Organizing for Radical Product Innovation

Rajesh K. ChandyUniversity of California, Los Angeles

Gerard J. TellisUniversity of Southern California

ISBM Report 104997

Institute for the Study of Business MarketsThe Pennsylvania State University

402 Business Administration BuildingUniversity Park, PA 16802-3004

(814) 863-2782 or (814) 863-0413 Fax

This publication is available in alternative media onrequest.The Pennsylvania State University is committed to the policy that all persons shallhave equal access to programs, facilities, admission, and employment without regardto personal characteristics not related to ability, performance, or qualifications asdetermined by University policy or by state or federal authorities. The PennsylvaniaState University does not discriminate against any person because of age, ancestry,color, disability or handicap, national origin, race, religious creed, sex, sexualorientation, or veteran status. Direct all inquiries regarding the nondiscriminationpolicy to the Affirmative Action Director, the Pennsylvania State University, 201Willard Building, University Park, PA 16802-2801; Tel. (814) 865-4700/V; (814) 865-1150/TTY.

U.Ed. BUS 98-0010

ORGANIZING FOR RADICAL PRODUCT INNOVATION

Rajesh K. Chandy and Gerard J. Tellis

June 9, 1997

Rajesh K. Chandy is a Visiting Assistant Professor at the Anderson Graduate School of Management, UCLA,110 Westwood Plaza, Los Angeles, CA 90095- 148 1 (Phone: 3 1 o-825-2502; Fax: 3 1 o-206-7422; Email:rchandy@anderson.ucla.edu). Gerard J. Tellis is the Jerry and Nancy Neely Professor of Marketing at theMarshall School of Business, University of Southern California, Los Angeles, CA 90089- 142 1 (Phone: 2 13-740-503 1; Fax: 213-740-7828; Email: tellis@rcf.usc.edu). This research is funded by grants from theInstitute for Study of Business Markets at Pennsylvania State University, the Marketing Science Institute, theCenter for International Business Education and Research at USC, and a dissertation support grant from theUSC Department of Marketing. The authors thank Bernie Jaworski, Jon Miller, and Allen Weiss for theirgenerous help and advice during the research. The authors also thank Kersi Antia, Teck Ho, Roy Howell,Shelby Hunt, Philip Bimbaum-More, Julia Leibeskind, Ofer Meilich, Om Narasimhan, Lars Pemer, AnilPeter, Jaideep Prabhu, and Pattana Thaivanich.

Report Summary

ORGANIZING FOR RADICAL PRODUCT INNOVATION

Radical innovations are critical to the long run success of a fI.rm. They can drastically change the competitive

position of Crms, weakening or destroying giants, encouraging new entrants, and promoting new market leaders. For

example, Xerox went from a nonentity to become a giant in the copier industry by its radical innovation in plain paper

copying. Similarly Kodak became a colossus of the photography industry with its radical innovation in celluloid roll

cameras. On the other hand, giants Cms such as Underwood and Remington lost their dominant market positions

because of their failure to cope with the next wave of technology.

Why are some fmns better at radical product innovation than others? The literature focuses on size as the

main cause of radical innovation. However, despite decades of research, authors are undecided whether it is small

firms, medium-sized firms or large fmns that are more likely to produce radical innova.ons. In contra& we propose

that certain organization factors are the key drivers of radical product innovation.

This study sought to test our hypotheses through a survey of 192 senior managers in the computer hardware,

telecommunications, and photonics industries. The results indicate that size by itselfhas no e&ct on radical product

innovation. The key factor that separates firms with strong radical product innovation records from others is their

willingness to cannibalize specialized investments. Specialized investments are investments whose value is strongly

tied to a particular product technology. Such investments could be in the form of physical assets, or less tangible .

organizational processes. A radical innovation has the potential to destroy the value of such investments. Innovating

organizations pursue the innovation even though doing so could involve cannibalizing specialized investments. Non-

innovators, however, are unwilling to cannibalize.

As firms build more specialized investments, they become less willing to cannibalize these investments.

Large, incumbent organizations in a product market possess many specialized investments in the older product

technology. Hence, other things being equal, they are less likely to be radical innovators.

