.g., Rothwell, 1972; Rubenstein, Chakrabarti, O'Keefe, Souder, & Young, 1976). The SAPPHO studies
(e.g., Rothwell, 1972; Rothwell et al., 1974) were conducted using 43 success and failure pairs among chemical and instruments firms within the
United Kingdom. The authors found that 41 factors, including understanding users' needs, attention to the market, efficient development, and senior leadership, were significantly related to successful product develop-

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

1995

349

Brown and Eisenhardt

0)~~~~~~~~~~~~~0
0

0
v

o
- ---

2o

P.

00

0
0

~ ~

4)~

00

00

U)

~~ ~ P.0~
~~~0

0)

0
U

00

40

04

~~~~~~~~~~~~~0
~ ~~
U

~~~02

0
14
)

t%

co

U~~~~0

1
0

4)

0~~~~~~~~~~~~~~~~~~~~~~~~

0
4)

ti

0 0

.0~00

>

4)5

)1)

41

-U

~~~~~~~~~0~

>~~~~~0

S
)s

'g
u~~,0

u a)
~
~ ~~~00~~~~~~~~~~~~00
0;
4)

00

-.~~~~mC:

4)

>0u

C,)

-0

0000

06
5-.

U)

0~~~~~

o* -

0P.,>

4)

U~

U~:3

?2

C)

~4

0~~~~~~~~~~~~~~~~~~~~~

U)

0
-

0
0

0
ull~
U)

).

o0

0)
4

uu

4)0

0-

L0

U)t00

90C
;dU)

'C

U)

0~~~~~~~~~~~~~~~~~~~~~~~~~~0

t~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
0a)

En

00

~~-e
(0~

C"04

U)~~~~~~~
~~~O

~~~~~

~ ~ ~~S 0~U

CO

~~~~~~~

CT)~~~~~~U
r-0i0
C")

U)
0

-0

0a)0)

U00

0~~~~~00
0

(2

a,O

-Q.

zN

CO

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

~~~~

350

Academy of Management Review

April

FIGURE 2
Rational Plan Model of Product Development
Team Composition
*Cross-functional1

Team Organization of Work

*Planning\
Team Process
*Cross-functional\
communication\\

\F\twithfrPerfopeance
Senior Management

Prft

*Support
share

Ioeme*Market
Product Effectiveness/
*Fit with market needs///
Unique benefits///
Quality///
*Hg growthit
Cost/
Clear concept/
*Fit with firm competencies/

//
//
//

Market/

*Large/
*Hfighgrowth/
*Low competition

/
/

Customer

ofinvolvement/

Supplier

oInvolvement

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

1995

Brown and Eisenhardt

351

ment. Although the large number of factors makes it difficult to hone

precise managerial und theoretical implications, this breadth of findings
yields a comprehensive view of the important issues within product development. The SAPPHO studies were then followed by similar studies in
other countries such as Finland (Kulvik, 1977), Hungary (Szakasits, 1974),
and West Germany (Gerstenfeld, 1976).
Subsequent research sharpened the emergent emphases on product
advantages, market attractiveness, and internal organization. Particularly important were two studies by Cooper (1979; Cooper & Kleinschmidt,
1987). The first, the NewProd study (Cooper, 1979), examined 102 successful and 93 failed products within 103 industrial firms in Canada. A subsequent study by Cooper and Kleinschmidt (1987) examined hypotheses
from the NewProd and other studies using 203 products in 125 manufacturing firms, including 123 successes and 80 failures. Data were gathered
from either the most knowledgeable manager or managers using a structured interview questionnaire. Success and failure were measured by 11
mostly financial measures, including profitability, payback, sales, and
market share.
The authors observed that the most important determinant of product
success was product advantage. The intrinsic value of the product, including unique benefits to customers, high quality, attractive cost, and
innovative features, was the critical success factor. Such products were
seen as superior to competing products and solved problems that customers faced.
Internal organization also was critical to product success. Particularly important was predevelopment planning. This included developing
a well-defined target market, product specifications, clear product concept, and extensive preliminary market and technical assessments. Other
internal organization factors also were important, including crossfunctional skills and their synergies with existing firm competencies. Top
management support also was important, but less so than these other
factors.
Finally, market conditions also affected product success. Cooper and
Kleinschmidt (1987) found that products that entered large and growing
markets were more likely to be successful. In addition, products introduced into markets with low overall intensity of competition were more
successful. However, they also noted that market characteristics were
less important to commercial success than were product and internal
organization factors such as product advantage, clear product concept,
and predevelopment planning.
More recently, Cooper and Kleinschmidt (1993) conducted another
NewProd study of product development in the North American and European chemical industries. The authors replicated some of their earlier
findings. Most notably, they once again found that product advantage
was most strongly associated with financially successful products. Contrary to their earlier study, the authors found in this case that market

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

352

Academy of Management Review

April

competitiveness had no relationship with product success. These results

suggest that the effect of market competitiveness on project outcomes
needs further investigation.
The Stanford Innovation Project also emphasized product advantages,
market attractiveness, and internal organization. Seventy product success/failure pairs initially were surveyed and, from these, 21 case studies
were subsequently conducted (Maidique & Zirger, 1984, 1985). The third
study expanded the first two by examining 86 success/failure product
pairs (Zirger & Maidique, 1990). The data were gathered using a questionnaire that asked respondents to compare a product success with a
product failure within their firms. Twenty-three items were examined using factor analysis and then 2-group discriminant analysis (Zirger &
Maidique, 1990). The respondents were senior executives in the electronics industry who were attending an executive education program. Success was measured by whether executives considered the product to be a
financial loss or a profit contributor. All hypotheses were supported.
read like a blueprint for a rationally
The authors' conclusions
planned product-development effort. First, excellent internal organization was important (i.e., smooth execution of all phases of the development process by well-coordinated functional groups). For example, the
authors wrote, "Products are more likely to be successful if they are
planned and implemented well" (Zirger & Maidique, 1990: 879). Products
that had top management commitment and were built on existing corporate strengths were also likely to be successful. In addition, product factors were critical. Successful products provided superior customer value
through enhanced technical performance, low cost, reliability, quality, or
uniqueness. Finally, market factors also affected product success. Early
entry into large, growing markets was more likely to lead to success.
More recently, other authors have identified specific aspects of rational planning, such as predevelopment planning (Dwyer & Mellor, 1991)
and a focus on marketing and R&D involvement (Hise, O'Neal, Parsuraman, & McNeal, 1990), that correlate with product success. Another trend
is to focus not on financial success but rather specifically on the speed of
product development (e.g., Cordero, 1991; Mabert, Muth, & Schmenner,
1992). For example, Gupta and Wilemon (1990) focused on accelerating
product-development pace. These authors polled 80 executives concerning factors that slowed or accelerated the development processes. Their
suggestions for fast product development emphasized internal organization, including the importance of early cross-functional, customer, and
supplier involvement in the process and visible top management support,
more resources, and better teamwork (Gupta & Wilemon, 1990).
Overall, according to this stream of research, successful product development is the result of rational planning and execution. That is, successful products are more likely when the product has marketplace advantages, is targeted at an attractive market, and is well executed
through excellent internal organization. Specifically, internal organiza-