Are incumbent firms with many specialized investments therefore doomed to die with their old technology?

Not necessarily. Results from the study indicate that such firms can overcome their natural unwillingness to

cannibalize by implementing three organizational features: 1) active internal markets, 2) influential product

champions, and 3) future market focus. Firms with active internal markets provide considerable autonomy to business

unit managers, and allow competition among their business units. Firms with influential product champions possess

organizational systems that promote the activities of champions. Firms with a future market focus emphasize future

customers and competitors over current customers and competitors in their market research and decision making.

Large firms that possess these features are highly willing to cannibalize, even though they own many specialized

investments. These results suggest a need to reconsider conventional wisdom on organizational size, cannibalization,

core competencies, and organizational synergy.

INTRODUCTION

Typewriters. Slide rules. Glass plate cameras. Each of these once dominant products are now

virtually extinct. They were swept away by radical innovations in the form of word processors, electronic

calculators, and cameras using celluloid film, respectively (Foster 1986; Utterback 1994). Interestingly, the

firms that introduced the new products in each of these categories were small new entrants into the market

(Utterback 1994). Yet they triggered the demise of the powerful firms that held sway in those markets.

The David versus Goliath scenario noted above is of great interest to managers for several reasons.

First, radical innovations have the capacity to destroy the fortunes of firms (Anderson and Tushman 1990;

Foster 1986; Tushman and Anderson 1986). Hard won customers quickly desert an incumbent firm when a

radical innovation provides better performance per dollar than the incumbent’s current products. Costly

investments made over the years may suddenly become useless because they cannot be applied to the new

generation of products. Skills that once led to the firm’s success may later cause it to become uncompetitive.

-cIc- : Second, radical product innovation can also be the source of competitive advantage to the innovator (e.g.,

Nason 1995; Wind and Mahajan 1997). The effects of radical innovation on a firm’s profits can be large,

positive, and long lasting (Geroski, Machin, and Van Reenen 1993). Both new and established firms can__

benefit from radical product innovation (e.g., Cooper 1994). Third, some authors argue that radical product

innovations are increasing in frequency (e.g., Foster 1986). If so, the need for radical product innovation

becomes particularly urgent.

In recent years, marketing researchers have increasingly focused on tactical decisions such as

promotion levels and package size (Bonoma 1988; 199 1). They have ceded the area of broad strategic issues,

such as radical product innovation, to researchers in other fields (Day 1996; Webster 1997). However, we

think this topic is very important to marketing research. First, radical product innovation is intrinsically

linked with a firm’s product-market strategy and can set the tone for the rest of the marketing program (e.g.,

Lehmann 1997; Gatignon and Xuereb 1997; Wind and Mahajan 1997). Second, in many high technology

firms we interviewed, marketing managers were directly confronted with these issues and took an active role

in related decisions.

The importance of radical innovation prompted ISBM to recently list “really new products” as one

of its priority research topics. In addition, in their introduction to the special issue of the Journd of

Marketing Research on innovation and new products, Wind and Mahajan (1997, p. 3) point to “the paucity of

research,” on breakthrough innovations and note that “the challenge is how to increase an organization’s

ability to develop breakthrough products.” What factors drive radical innovation?

Much research on the causes of organizational innovation is rooted in the seminal work of Joseph

Schumpeter (1912; 1942). His notion of “creative destruction,” whereby innovations destroy the market

positions of firms committed to old technology, first drew attention to the powerful effects of radical

innovation on the economy and the fortunes of individual firms. A key Schumpeterian hypothesis is that

large firms innovate more “intensively” than small firms (Acs and Audretsch 199 1; Scherer 1992, p. 1422).

- --. : This hypothesis has been the subject of extensive and continuing research. Over a hundred research articles

have studied the effects of size on innovation (Acs and Audretsch 1991; Baldwin and Scott 1987; Cohen and

Levin 1989). However, results of the research have been decidedly mixed (Scherer 1980; 1984).

Authors such as Galbraith (1952; 1968) and Ali (1994) build on Schumpeter’s basic arguments and

suggest that large firms have many advantages over small ones in their ability to produce radical innovations.