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

Brown and Eisenhardt

1995

353

tion is conceptualized as carefully planned predevelopment activities,

execution by competent and well-coordinated cross-functional teams
playing on the synergies of the firm, and significant support from top
management.
This broad-brush approach leads to an excellent and a comprehensive overview of the product-development process, which emphasizes
features of the product, internal organization, and the market. This same
breadth, however, also somewhat undermines the contributions of the
stream. To use a colloquialism, it is often difficult to observe the "new
product development" forest amid myriad "results" trees. The findings of
many studies read like a "fishing expedition"-too
many variables and
too much factor analysis. In this research stream, it is not uncommon for
a study to report 10 to 20 to even 40 or 50 important findings (e.g., Hise et
al., 1990; Rubenstein et al., 1976). Further, extensive bivariate analysis is
commonplace, and this blurs possible multivariate relationships.
Second, the research stream relies heavily on retrospective sense
making of complex past processes, usually by single informants. Individuals often are asked to quantify subjective judgments surrounding long
lists of success and failure factors. The frequent use of single informants
simply exacerbates these methodological problems. Thus, the research
results are likely to suffer from a host of attributional and other biases,
memory lapses, and myopia, which are associated with subjective, retrospective sense-making tasks.
Most important, the research in this stream often presents results
without relying on well-defined constructs. Thus, is it surprising that better products are more likely to be successful or that well-executed processes are likely to produce more successful products? Rather, the next
step is figuring out just what is a "better" product or just how do people go
about the "effective" execution to develop such a product. Research in this
stream is largely atheoretical as well, and so it fails to take the next
theory-building step. For example, Zirger and Maidique (1990) found that
entry into large, growing markets improves a project's performance. However, this result is not theoretically integrated with existing research that
warns of first-mover disadvantages (Lieberman & Montgomery, 1988) or
describes the power of imitation strategy (Bolton, 1993). Moreover, given
their often exploratory nature, studies in this stream often do not report
nonsignificant findings, which further inhibits theory building. Nonetheless, despite these shortcomings, this research stream has been enormously important, particularly in creating an early and a broad underof which factors are essential
standing
for successful
product
development and for emphasizing the role of the market in what is often
conceived of as a purely technical or organizational task.
Product Development

As Communication Web

A second stream of product-development research centers on communication. This research stream has evolved from the pioneering work of

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

354

Academy of Management Review

April

Allen at MIT (1971, 1977). The underlying premise is that communication

among project team members and with outsiders stimulates the performance of development teams. Thus, the better that members are connected with each other and with key outsiders, the more successful the
development process will be.
In contrast to the first perspective, this stream is narrowly focused on
one independent variable-communication.
Thus, these studies emphasize depth, not breadth as in the rational plan, by looking inside the
"black box" of the development team. They complement the rational lens
by including political and information-processing aspects of product development. The result is excellent theoretical understanding of a narrow
segment of the phenomenon. In this case, there also is greater methodological sophistication (e.g., multiple informants, multivariate analysis)
than in the first stream. Selected studies in this stream are highlighted in
Table 3, and a model is presented in Figure 3.
Some of the earliest empirical research along these lines was focused
on the flow of information in R&D groups (e.g., Allen, 1971, 1977). Often the
approach used was to have professionals keep track of their communications for some period of time. For example, Katz and Tushman (1981)
studied the performance of project groups in the R&D facility of a large
corporation. In 60 project groups, a total of 345 professionals kept track of
their communications for a randomly chosen day each week for 15 weeks.
The results of these early studies highlight the importance of external
communication to success. Specifically, these studies observed the presence of "gatekeepers"-(i.e.,
high-performing individuals who also communicated more often overall and with people outside their specialty)
(Allen, 1971). These gatekeepers brought information into the organization and dispersed it to fellow team members. The authors noted that
gatekeepers not only gathered and translated external information, but
they also facilitated the external communication of their fellow team
members (Katz & Tushman, 1981). Further, members of development
projects with gatekeepers performed better than those without, even after
accounting for the direct effect of the gatekeeper's high, personal performance. Teams with gatekeepers communicated more externally, leading
to improved project performance (Katz & Tushman, 1981). Finally, Von
Hippel (1986) noted how important communication with key customers
was regarding better product designs.
Other authors have built on this early work by Allen and colleagues.
For example, the content of external communication has been examined
closely by Ancona and Caldwell (1990, 1992a,b). These authors collected
questionnaires from 409 members of 45 new product-development teams
in 5 companies on communication and success patterns. Success was
measured by subjective team and management ratings of performance.
The authors found that team members communicated more with outsiders
who had similar functional backgrounds. Thus, when there were more
functions represented on the team, there was more external communica-

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

355

Brown and Eisenhardt

1995

~~~~~
-~~~~
~~~~~~

0
E

A.

o
0

tm0

o0

ID

.0
00

0)4

000

~ ~~0
X~~~~~~~

4)

.~~~~~~~~~r

).

0(D

0 .0

-~~~

~~~~~0)0

0)00

~~~00 000

00

00t.4

0n

~ ~~

0
b..
>..

0)

0~~~~~~~~~~~
a 4) 0
4)

rn

I!

V
0

0..

2)

&.

0.

o.

04

I0

j-0

2
0

V.

-..2

1:~~~m
0

t..
4)

0
2

00-

>~~~~~~~

cn

0~

cl

Old
C/i0

tN

U)o

.o

03

0
0)-

~ ~~~~0

O)

0~~~~~~~~~~~~~~~~.
0

020 O

5)

0t.

dO

N0)

u
.0~~~~~~.
ti

.9..
~~0

00Gi.

2n

CZ

04t

0)0

0l

O)

~~0

000

o2

-d0

0-)

0.

016~~~~~~~~~~~~~~~~~
0

2220
0

En

0 0
b-, 0
U) " E

0.
0
0)
r.)

N
O
0

0)
L)

0.

0.

LI)

~ ~

0
0.0U

0.

>~~~~~~~~~~

>
4)0

-4

00

O(C0

0~

l)0

00~~~~~~~~
0'~~~~~~~~~~C40)t
0)

0)
E-4

od

00~~~~

rnN

04

C5

0'
0

)0

ty)~~0

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

356

Academy of Management Review

Communication

April

FIGURE 3
Web Model of Product Development
Team
Composition
|Gatekeepers

*Moderaltetenure

Team\
Internal Communication
Performance
*High
*Experiential
*Iterative
*Nonroutine

*Various, frequently
perceptual measures

Project Leader
*Power
Team
External Communication
*High
*Ambassadorial,
task coordination

tion by the team as a whole and better management-rated performance

(Ancona & Caldwell, 1992a).
More important, the authors (Ancona & Caldwell, 1990) developed a
typology of external communication or "boundary-spanning" behaviors.
Ambassador activities consisted of political activities such as lobbying
for support and resources as well as buffering the team from outside
pressure and engaging in impression management. Task coordination
involved coordination of technical or design issues. Scouting consisted of
general scanning for useful information, whereas guard activities were
those intended to avoid the external release of proprietary information.
Ancona and Caldwell (1992b) subsequently identified group-level
strategies employed by the 45 product-development teams in their sample. One of this study's most interesting findings was that the frequency
of external communications was not a significant predictor of team performance. Rather, communication strategy was germane. The most successful product-development teams engaged in a comprehensive external
communication strategy, combining ambassador and task-coordination
behaviors that helped these teams to secure resources, gain task-related
information, and so enhance success. In contrast, less successful productdevelopment teams used strategies involving fewer types of external

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

1995

Brown and Eisenhardt

357

communication activities and less overall external communication. Thus,

more effective teams engaged in both political and task-oriented external
communications, suggesting that product-development teams must attend not only to the frequency of external communication, but also to the
nature of that interaction (Ancona & Caldwell, 1992b). Work by Benjamin
(1993) has led to similar conclusions regarding the importance of external
political communication and impression management.
There also has been interest in internal communication among team
members. For example, in his study of 32 project groups, Keller (1986)
found that internal group cohesion helped performance. Similarly, Ancona and Caldwell (1992a) found that teams with more thorough internal
communication (e.g., they defined goals better, developed workable
plans, and prioritized work) had superior performance. Also, Joyce (1986)
observed that powerful project leaders in matrix organizations often enhanced the quantity of internal communication but decreased its quality.
The critical cross-functional issues of internal communication have
been explored by Dougherty (1990, 1992; Dougherty & Corse, In press).
This research consisted of a multiple case, inductive study of the barriers
that hinder cross-functional communication. The study examined 18 new
product-development efforts in 5 firms. Failed products were those that
were canceled after introduction. Successful products were those that
were meeting or exceeding expectations after introduction. Later research
(Dougherty & Corse, In press) extended the sample to 40 projects, 134
people, and 15 firms. A key feature is emphasis on innovative product
development in large, established firms.
Dougherty (1990) demonstrated that various functional departments
were tantamount to "thought worlds," each with its own "fund of knowledge"-what
members know-and
"system of meaning"-how
members
know (Dougherty, 1992). Not surprisingly, individuals from different departments understood different aspects of product development, and they
understood these aspects in different ways. This difference led to varying
interpretations, even of the same information. Interestingly, what distinguished successful projects was not the absence or presence of these
barriers, but rather how they were overcome. For successful products,
cross-functional personnel combined their perspectives in a highly interactive, iterative fashion (Dougherty, 1990). This type of internal communication appeared to increase information content. In contrast, failed
products were characterized by sequential attention by functional groups
such that each departmental view dominated a particular phase of the
project.
Project teams also overcame cross-functional communication barriers
when team members participated in concrete tasks together and violated
routines such as usual relationships and divisions of tasks (e.g., Dougherty, 1992; Dougherty & Corse, In press). These tactics for organizing internal communication appeared to increase the information flow during
the communication process. Dougherty (1992) described, for example, how

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

358

Academy of Management Review

April

one team met with customers directly in focus groups to achieve a common team understanding of who the customer was. This common experience improved the information content of the communication as team
members developed a common understanding of the customer while
working together.
Finally, researchers also have been interested in how communication
affects the performance of teams over time. For example, Katz (1982) explored the relationship among the mean tenure of a team, the degree of
external communication, and performance. In his study of 50 productdevelopment teams in a large American corporation, he found that initially group performance increased with increasing mean tenure of the
group, but this relationship reversed and performance dropped off after
five years. The decline in performance was significantly correlated with
a decline in external communication.
In summary, the results indicate that external communication is critical to successful product development. However, this stream goes beyond this rather intuitive result to illustrate how teams increase their
external communication and what types of communication are important.
Specifically, successful product-development teams include gatekeepers,
who encourage team communication outside of their groups, and powerful project managers, who communicate externally to ensure resources for
the group. In addition, such teams also engage in extensive political and
task-oriented external communication. The underlying rationale is that
politically oriented external communication increases the resources of
the team, whereas task-oriented external communication increases the
amount and variety of information. These types of communication, in
turn, aid the development-process performance.
Similarly, internal communication improves development-team performance. For example, managers who are inwardly focused on the technical issues of the project will enhance internal communication and improve team performance. Cross-functional teams that structure their
internal communication around concrete tasks, novel routines, and fluid
job descriptions also have been associated with improved internal communication and successful products. These observations on how to break
down cross-functional barriers are particularly critical insights. Thus,
high internal communication increases the amount and variety of internal
information flow and, so, improves development-process performance.
Overall, two theoretical themes emerge in the literature. One, an
information-processing
view, emphasizes that frequent and appropriately structured task communication (both internal and external) leads to
more comprehensive and varied information flow to team members and,
thus, to higher performing development processes. The second, a resource dependence view, emphasizes that frequent political communication (typically external) leads to higher performing development processes by increasing the resources (e.g., budget, personnel, equipment)
available to the team.

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

1995

Brown and Eisenhardt

359

In contrast, the principal shortcoming of this perspective is that it is

so focused on communication by project team members that other factors
(e.g., organization of the work, product attributes, market attractiveness)
are neglected. There are also other problems. For example, although this
research is more methodologically sophisticated than the first stream
(e.g., multiple informants, multivariate analysis, tighter constructs), performance measures frequently are very subjective, and so it is difficult to
know whether the results would replicate for more objective measures of
performance, such as product profitability. In addition, the research in
this stream does not distinguish between different types of products, such
as incremental versus breakthrough versus platform products. However,
as Katz (1982) observed, such distinctions may affect appropriate types of
communication.
Despite these problems, this research stream has been influential,
particularly in highlighting the political and information-processing dynamics underlying the communication processes of successful productdevelopment teams. Thus, the in-depth focus of the communication web
complements the sweeping perspective of the rational plan.
Product Development

as Disciplined Problem Solving

A third stream of research is what we have termed the disciplined

problem-solving perspective. This stream evolved from studies of Japanese product-development practices in the mid-1980s (e.g., Imai et al.,
1985; Quinn, 1985). In this case, successful product development is seen as
a balancing act between relatively autonomous problem solving by the
project team and the discipline of a heavyweight leader, strong top management, and an overarching product vision. The result is a fast, productive development process and a high-quality product concept. Selected
studies are summarized in Table 4, and a corresponding model appears
in Figure 4.
Case-based research (Imai et al., 1985; Quinn, 1985; Takeuchi & Nonaka, 1986) laid the groundwork for this stream of research. For example,
Imai and colleagues (1985) studied seven successful product-development
efforts in five different Japanese companies across several industries. The
seven products included Fuji-Xerox's FX-3500 copier, the City box-car by
Honda, and the Canon Auto Boy (Sure Shot) camera. Performance was
defined in terms of speed and flexibility of development.
The authors found several management practices that were particularly effective for fast, efficient product development. One was the extensive use of supplier networks. The researchers observed that strong formal ties to suppliers and R&D networks were very important to the
product-development process. In such networks, suppliers can acquire a
very high level of technical skill in a specialized area, which allows them
to fulfill sudden or unusual requests quickly and effectively.
Imai and colleagues (1985) also observed a problem-solving strategy
involving cross-functional
development
teams that aided effective

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

Academy of Management Review

360

En
(n

En

::s

U)

En
En

O.

4)

.-u0 O)Z
u -a >

U0

>
OZ 4)

0u 0

0>
a)

0)

::s
>
O

t3)

-0
0

0 En
ty) u 0

O)

U)
m

u 0

u
0.

0
0

0
.0

0
.0

En

O)

O)

r)

u
O)

O)

En

U)
0

a)

ui
O)

u u O)

u
O

u
En

::s0

>

-ty)

En
0)

En
En

::1

April

En
u

En

-2

-6
u

>
--I
0

cr

En
0

a)

::s 0

::10 0

u
C.D
CO

C)
CN

L-O

FA
u

0
rn

P.,

rn '24 u

0
0

0
U

> 0

En U
CIQ

CN

o-6

00
En

OC)

tl r)
-.4

CN

K4
0c)

'E3E. z

u
0

0 C
a) 00

cli

0 a)
E

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

z
0

361

Brown and Eisenhardt

1995

0~~~~~~~~~~~

~~~~~~~~~~~~~04

0~~~~~~~~~
34~~~~~~~~~~~~~~u:
U

O 0

*.

3.4.4

0~~~~~~
0~~~~~~~~~~~~~~~0

3P.4

-.4

0,

04

~ ~

4)~~.4

~0

~ ~
~~~~~~ 26~~~~~~~~~~~~6