They note that large firms enjoy economies of scale in R&D, can spread risks widely (Arrow 1962), and have

greater access to financial resources. Other researchers argue that as firms become large, they become more

bureaucratic, slow to react, and less willing to take risks (e.g., see Mitchell and Singh 1993; Tornatzky and

Fleischer 1990). As a result, they are less Zikely to produce radical innovations than smaller firms that do not

possess these handicaps. These two views are contradictory. In addition, some authors suggest that the

relationship between innovative productivity and size is bell shaped (e.g., Ettlie and Rubenstein 1987).

Medium sized firms are best positioned for radical product innovation, since unlike small firms they possess

2

the critical mass for research, but do not suffer from the bureaucratic inertia of large firms. Conversely,

Pavitt (1990) argues for a U-shaped curve. He suggests that the “proportions of significant innovations made

by both large and small firms have been increasing at the expense of the medium sized firms in between” (p.

23). Perhaps medium sized firms have the liabilities of large and small firms, but few of their strengths.

Decades of research on the effect of size on radical innovation indicates that that factor may not

throw light on the causes of radical product innovation. Authors have not reached a consensus about the role

of size. Contradictions abound. Managerially useful generalizations are rare.

In its focus on size, the literature neglects attitudinal and organizational factors that may more

strongly influence radical innovation in firms (e.g., see Acs and Audretsch 199 1; Cohen 1995; Scherer 1992).

Ironically, Schumpeter (1942) himself noted that “mere size is neither necessary nor sufficient” for the

superior innovative performance of the large firm (Schumpeter 1942, p. 101). Anecdotal evidence suggests

that small, medium, and large firms can all be radical innovators. For example, the first celluloid roll camera

was introduced by Kodak, when it was a small firm. The first air-conditioner was introduced by the Buffalo

Forge company, a medium sized firm. The first microwave oven was introduced by Raytheon, a large firm.

This paper proposes an alternative explanation for radical innovation, based on organizational and

attitudinal factors. It suggests that the effect of size may be weak once these factors are accounted. We define

radical product innovation as the propensity of a firm to introduce new products that: 1) incorporate

substantially different technology from existing products, and 2) can fulfill key customer needs much better

than existing products. The next section explains our theory and hypotheses, which are based on insights

from our in-depth field interviews and the history of radical product innovations. The subsequent sections

describe the method and results. The final section discusses the implications of the results.

Beyond Size: What Really Drives Radical Product Innovation?

We argue that the key determinant that separates radical product innovators from others is the

willingness of the former to cannibalize their specialized investments. Willingness to cannibalize is critical

because firms that dominate markets are often reluctant to embrace or foster radical innovations in their

markets. Their reluctance derives from their established base of specialized investments with which they

serve such markets. Nevertheless, contrary forces can overcome this reluctance. In particular, the presence of

internal markets, influential product champions, and future market focus can overcome reluctance to

cannibalize specialized investments, and can motivate radical product innovations. While marketing

managers have little control over their firm’s size, they have considerable influence over these organizational

factors (e.g., Chandy and Stringfellow 1995; Kohli and Jaworski 1990; Menon, Bharadwaj, and Howell

1996). The subsequent sections explain each of these concepts of our argument.

RADICAL PRODUCT INNOVATIONSThe most popular product classification scheme in marketing, that of Booz, Allen & Hamilton

(1982), does not take into account a critical factor relevant to radical product innovation - technology. Yet

managers’ perceptions of alternate technologies play an important role in their decisions to pursue radical

product innovations (Henderson and Clark 1990). To overcome this limitation and account for the variety of

innovations, we develop a simple classification based on two key factors that drive product innovation -

technology (e.g., Capon and Glazer 1987; Ellul 1964; Veblen 1914) and markets (e.g., Urban, Weinberg, and

Hauser 1996; von Hippel 1986). The first factor determines the extent to which the technology involved in a

new product is different from current technologies. The second factor determines the extent to which the new

product fulfills key customer needs better than existing products, on a per dollar basis. Considering two *

levels (low and high) on each factor leads to four types of product innovations (see Table 1): incremental

innovations, market breakthroughs, technological breakthroughs and radical innovations.