p4~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4
.-~~~~~~~~~~~~~~~r
Ow

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

362

Academy of Management Review

April

product development. When the development team was composed of

members with varied functional specializations, team members had access to more diverse information. In addition, cross-functional teams permitted the overlap of development phases, which also quickened the pace
of product development. Furthermore, the authors observed that product
development was accelerated by overlapping of development phases and
cross-functional teams only if supported by continuous communication
among project members. This communication increased the information
flow among team members, making it easier for team members to understand each other's specialties and to coordinate overlapped development
phases. The creation of redundant information (Nonaka, 1990) and an
emphasis on extensive multifunctional training (Imai et al., 1985) also
were related to this style of problem solving.
Finally, Imai and colleagues (1985) noted that rather than playing just
a supportive role, as suggested by the rational plan studies (e.g., Cooper
& Kleinschmidt, 1987; Gupta & Wilemon, 1990; Zirger & Maidique, 1990),
senior management should engage in "subtle control." The key idea behind subtle control is that members of successful project teams maintain
a balance between allowing ambiguity, such that creative problem solving can flourish at the project team level, and exercising sufficient
control, such that the resulting product fits with overall corporate competencies and strategy. The authors found that, for the best performance
results, senior management engaged in subtle control by communicating
a clear vision of objectives to their teams while simultaneously giving
team members the freedom to work autonomously within the discipline of
that vision.
Later research replicated and extended this early work. For example,
several Harvard researchers
studied the management
of productdevelopment projects in the auto industry (Clark et al., 1987; Clark &
Fujimoto, 1991; Hayes, Wheelwright, & Clark, 1988). These very impressive data consist of in-depth case studies of 29 major car-development
projects across 20 companies-3
American, 8 Japanese, and 9 European.
Similar to the aforementioned study by Imai and his colleagues, this
research examined new models of established products, not breakthrough products. This auto industry data measured the performance of
the product-development process along three dimensions: total product
quality, lead time, and productivity.
The authors (e.g., Clark et al., 1987; Clark & Fujimoto, 1991) replicated
earlier findings. They reported that extensive supplier networks coupled
with overlapping product-development
phases, communication, and
cross-functional groups (what they term integrated problem solving) improved the performance of development teams. The authors also introduced two central concepts: heavyweight team leaders and product integrity. These two concepts help to clarify the meaning of subtle control.
"Heavyweight" team leaders are powerful "linking pins" who, on the
one hand, coordinate the activities of a product-development team and,

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

1995

Brown and Eisenhardt

363

on the other hand, work with senior management to create an overarching

product concept. Thus, senior management can exercise subtle control
through the use of such leaders who manage their teams in the context of
a product vision. The findings indicate that these team leaders are able to
gain resources, command respect, and break down traditional functional
allegiances while simultaneously building a strong product vision (Clark
et al., 1987; Clark & Fujimoto, 1991; Hayes et al., 1988).
"Product integrity" (Clark & Fujimoto, 1991), unlike the dimensions of
product attractiveness in the rational plan perspective such as low price,
describes the notion of a product's being consistent with the corporate
image. Product integrity implies a clear vision of the product's intended
image, performance, and fit with corporate competencies and customers.
By focusing on establishing product integrity, senior management can
ensure that an overall vision for the product is communicated to the project team and, thus, balance the autonomy gained through heavyweight
leadership.
Finally, this study also added an emphasis on predevelopment activities. Hayes and others (1988) described how bringing conflicts to the
surface early in the development process was an important factor in successful development projects. By resolving conflicts through mutual accommodation at low levels in the organizational hierarchy, a clear project
vision was established early on, which subsequently sped up the development process.
A study related to the Harvard auto industry research, the MIT International Motor Vehicle Program, examined lean versus mass production
in the auto industry. This study (Womack, Jones, & Roos, 1990) presented
Clark and Fujimoto's (1991) data along with some consolidated data from
secondary sources. Not surprisingly, many of their conclusions, such as
the importance of powerful leaders, overlapping development phases,
predevelopment activity, and internal communication to product development, replicated those of the Harvard study (Clark et al., 1987; Clark &
Fujimoto, 1991; Hayes et al., 1988).
More recent research in this stream emphasized industries with less
domination by Japanese competitors and more scientific content. For example, Iansiti (1992, 1993) deductively examined the mainframe computer
industry. His data consisted of 27 in-depth studies that represent all the
major products developed by the 12 chief competitors in the mainframe
computer industry (from Japan, the United States, and Europe) during the
1980s. The author focused on the development of technologies associated
with the packaging and interconnect system of the mainframe processor.
The primary result is that a high system focus (i.e., a combination of
technical integration, exposure to systems integration, and accumulation
of interaction knowledge) predicted both lead time and productivity. Similar to product integrity, system focus implied concern for how technology
choices for a given component fit with the product as a whole. Like predevelopment activities, system focus also involved early planning for the

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

364

Academy of Management Review

April

integration of new technology, product expectations, and manufacturing

systems into the problem-solving process. Thus, this result strengthened
earlier findings that predevelopment planning and product integrity enhanced the performance of product-development teams.
A study by Eisenhardt and Tabrizi (In press) examined the computer
industry in Japan, Europe, and the United States. This deductive study
considered 72 products in 36 firms, including the personal, workstation,
mainframe, and peripherals segments of the industry. Thus, the average
product life cycle was substantially shorter than in the mainframe computer or automobile industry studies. The data were collected on site
using research teams who helped firm personnel gather data. The focus
was on speed of development as the performance measure. Specifically,
the authors compared a compression model of fast product development
with an experiential approach.
The unique insight of this study is that fast product development was
associated with the experiential approach (Eisenhardt & Tabrizi, In
press). Product teams who engaged in more experiential or improvisational product design through frequent iterations, more testing, frequent
milestones, and powerful leadership developed products more quickly. In
contrast, attempts simply to compress the product-development cycle
through the use of computer-aided design (CAD), rewards for schedule
attainment, supplier involvement, overlapping development stages, or
extensive planning not only did not accelerate pace, but, in fact, often
slowed it. The exception was the mainframe segment in which the use of
CAD, supplier involvement, and overlap effectively accelerated the pace.
Thus, these results suggest that there are two relevant problem-solving
models for organizing product development. One focuses on factors such
as planning and overlap that are relevant for more stable products in
mature settings (e.g., Clark & Fujimoto, 1991; Iansiti, 1992), and the other
focuses on experiential product design that is relevant for less predictable products in uncertain settings, such as personal computers, workstations, and peripherals. Finally, as in the previous research, crossfunctional teams sped up development for all industry segments.
In summary, this stream of research envisions successful product
development as disciplined problem solving. That is, successful product
development involves relatively autonomous problem solving by crossfunctional teams with high communication and the organization of work
according to the demands of the development task. This perspective also
highlights the role of project leaders and senior management in giving
problem solving a discipline-a
product vision. There is an emphasis on
both project and senior management, on the one hand, to provide a vision
or discipline to the development efforts and yet, on the other hand, to
provide autonomy to the team. Thus, this stream portrays product development as a balancing act between product vision developed at the executive level and problem solving found at the project level.
In contrast to the rational plan stream, this stream is more specific

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

1995

Brown and Eisenhardt

365

about the effective organization of work and is more focused on the development process and product concept than on the financial success of
the product. In contrast to the communication web perspective, this
stream has a broader scope and considers the role of suppliers and senior
management in addition to project leaders and teams. Methodologically,
the data are much richer and more detailed than the single-informant
information that underlies much of the rational stream. Theoretically, this
perspective extends the information-processing view of the communication web research by emphasizing not only the amount and variety of
information, but also its organization into problem-solving strategies.
However, the stream suffers from several shortcomings. One is that
there is a lack of political and psychological realism. In comparison with
the communication web, there is naive understanding of the political
realities of product development, such as the dependence of project teams
on external actors for resources. From a psychological standpoint, there is
little appreciation of the problems of actually motivating people and making cross-functional teams, high communication, and overlapping work.
Moreover, the heavyweight leaders seem almost "superhuman" in their
skills and duties.
Second, some of the constructs are challenging to comprehend. For
example, subtle control, product vision, and system focus are vague concepts. Even with attempts at clarification (e.g., Clark & Fujimoto, 1990;
Clark & Wheelwright, 1992), product integrity and heavyweight project
leader concepts also remain hazy. Although this lack of clarity may reflect
the complexity of the subject (or perhaps the often non-Western origin of
the ideas), it also impairs the usefulness of the perspective.
Finally, there is an extensive reliance on a Japanese viewpoint. Even
though Japanese comparisons have been critical to improving thinking,
Japanese industrial dominance sometimes makes it unclear which features are important to product development and which are simply Japanese. This concern seems particularly relevant to the supplier network
findings (e.g., Clark & Fujimoto, 1991; Imai et al., 1985), which are dependent on the specifics of the Japanese industrial infrastructure. Nonetheless, the image of product development as disciplined problem solving is
a powerful and sophisticated metaphor for successful product development.