Incremental innovations involve relatively minor changes in technology, and provide relatively low

incremental customer benefits per dollar. An example is Fuji’s “environmentally friendly” photographic

film which is enclosed in paper canisters instead of the plastic canisters that hold current film (Discount

Store News 199 1). Market breakthroughs are based on similar technology as existing products but provide

substantially higher customer benefits per dollar. An example is cable television with signal compression

systems that allow a substantially higher number of channels to be transmitted to customers using the

existing cable technology (e.g., Goldman 1995). Technological breakthroughs adopt a substantially

different technology from that used by existing products but do not provide superior customer benefits per

dollar. For example, electronic imaging uses a substantially different technology from celluloid film but does

not currently provide as good a resolution. In contrast to the above three, radical innovations involve

substantially new technology and provide substantially greater customer benefits per dollar relative to

existing products. An example is today’s celluloid roll film which displaced the glass plates used previously

(Coe 1989).

These different types of innovations are also related by an important dynamic, a series of S-curves of

technological innovation (see in Figure 1; Foster 1986; Sahal 1985; Utterback and Abernathy 1975). The S-

shape emerges because an existing technology (Tl) initially improves rapidly in benefits to consumers when

-0 C’ _.. the technology is new, and then more slowly as the technology matures. At some point in the history of that

technology, a new technology, T2, emerges. Initially, T2’s benefits are inferior to those of Tl, and the new

product based on T2 may be called a technological breakthrough (a). However, with research, T2 begins to

improve rapidly in consumer benefits and it ascends on its own S-curve. A point comes (b) when T2 passes

the existing technology Tl in benefits. At this point, the product based on T2 becomes a radical product

Iinnovation.

Faced with the threat of T2, supporters of T1 do make fresh efforts to improve the benefits from T1.

These efforts may yield some product improvements. These improvements may be a market breakthrough (c)

or just incremental innovations (d). However, the improvements cannot keep pace with the much faster rise

of T2 which ultimately replaces Tl as the dominant technology.

In time, T2 itself is threatened by a new technological breakthrough, and the cycle is repeated. The

history of technologies shows that innovation generally proceeds non-monotonously as a series of S-curves

(Foster 1986; Utterback and Abernathy 1975).

Although not every technological breakthrough becomes a radical product innovation, radical

product innovations can often render existing products obsolete. For this reason, a firm’s willingness to

cannibalize is a vital factor that drives the extent to which managers support radical product innovations.

Willingness to Cannibalize and Radical Product InnovationWillingness to cannibalize refers to the extent to which a firm is prepared to reduce the value of its

current specialized investments. Why does willingness to cannibalize drive radical product innovation? To

answer this question we need to understand managerial behavior with respect to the S-curves.

Note that while the new technology T2 is in its embryonic stage, an incumbent faces two options: it

can either choose to advance the current technology, T2 or switch to the new technology, Tl. A firm using

the current technology possesses many specialized investments whose value is strongly tied to that

technology (Ghemawat 1991 b; Leonard-Barton 1992; Teece 1987). Continuing with the current technology

would allow it to capitalize on its specialized investments. Switching to the new technology would require it_cI -- _..

to makes fresh investments in the new technology, potentially rendering the current investments obsolete. A

firm that is reluctant to cannibalize will continue with the earlier technology, TI. Initially, its efforts may__

yield some market breakthroughs as explained above. These achievements give the impression of success,

but they are the last gasp of a declining technology. While investing in this technology, the firm ignores the

new one, T2, and fails to create the environment for radical product innovations. Indeed, its adherence to the

old technology may lead to its ultimate demise (Foster 1986; Saxenian 1994; Utterback 1994).

In contrast, a firm willing to cannibalize will encourage employees to work on new technologies. It

would allocate adequate resources to radical product innovation. Such a nurturing atmosphere would be more

likely to lead to a breakthrough technology (Crawford 1994). Once a breakthrough technology has been

developed into a product, a firm that is willing to cannibalize would be likely to introduce it to the

marketplace before others. So, we expect:

PI: Firms that exhibit high willingness to cannibalize are more likely to be radical product6

innovators than other firms.