TOWARDAN INTEGRATIVE
MODELOF PRODUCTDEVELOPMENT
In the previous section, we described three streams of productdevelopment research. These streams evolved from different sources and
focused on somewhat different aspects of product development. However,
as we observed previously, they also offer complementary and sometimes
overlapping insights into product development. In this section, we rely on
these insights to provide the basis of an integrative model.
Key to developing this model is the observation that the three streams
focus on both overlapping and complementary sets of constructs. The

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

Academy of Management Review

366

April

rational plan perspective contributes a sweeping view of product development, including team, senior management, market, and product characteristics to predict financial success. In contrast, the problem-solving
perspective has a more deeply focused view on the actual development
process (i.e., those project team and management factors that contribute
to a better product-development process and a more effective product
concept). The concepts are more elaborated than in the rational plan
perspective so that, for example, the "effective execution" of the rational
plan is given more concrete and detailed development in the problemsolving perspective. The communication web perspective is narrower
still. This research centers on a very specific, although important, aspect
of product development, namely on internal and external communication
by project team members.
Also key to developing an integrative model is the observation that
the streams have complementary theoretical approaches. The rational
plan perspective is largely atheoretical, consisting of collections of associations. In contrast, the product-solving perspective has a cognitive theoretical orientation, which links ideas about information and its organization to effective problem solving. Finally, the communication perspective
relies on a simpler but consistent theoretical view of information (i.e., the
amount and variety of information) and a complementary political perspective, which emphasizes the need for resources.
These overlapping and complementary focal interests as well as the
theoretical complementarities suggest that the streams are ready for synthesis into an integrative model. In this section, we craft such an integrated conceptual perspective that summarizes key findings within the
literature and provides a departure point for future studies of product
development. The resulting model is depicted in Figure 1.
Model Overview
The organizing idea behind the model in Figure 1 is that there are
multiple players whose actions influence product performance. Specifically, we argue that (a) the project team, leader, senior management, and
suppliers affect process performance (i.e., speed and productivity of product development), (b) the project leader, customers, and senior management affect product effectiveness (i.e., the fit of the product with firm
competencies and market needs), and (c) the combination of an efficient
process, effective product, and munificent market shapes the financial
success of the product (i.e., revenue, profitability, and market share).
Underlying these relationships are the theoretical underpinnings that
we have identified from the combined research streams. Thus, process
performance is driven by the amount, variety, and problem-solving organization of information and by the resources available to the team. Product effectiveness is driven by the input of leaders, senior management,
and customers into the formation of a clear product vision (a less wellunderstood process). Both product effectiveness and process performance

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

Brown and Eisenhardt

1995

367

influence the financial success of the product. We turn now to a description of the model, arranged according to the key players.
Project Team
The heart of the product-development process and the focus of much
research is the project team. Project team members are the people who
actually do the work of product development. They are the people who
transform vague ideas, concepts, and product specifications into the design of new products. Not surprisingly, then, the project team is central to
our model of product development. Specifically, we argue that the composition, group process, and work organization of the project team affect
the information, resources, and problem-solving style of the team. These,
in turn, ultimately influence process performance (i.e., speed and productivity of the process) (Figure 1).
Regarding composition, consistent with all three streams, crossfunctional teams are critical to process performance (e.g., Clark & Fujimoto, 1991; Dougherty, 1992; Zirger & Maidique, 1990). We define crossfunctional teams as those project groups with members from more than
one functional area such as engineering, manufacturing, or marketing.
The underlying reasoning is that the functional diversity of these teams
increases the amount and variety of information available to design products. This increased information helps project team members to understand the design process more quickly and fully from a variety of perspectives, and thus it improves design process performance. Moreover,
the increased information helps the team to catch downstream problems
such as manufacturing difficulties or market mismatches before they happen, when these problems are generally smaller and easier to fix. Thus,
consistent with the empirical support in all three research streams, crossfunctional teams are associated with high-performing processes. Moreover, this link (1) in the model is among the most important and empirically robust (e.g., Clark & Fujimoto, 1991; Dougherty, 1992; Zirger &
Maidique, 1990).
Gatekeepers are another important facet of project team composition.
Gatekeepers are individuals who frequently obtain information external
to the group and then share it within the project team. Similar to crossfunctional teams, they affect process performance by increasing the
amount and variety of information available in the design process. Gatekeepers expose the project team members to more and diverse information such as new technical developments occurring outside the group.
Although gatekeepers are probably less important in cross-functional
teams because the members of such teams have natural outside contacts
in their functional homes, gatekeepers clearly increase the external information reaching the project team (link 1 in the model) (Allen, 1971, 1977;
Katz & Tushman, 1981).
Finally, team tenure is a third composition factor that plays a role in
influencing process performance. Teams with a short history together

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

368

Academy of Management Review

April

tend to lack effective patterns of information sharing and working together (Katz, 1982). Thus, the amount and variety of information that can
be communicated among project team members is limited by this unfamiliarity. In contrast, teams with a long tenure together tend to become
inward focused and neglect external communication (Katz, 1982). This
tendency restricts the information and resources from outside the team
that the team members receive. Because neither of these situations is
desirable, process performance is highest when team tenure is at moderate levels. At this level of tenure, team members are most likely to
engage in both extensive internal and external communication and,
therefore, to receive maximum benefit. As Katz found (1982), this leads to
higher project performance (link 1 in the model).
Another important factor affecting process performance is group process, especially communication. Results from all three research streams
indicate that effective group processes, particularly those related to comincrease information and so are essential
munication,
for highperforming development processes (e.g., Imai et al., 1985; Katz, 1982; Zirger & Maidique, 1990). In the case of internal communication (e.g.,
Dougherty, 1990; Keller, 1986), frequent communication increases the
amount of information directly in that more communication usually yields
more information. More subtly, frequent communication also builds team
cohesion, which then breaks down barriers to communication and so increases the amount of information as well (Keller, 1986). Moreover, especially when this communication is effectively structured (e.g., includes
concrete communication surrounding shared group experiences and nonroutine rule breaking [Dougherty, 1992]), it cuts misunderstandings and
barriers to interchange so that the amount of information conveyed is
increased. This, in turn, improves the speed and productivity of the entire
development process (Dougherty, 1992).
In the case of external communication, frequent communication with
outsiders such as customers, suppliers, and other organizational personnel opens the project team up to new information (e.g., Clark & Fujimoto,
1991; Imai et al., 1985; Katz, 1982; Katz & Tushman, 1981). When this external communication is task oriented, team members gain information
from diverse viewpoints beyond those of the team. Further, when the
communication is frequent, project teams are likely to develop an absorptive capacity such that they become more efficient in gaining and using
the information being conveyed. Both of these factors should improve the
productivity and pace of the development process. Also important is external communication in the form of political activities such as lobbying
for resources, engaging in impression management, and seeking senior
management support for the project (Ancona & Caldwell, 1992b). Taken
together and consistent with previous research, internal and external
communication both increase the amount and variety of information and
the resources available to the project team. These, in turn, improve process performance. These links (link 2 in the model) are also among the

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

1995

Brown and Eisenhardt

369

most empirically robust (e.g., Ancona & Caldwell, 1992b; Clark & Fujimoto, 1991; Imai et al., 1985; Katz, 1982; Katz & Allen, 1985; Katz & Tushman, 1981).
The third team feature of the model is the problem-solving strategy by
which team members organize their work. As described in the problemsolving stream, several empirical findings suggest a contingent set of
problem-solving strategies that are appropriate for different types of
tasks (Eisenhardt & Tabrizi, In press). Each strategy represents a different
structuring of information. For stable and relatively mature products such
as automobiles (Hayes et al., 1988; Womack et al., 1990) and mainframe
computers (Eisenhardt & Tabrizi, In press; Iansiti, 1992), product development is a complex task for which tactics such as extensive planning and
overlapped development stages are appropriate. These tactics assume a
certain but often complex problem-solving task that can be rationalized.
For example, consistent with the rational and problem-solving perspectives, planning (e.g., Cooper & Kleinschmidt, 1987; Hayes et al., 1988;
Iansiti, 1992; Zirger & Maidique, 1990) improves the speed and productivity of the development process by eliminating extra work, rationalizing
and properly ordering the steps of the process, and avoiding errors. Similarly, overlapped development stages improve process performance
(Clark & Fujimoto, 1991; Imai et al., 1985) by allowing at least partially
simultaneous execution of development steps such as design and test,
which often are viewed as sequential. In effect, the development process
is squeezed together (e.g., Clark & Fujimoto, 1991).
In contrast, when there is more uncertainty in the design process,
such as in rapidly changing industries (e.g., microcomputers), more experiential tactics, including frequent iterations of product designs, extensive testing of those designs, and short milestones (i.e., short time between successive
milestones)
improve process performance (e.g.,
Eisenhardt & Tabrizi, In press). The underlying idea is that under conditions of uncertainty it is not helpful to plan. Rather, maintaining flexibility and learning quickly through improvisation and experience yield effective process performance (e.g., Eisenhardt & Tabrizi, In press; Miner &
Moorman, 1993; Weick, 1993). However, although these ideas may be attractive, they have not been examined extensively, and the contingencies
between these factors and the previously mentioned factors of overlap
and planning have not been well explored empirically (link 3 in the
model). As we will discuss, further elaboration and testing of this part of
the model are prime opportunities for future research.
Project Leader
Even though the cross-functional team is the heart of efficient product
development, the project leader is the pivotal figure in the development
process. Consistent with the communication and problem-solving perspectives, the project leader is the linking pin or bridge between the
project team and senior management. Therefore, as indicated in Figure 1,

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

370

Academy of Management Review

April

the project leader critically affects both the process performance (i.e.,
speed and productivity of the development process) and the effectiveness
of the product (e.g., Clark & Fujimoto, 1991; Joyce, 1986; Katz & Allen,
1985). Several characteristics of the project team leader are particularly
germane.
One central characteristic is the power of the project leader. By powerful leaders we mean those project leaders with significant decisionmaking responsibility, organizationwide authority, and high hierarchical
level. Such leaders are particularly able to improve process performance.
The underlying rationale is that such leaders are highly effective in obtaining resources such as more personnel and larger budgets for the project team. As Ancona and Caldwell (1992b) observed, powerful project
leaders are particularly effective politicians in lobbying for resources,
protecting the group from outside interference, and managing the impressions of outsiders. In contrast, less powerful project leaders are likely to
be less successful in gaining needed talent and financial support and in
shielding the team from outside interference. In addition, powerful leaders also may command greater respect and, thus, may be able to attract
better project team members to the group and to keep groups focused and
motivated (Clark & Fujimoto, 1991). These qualities, in turn, create a
faster and more productive development process. Moreover, because the
importance of a powerful leader has been demonstrated in both the communication and problem-solving streams (e.g., Clark & Fujimoto, 1991;
Katz & Allen, 1985) (link 4 in the model), it is a robust link.
A second, important characteristic of the project leader is vision. Vision involves the cognitive ability to mesh a variety of factors together to
create an effective, holistic view and to communicate it to others. Specifically, in the case of product development, this means meshing together
firm competencies (e.g., particular technical, marketing, or other skills)
and strategies with the needs of the market (e.g., consumer preferences
for style and cost) to create an effective product concept. As described in
the problem-solving perspective, this is a critical characteristic of project
leaders because they often are central to the creation of the product concept. Project leaders, together with senior management, frequently shape
the overall product concept and communicate it to project team members
(Clark & Fujimoto, 1991). However, even though this aspect of project
leadership is, we think, compelling, our understanding of exactly what
vision is, what an effective product is, and the theoretical links between
the two is very weak (link 5 in the model). As we discuss further in our
suggestions for future research, understanding this creative task is an
important research direction.
Finally, project team leaders also are small-group managers of their
project teams. Surprisingly, however, there is very little research about
appropriate internal management skill for project leaders beyond studies
of matrix communication and leadership (Joyce, 1986; Katz & Allen, 1985).

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

Brown and Eisenhardt

1995

371

Perhaps, simply, the general research on leadership of small groups applies here.
Senior Management
Although the team and project leader are critical in the productdevelopment process, senior management is important as well. Consistent with studies in the rational stream, senior management support is
critical to successful product-development
processes (e.g., Cooper &
Kleinschmidt, 1987; Gupta & Wilemon, 1990; Rothwell, 1972; Zirger &
Maidique, 1990). By support we mean the provision of resources to the
project team, including both financial and political resources. The underlying reasoning is that this support is essential for obtaining the resources necessary to attract team members to the project, to gain project
approval to go ahead, and to provide the funding necessary to foster the
development effort. Thus, as shown in Figure 1, senior management support is essential for fast and productive product development. This link (6
in the model) is well supported in the literature (e.g., Cooper & Kleinschmidt, 1987; Gupta & Wilemon, 1990; Katz & Allen, 1985; Zirger & Maidique, 1990).
Second, as described by Imai and colleagues (1985) in the problemsolving perspective, the ability of senior management to provide what
they term subtle control also is important to both superior process performance and effective products. Much like the vision of project leaders,
subtle control involves having the vision necessary to develop and communicate a distinctive, coherent product concept. As we described previously, senior management and project leaders often work together to develop such a product concept. At the same time, subtle control also
involves delegation by senior management to project teams such that
they have enough autonomy to be motivated and creative. We argue that
such creativity and motivation are likely to yield a better development
process. However, as in the case of project leader vision, the subtle control and product effectiveness concepts and their theoretical links are
blurred and lack rigorous empirical examination (link 7 in the model). We
will discuss this lack of clarity in our suggestions for future research.
Suppliers and Customers
The final key players in the model are suppliers and customers. Previous research has associated a faster development process with early
(Gupta & Wilemon, 1990) and extensive (Clark & Fujimoto, 1991; Imai et
al., 1985) supplier involvement. As explained in the problem-solving
stream, extensive supplier involvement in product design can cut the
complexity of the design project, which in turn creates a faster and more
productive product-development
process. Such involvement also can
alert the project team to potential downstream problems early on, at a
point when they are easier and faster to fix. Customer involvement also

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

Academy of Management Review

372

April

has been shown to improve the effectiveness of the product concept in the
rational plan stream (e.g., Cooper & Kleinschmidt, 1987; Zirger & Maidique, 1990). However, it is not clear exactly how or when suppliers and
customers are appropriately involved in the development process, and
the evidence is not unaminous (e.g., Eisenhardt & Tabrizi, In press). Thus,
even though the participation of these outside constituents is probably
important, the empirical literature is imprecise (links 8 and 9 in the
model). Again, we address this in our discussion of future research.
Financial Success
The previous discussion linked the key players in product development to process performance and product effectiveness. The final portion
of the model combines these two factors with characteristics of the market
to predict the financial performance of the product.
The underlying rationale for the link (10 in the model) between process performance and financial success of the product is twofold. A productive process means lower costs and thus, lower prices, which, in turn,
should lead to greater product success. Second, a faster process creates
strategic flexibility and less time to product launch, both of which may
lead to financially successful products. The second predictor of the
financial success of the product is product effectiveness (link 11 in the
model). As shown in the rational plan literature (e.g., Cooper & Kleinschmidt, 1987; Zirger & Maidique, 1990), product characteristics such as
low-cost and unique benefits and fit-with-firm competencies create financially successful products. Presumably, such products are more attractive
to consumers.
Finally, consistent with research in the rational stream (Cooper &
Kleinschmidt, 1987; Zirger & Maidique, 1990), the third link (12 in the
model) ties a munificent market to financial success. Specifically, we
define a munificent market as one that is large and growing and has low
competition. The reasoning is that such markets offer the possibility of
large sales and, in the case of growing markets, competitive instability
that may favor new products.
Overall, the argument is that a strong product-development process,
an attractive product, and a munificent market should lead to a financially successful product. However, although these links have some substantiation in the rational plan literature and seem plausible, they have
received little rigorous testing and rely on limited theoretical logic. As we
will discuss, this portion of the model presents another excellent opportunity for future research.
AGENDA FOR FUTURE RESEARCH
As noted in the previous section, many of the concepts and theoretical
links presented in the model in Figure 1 have been well studied. However,