The relationship between willingness to cannibalize and radical product innovation may appear

intuitive, or the two concepts may appear indistinguishable. However, note that:

l Radical innovation is a behavioral variable, while willingness to cannibalize is an attitudinal one.l Other environmental or structural factors could also drive radical innovation. For example, the past

literature focused on a structural variable, size.l Radical innovations does not always result in cannibalization or obsolescence of related products. For

example, the television did not render the radio useless. Microwave ovens did not render conventionalovens useless. Movie videos did not render theatrical releases obsolete.

More often however, dominant firms are unwilling to cannibalize their specialized investments until

it is too late. Indeed, the history of innovation indicates that a new innovation (e.g. T2) is often introduced

not by the managers of Tl, but by some other party, sometimes even an outsider (e.g., Christensen and

Rosenbloom 1995; Foster 1986). In contrast, the managers of Tl tend to fight T2 by making further

investments in Tl. Why are intelligent managers reluctant to cannibalize their specialized investments? We

think the answer lies in understanding the concept of cannibalization.

A BROADER VIEW OF CANNIBALIZATIONTraditionally, marketers use the term cannibalization to refer to sales cannibalization, which is a

- reduction in sales of a firm’s older product due to sales of its new product (e.g., Mason and Mime 1994;

Carpenter and Hanssens 1994; Moor-thy and Png 1992). As such, marketers often consider cannibalization as

a problem to guard against. For example, Copulsky (1976) discusses how companies can avoid the “error of

cannibalism” in product development (p. 103). Kerin, Harvey and Rothe (1978) note how “cannibalism...

may be the unwelcome consequence of new product development” (p. 25).

In contrast, we view cannibalization as a desirable trait that can promote radical product innovation,

and the longterm success of the firm. We begin by defining the term broadly as cannibalization of

specialized investments, which includes actions by the firm that reduce the value of its investments in the

older product technology. Specialized investments may be in the form of assets or organizational routines.

Assets may be tangible, such as manufacturing or processing equipment, or intangibles such as

7

knowledge and expertise in the firm, (e.g., Glazer 1991; Henderson and Clark 1990; Tang 1988) or access to

distribution channels (e.g., Mitchell 1989; 1991). The success of a new technology can cause a firm’s assets

in a prior technology to become obsolete. For example, if Kodak’s current efforts in the area of electronic

imaging take off, they will substantially reduce the value of the organization’s prior assets in research,

production, processing, and marketing of celluloid film (Maremont 1995).

Organizational routines are established procedures used by a firm to carry out its day-to-day

activities. Such routines develop to enable a firm to bring a technology to market, but may be irrelevant for a

new technology. The following quote illustrates this point:

We had constructed over the years a management apparatus that was right for its times...It was rightfor the ‘70s...a growing handicap in the ‘~OS... and it would have been a ticket to the boneyard in the‘90s. So we got rid of it.... (General Electric CEO John Welch, quoted in Ghemawat 1991a, p. 152).

This broadened view of investments covering assets and routines is essential to explain willingness

to cannibalize.

-_) -- Specialized lnvestments and Willingness to CannibalizeAs explained above, firms that are successful with their current products possess many specialized

investments, frequently the fruit of years of labor of the firm’s senior managers. As such, managers develop a

strong professional and personal comm itment to them, which can lead to sub-optimal or even irrational

decisions. Boulding, Morgan, and Staelin (1997) note that such behavior is a major problem in managing

new products. Literature in at least three disciplines has addressed this issue. Psychologists have studied why

decision makers continue with past courses of action even when such actions may be irrational (e.g., Aronson

1968; Barnes 1984; Staw 1981). Economists and sociologists have likewise studied why firms are unable to

change their previous modes of action to respond to changed environments (e.g., Hannan and Freeman 1977;

Huff, Huff, and Thomas 1992; Rumelt 1995; Leonard-Barton 1992). This literature can be summarized in

terms of four biases that afflict firms with large investments in current technology: cognitive dissonance,

self-justification, external justification, and sunk cost fallacy.

Cognitive dissonance (Festinger 1957) and self-justification (Aronson 1968) theories imply that

individuals have a need to justify past actions to themselves, leading them to further increase commitment to

the actions. External justification suggests that individuals persist in actions to “save face” before others

(e.g., see Caldwell and O’Reilly 1982). The sunk cost fallacy suggests that individuals’ future actions are

influenced by expenditures that have already been made, even though they should not rationally be taken into

account (Kahneman and Tversky 1979).