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

1995

Brown and Eisenhardt

373

some concepts are less sharply defined, and some theoretical links are
not well tested. These shortcomings present research opportunities.
One research opportunity is to examine the primary links of the
model-that
is, the links among process performance, effective product,
market factors, and financial performance. As was noted, these links
have been primarily empirically examined in the rational plan research
stream. However, because the methodology in this research stream so
often involves subjective, retrospective responses by single informants
and bivariate analysis, the validity of these links is tenuous. Thus, a test
of these fundamental theoretical links would be useful. A related research opportunity is determining the relative importance of these factors. For example, although Cooper and Kleinschmidt (1987, 1993) found
that the market was not as relevant as other variables in predicting product success, it would be useful to examine the robustness of this claim.
Another related opportunity is to examine whether process performance, product effectiveness, and munificent markets are actually independent variables. For example, market factors may moderate the relationship of process performance and product effectiveness to financial
success.2 It may well be that process performance and product effectiveness are important predictors of financial success only in poor markets,
whereas most products will be successful in munificent markets. This
reasoning reemphasizes the need to garner a better understanding of the
relative importance that these factors have in driving financial performance.
A second area of research is the organization of work. As was noted,
two models have emerged to describe alternative organizations of work.
One is the fairly well-studied model that includes extensive planning and
overlapped development stages (e.g., Hayes et al., 1988; Iansiti, 1993) that
was developed in the context of complex products in mature markets. A
more recent model, related to improvisational thinking, emphasizes experiential product development such as frequent iterations, testing, and
milestones (Eisenhardt & Tabrizi, In press), yet this second model has
received only limited empirical examination. Indeed, other research suggests that our understanding is incomplete. For example, Tyre and Hauptman (1992) argued that the degree of system-level change is another contingency factor in the organization of work. In addition, a provocative
study by Benghozi (1990) of a massive, long-term innovation project in the
French telecommunication industry sketches a third model. For huge and
lengthy projects, Benghozi (1990) suggested that innovation routines,
which include dynamic planning, monitoring, and scheduling projects
over time as the environment changes, are needed. Overall, the point is
that exploring contingent models for the organization of work is an important path for future research.

We thank an anonymous AMR reviewer for this insightful suggestion.

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

374

Academy of Management Review

April

Third, our understanding of how senior managers affect development

is incomplete. They are consistently found to be important contributors to
project success (e.g., Clark & Fujimoto, 1991; Cooper & Kleinschmidt, 1987;
Zirger & Maidique, 1990). However, the management-related concepts in
Figure 1 such as vision, subtle control, and even support are vague. There
is also little understanding of the links between product effectiveness and
the creative processes by which senior managers and others match firm
competencies with market needs to create an effective product concept.
This process has been virtually unexplored. In addition, previous research (and thus our model) is vague regarding how the responsibilities
of senior management are distinct from the responsibilities of project
leaders. For example, should senior managers or project leaders be responsible for ensuring that products are synergistic with the core competencies of the firm (Leonard-Barton, 1992)? Most important, the model
echoes the product-development literature in conceptualizing only two
levels of management: a project leader and senior management, yet in
reality, frequently there are several levels of management with presumably different responsibilities. How do these levels interact? How do these
members of different levels of management affect technology strategies
and the management of multiple product-development projects for which
they are responsible? Thus, the concepts surrounding senior management and its link to product effectiveness offer prime opportunities for
future research.

CONCLUSION
We began this paper by noting that product development is the nexus
of competition for many firms as well as the central organizational process for adaptation and renewal. Simply put, product development is
critical to the viability of firms and an important core competence, yet it
is challenging to understand the findings of the related empirical research because this literature is so fragmented and varied.
The article has three conclusions.
One is that the productdevelopment literature can be organized into three streams of research:
product development as rational plan, communication web, and disciplined problem solving. We have highlighted these three streams of research as well as their key findings, strengths, and weaknesses. Second,
we conclude that these streams can be synthesized into a model of factors
affecting product-development success. Given the complementary nature
of past research, we were able to craft an integrated conceptual perspective that combines many of the empirical findings. Third, we conclude
that there are research implications for the future based on the mixture of
support for various findings in the model. We developed a brief outline of
potential research paths based on constructs and theoretical links that
are fuzzy or less explored in our model and the literature as a whole.
Overall, this article attempts to contribute an understanding of past literature, a model of current thinking, and a vision for future research.

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

1995

Brown and Eisenhardt

375

We end by looping back to the wider innovation literature. As we

noted at the outset, there are two broad areas of inquiry (e.g., Adler, 1989)
that complement one another. The economics-oriented branch offers an
understanding of innovation across industries and the evolution of technologies that provides a useful context for thinking about product development. However, even in more organizational-level work in this branch
(e.g., Nelson & Winter, 1982), the actual process of product development is
still largely a "black box." At best, this work simply describes the evolution of idiosyncratic innovation routines within organizations (Nelson &
Winter, 1982). More often, there are no organizational effects at all. Complementary
to this branch of innovation
the productresearch,
development literature opens up that black box by providing depth and
rich understanding of how actual products are developed within firms, a
critical core capability for many firms. This research indicates the organizational structures, roles, and processes that are related to enhanced
product development. Although much remains to be explored, the product-development branch remains essential for a complete picture of innovation.

REFERENCES
Adler, P. 1989. Technology strategy: A guide to the literatures. In R. Rosenbloom & R. Burgelman (Eds.), Research on technological innovation, management, and policy, vol 4: 25151. Greenwich, CT: JAI Press.
Allen, T. J. 1971. Communications,
R&D Management, 1: 14-21.

technology transfer, and the role of technical gatekeeper.

Allen, T. J. 1977. Managing the flow of technology. Cambridge, MA: MIT Press.
Ancona, D. G., & Caldwell, D. F. 1990. Beyond boundary spanning: Managing external
dependence in product development teams. Journal of High Technology Management
Research, 1: 119-135.
Ancona, D. G., & Caldwell, D. F. 1992a. Demography and design: Predictors of new product
team performance. Organization Science, 3: 321-341.
Ancona, D. G., & Caldwell, D. F. 1992b. Bridging the boundary: External process and performance in organizational teams. Administrative Science Quarterly, 37: 634-665.
Benghozi, P. 1990. French telecom. Organization Studies, 4: 531-554.
Benjamin, B. 1993. Understanding the political dynamics of developing new products. Working paper, Stanford University Graduate School of Business, Stanford, CA.
Bolton, M. K. 1993. Imitation versus innovation: Lessons to be learned from the Japanese.
Organizational Dynamics (3): 30-45.
Brown, S. L., & Eisenhardt, K. M. 1995. Core competence in product innovation: The art of
managing in time. Working paper, Department of Industrial Engineering and Engineering Management, Stanford University, Stanford, CA.
Burgelman, R., & Sayles, L. 1986. Inside corporate innovation. New York: Free Press.
Business Week. 1992. Inside Intel. June 1: 86-90, 92, 94.
Clark, K. B., Chew, W. B., & Fujimoto, T. 1987. Product development
industry. Brookings Papers on Economic Activity, 3: 729-781.

in the world auto

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

376

Academy of Management

April

Review

Clark, K. B., & Fujimoto, T. 1990. The power of product integrity. Harvard Business Review,
68(6): 107-118.
Clark, K. B., & Fujimoto, T. 1991. Product development performance. Boston: Harvard Business School Press.
Clark, K. B., & Wheelwright, S. C. 1992. Organizing and leading "heavyweight" development teams. California Management Review, 34(3): 9-28.
Cooper, R. G. 1979. The dimensions of industrial new product success and failure. Journal of
Marketing, 43: 93-103.
Cooper, R. G., & Kleinschmidt, E. J. 1987. New products: What separates winners from losers? Journal of Product Innovation Management, 4: 169-184.
Cooper, R. G., & Kleinschmidt, E. J. 1993. Major new products: What distinguishes the winners in the chemical industry? Journal of Product Innovation Management, 10: 90- 111.
Cordero, R. 1991. Managing for speed to avoid product obsolescence:
Journal of Product Innovation Management, 8: 283-294.

A survey of techniques.

David, P. 1985. Clio and the economics of QWERTY. American Economic Review, 75: 332337.
Dosi, G. 1988. Sources, procedures, and microeconomic
Economic Literature, 26: 1120-1171.
Dougherty, D. 1990. Understanding
Journal, 11: 59-78.

effects of innovation. Journal of

new markets for new products. Strategic Management

Dougherty, D. 1992. Interpretive barriers to successful

Organization Science, 3: 179-202.

product innovation in large firms.

Dougherty, D., & Corse, S. M. In press. When it comes to product innovation, what is so
"bad" about "bureaucracy?" Journal of High Technology Management Research.
Dumaine, B. 1991. Closing the innovation gap. Fortune, December 2: 56-59, 62.
Dwyer, L., & Mellor, R. 1991. Organizational environment, New product process activities,
and project outcomes. Journal of Product Innovation & Management, 8: 39-48.
Eisenhardt, K. M., & Tabrizi, B. In press. Accelerating adaptive processes: Product innovation in the global computer industry. Administrative Science Quarterly.
Gerstenfeld, A. 1976. A study of successful projects, unsuccessful projects and projects in
process in West Germany. IEEE Transactions in Engineering Management, 23: 116-123.
Gupta, A. K., & Wilemon, D. L. 1990. Accelerating the development of technology-based
products. California Management Review, 32(2): 24-44.

new

Hargadon, A. B., & Sutton, R. I. 1994. Exploiting exploration: The routinization of innovation
in the product-development process. Working paper, Department of Industrial Engineering and Engineering Management, Stanford University, Stanford, CA.
Hayes, R. H., Wheelwright, S. C., & Clark, K. 1988. Dynamic manufacturing. New York: Free
Press.
Hise, R. T., O'Neal, L., Parasuraman, A., & McNeal, J. U. 1990. Marketing/R&D interaction in
new product development: Implications for new product success rates. Journal of Product Innovation Management, 7: 142-155.
Iansiti, M. 1992. Science-based product development: An empirical study of the mainframe
computer industry. Working paper, Harvard Business School, Cambridge, MA.
Iansiti, M. 1993. Real-world R&D: Jumping the product generation gap. Harvard Business
Review, 71(3): 138-147.
Imai, K., Ikujiro, N., & Takeuchi, H. 1985. Managing the new product development

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

process:

1995

Brown and Eisenhardt

377

How Japanese compaides learn and unlearn. In R. H. Hayes, K. Clark, & Lorenz (Eds.),
The uneasy alliance: Managing the productivity-technology dilemma: 337-375. Boston:
Harvard Business School Press.
Joyce, W. F. 1986. Matrix organizations:
Journal, 3: 536-561.

A social experiment.

Academy of Management

Katz, R. 1982. The effects of group longevity on project communication

Administrative Science Quarterly, 27: 81-104.

and performance.

Katz, R., & Allen, T. J. 1985. Project performance and the locus of influence
matrix. Academy of Management Journal, 28: 67-87.

in the R&D

Katz, R., & Tushman, M. L. 1981. An investigation into the managerial roles and career paths
of gatekeepers and project supervisors in a major R&D facility. R&D Management, 11:
103-110.
Keller, R. T. 1986. Predictors of the performance of project groups in R&D organizations.
Academy of Management Journal, 29: 715-726.
Kulvik, H. 1977. Factors underlying the success and failure of new products. (Report No. 29).
Helsinki, Finland: University of Technology.
Lieberman, M., & Montgomery, D. 1988. First-mover advantages.
Journal, 9: 41-58.

Strategic Management

Leonard-Barton, D. 1992. Core capabilities and core rigidities: A paradox in managing new
product development. Strategic Management Journal, 3: 111-125.
Mabert, V. A., Muth, J. F., & Schmenner, R. W. 1992. Collapsing new product development
times: Six case studies. Journal of Product Innovation Management, 9: 200-212.
Maidique, M. A., & Zirger, B. J. 1984. A study of success and failure in product innovation:
The case of the U.S. electronics industry. IEEE Transactions in Engineering Management, 4: 192-203.
Maidique, M. A., & Zirger, B. J. 1985. The new product learning cycle. Research Policy, 14:
299-313.
Miner, A. S., & Moorman, C. 1993. Organizational improvisation in new product development and introduction. Proposal for research, School of Business, University of Wisconsin, Madison.
Myers, S., & Marquis, D. G. 1969. Successful industrial innovations. (NSF 69-17). Washington, DC: National Science Foundation.
Nelson, R., & Winter, S. 1977. An evolutionary
England: Belknap Press.

theory of economic change.

Cambridge,

Nonaka, I. 1990. Redundant, overlapping organization: A Japanese approach to managing

the innovation process. California Management Review, 32(3): 27-38.
Quinn, J. B. 1985. Managing innovation: Controlled chaos. Harvard Business Review, 63(3):
73-84.
Rothwell, R. 1972. Factors for success in industrial innovations from project SAPPHO-A
comparative study of success and failure in industrial innovation. Brighton, Sussex,
England: S.P.R.U.
Rothwell, R., Freeman, C., Horsley, A., Jervis, V. T. P., Robertson, A., & Townsend, J. 1974.
SAPPHO updated-Project
Sappho phase II. Research Policy, 3: 258-291.
Rubenstein, A. H., Chakrabarti, A. K., O'Keefe, R. D., Souder, W. E., & Young, H. C. 1976.
Factors influencing success at the project level. Research Management, 16: 15-20.
Schoohoven, C. B., Eisenhardt, K. M., & Lyman, K. 1990. Speeding products to market: Waiting time to first product introduction in new firms. Administrative Science Quarterly, 35:
177-207.

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions

378

Academy of Management Review

April

Schlender, B. R. 1992. How Sony keeps the magic going. Fortune, February 24: 76-79, 82, 84.
Szakasits, G. G. 1974. The adoption of the SAPPHO method in the Hungarian electronics
industry. Research Policy, 3: 18-28.
Takeuchi, H., & Nonaka, I. 1986. The new new product development game. Harvard Business
Review, 64(1): 137-146.
Tyre, M., & Hauptman, 0. 1992. Effectiveness of organizational responses to technological
change in the production process. Organization Science, 3: 301-320.
Urabe, K., Child, J., & Kagono, T. 1988. Innovation and management: International comparisons. Berlin: de Gruyter.
Weick, K. E. 1993. The collapse of sensemaking in organizations: The Mann Gulch disaster.
Administrative Science Quarterly, 38: 628-652.
Womack, J. P., Jones, D. T., & Roos, D. 1990. The machine that changed the world. New York:
HarperCollins.
Zirger, B. J., & Maidique, M. 1990. A model of new product development: An empirical test.
Management Science, 36: 867-883.
Shona L. Brown received her Ph.D. from Stanford University. She is a postdoctoral
research associate in the School of Engineering at Stanford. Her research interests
include the formation of strategic capabilities, learning within and across firms, and
product innovation.
Kathleen M. Eisenhardt received her Ph.D from Stanford University. She is an associate professor of strategy and organization in the School of Engineering at Stanford. Her research interests center on the management of high-technology firms,
including evolutionary change, product innovation, strategic decision making, and
entrepreneurship.

This content downloaded from 217.20.216.203 on Sat, 07 Mar 2015 13:11:23 UTC
All use subject to JSTOR Terms and Conditions