Note that a firm may be affected by these biases even if top managers do not necessarily suffer from

them. Operating managers who could also suffer from these biases can manipulate information to justify

commitment to the current investments (see Barnes 1984). Moreover, as the cumulative investments in a

particular course of action increases, these biases become more pronounced (Brockner and Rubin 1985).

The S-curve of technological innovation can accentuate these biases. Manager’s commitment to a

current technology can be very high, if they were involved with the current technology and rode its own S-

- -_=e .L- curve to success as it replaced an older technology (see Figure 1). In that case they will be less open to a new

or future innovation. Overall, the above arguments suggest that:

P2: The more specialized investments a firm has in a current technology, the lower itswillingness to cannibalize those investments.

Do large firms generally have more specialized investments than small firms?

SIZE AND SPECIALIZED INVESTMENTS -We discuss the effects of firm size on the two types of specialized investments we noted earlier: assets and

organizational routines.

Size and Speciahed Assets: To stay competitive, firms within a product category have to make many

investments in assets. These assets are often specialized to a particular technology and cannot be applied to a

different technology. Large firms are more likely to accumulate large quantities of such assets because they have

greater access to resources, and have a need to sustain large volumes of sales relative to smaller firms (e.g., Fisher

and Temin 1973). Thus, large firms generally have more specialized assets than smaller firms.

Size and Organizational Routines: Greater size results in increased problems in communication and

coordination within the firm (Blau 1970; Blau and Schoenherr 197 1; Klatsky 1970). These factors, in turn, lead to

an increase in the size of the firm’s administrative component, the number of levels in its hierarchy, and rules and

procedures (see review by Kimberly 1976). Thus, greater size is generally associated with more organizational

routines. So we expect:

P3: The greater the size of the firm, the greater is its level of specialized investments in theform of assets and organizational routines.

Are large firms with many specialized investments doomed to die with existing technologies? While

the literature generally suggests this conclusion (e.g., see Henderson 1993), a careful look at history suggests

that this need not be the case. For example, Hewlett-Packard, a dominant firm in the laser printer market is

also on the forefront of the next generation of printers, the ink-jets. The key question then becomes, what

factors enable a firm to overcome commitment to current investments? We think that three such factors are

internal markets, strong product champions, and a focus on future markets.

Internal MarketsThe idea of internal markets has a fairly long history in the organizational literature (e.g., see

- For-rester 1965; Halal, Geranmayeh and Pourdehnad 1993). We consider a firm to have an active internal

market if it possesses high levels of: a) internal autonomy, and b) internal competition. Internal autonomy

refers to the extent of authority the business unit manager enjoys, relative to the corporate office, in making

decisions relating to his or her business unit (e.g., see Davis and Schul 1993; Hage and Aiken 1967; Gupta

1987). Internal competition refers to the extent of rivalry among business units in a corporation. In this

paper, an active internal market refers to an organizational structure with high levels of internal autonomy

and internal competition.

High levels of internal autonomy and competition hold the potential for dysfunctional conflict in the

firm. Moreover strong internal markets can lead to duplication of scarce resources. This is especially true in

the area of research and developments, where rival business units may have to duplicate research facilities in

10

order to develop alternate technologies in secrecy and competition with each other. For these reasons, authors

in recent years have urged firms to extract synergies across business units, and to promote cooperation rather

than competition among business units (e.g., Vizjak 1994; see discussion in Menon, Bharadwaj and Howell

1996).

However, we argue that active internal markets provide a critical benefit to firms (Menon,

Bharadwaj, and Howell 1996). They increase willingness to cannibalize, and thereby cause a firm to be more

radically innovative, for two reasons. First, in firms with active internal markets, the key decision maker for

product related decisions is the manager of each business unit. In considering the decision to switch to the

new product technology, the business unit manager takes into account the specialized assets owned by the

business unit. Hence, the decision to cannibalize is easier for the business unit manager than it is for top

management.

Second, when a firm is large, successful and dominant, it tends to belittle its competition. In the

perceived absence of external competition, a manager may not be eager to cannibalize his or her own assets

to be innovative. However, in the presence of internal markets, the business unit manager has to compete not

only with external competitors (i.e., other firms), but also other managers within the firm. If a particular

business unit manager does not support a radical new product that originates within the firm, a competing

manager may. Hence, the business unit manager has a strong incentive to aggressively seek and support

radical new products within the firm. Evidence from case studies suggests that internal markets of the form *

portrayed above may explain the strong innovative records of firms such as 3M and Johnson & Johnson (e.g.,

Peters and Waterman 1982). So we hypothesize:

P4: Firms with active internal markets exhibit greater willingness to cannibalize than otherfirms.

Product Champion InfluenceThe product champion has long been cited as a preeminent factor in producing radical product

innovations (e.g., Ettlie, Bridges, and O’Keefe 1984; Fischer et al. 1986; Rothwell 1977; Schon 1963). But

11

past research has focused on the traits of individual product champions (e.g., Chakrabarti 1974; Howell and

Higgins 1990), and their role in the innovation process (e.g., Ettlie, Bridges, and O’Keefe 1984; Rothwell

1977). The product champion is generally a person with drive, aggressiveness, political astuteness, technical

competence, and knowledge about the market (Fischer et al. 1986; Nakata and Sivakumar 1996). Most

accounts imply that the match of the product champion with the innovative project is generally due to

happenstance; the product champion is generally portrayed as being at the right place at the right time.

We take a different approach. We focus on the influence of product champions in a firm rather than

on their individual traits. Indeed, individuals who fit the personality traits of a product champion may be

equally likely to be present in radically innovating and non-innovating firms. However, the product

champions in innovating firms wield substantially higher influence. The reward systems (Jaworski 1988),

organizational legends (Deshpande and Webster 1989; Smircich 1983), and training programs at these firms

promote the influence of product champions. Top management actively supports the activities of product

- _‘ :. champions. Non-traditional ideas, which typically meet intense resistance within firms, stand a better chance

in firms with influential product champions who aggressively push such ideas (see discussion on new

product creativity in Moorman 1995; Moorman and Miner 1997). For example, Burgelman and Maidique-_

(1988) note the high levels of influence enjoyed by product champions at 3M. In such firms, key decision

makers become aware of the advantages of radical products, and the need to cannibalize past investments to

function effectively in the new product technology. Overall, the above arguments suggest that firms with

influential product champions would be willing to cannibalize even if their specialized investments are high,

since the opportunities and threats of potential new products are salient to them. Thus:

Pg: The greater the influence of product champions in a firm, the greater thewillingness to cannibalize.

Focus on Future MarketsTraditionally, marketing has always emphasized that firms be market oriented. Indeed, market

orientation has become the foundation of marketing strategy. However, some authors have asserted that a

12

strong market orientation “inhibits organizations from developing truly breakthrough innovations,” (see

discussion in Kohli and Jaworski 1990, p. 13; also see Bennett and Cooper 198 1; Kaldor 197 1; Tauber 1974).

Recent empirical research appears to validate this assertion. Gatignon and Xuereb (1997) find that a strong

customer orientation led to less radical innovation among firms in their survey sample. Christensen and his

colleagues (Bower and Christensen 1995; Christensen and Rosenbloom 1995) argue that dominant firms in

the rigid disk-drive, copier, tire, minicomputer, and mainframe computer markets stayed too close to their

customers, and consequently lost their market position in the new generation of products. This literature

might suggest that radically innovative firms need to “ignore their customers” (Martin 1995, p. 12 1). Which

of these two positions is right?

Our field interviews and analysis of case studies indicate that radically innovative firms pay close

attention to their markets. However, the difference lies in distinguishing between current andfiture markets.

The negative findings about market orientation discussed above may be of firms that stayed close to their

-c) -. :. current markets. The radically innovative firms we interviewed emphasizedfirture customers and

competitors relative to current customers and competitors. We term this emphasis a future market focus.

Such a focus can involve considering needs of customers and competitors who will constitute the future

market of the firm. It can also involve considering future needs of existing customers, or considering future

actions of current or potential competitors.

The availability heuristic explains the important influence of future market focus on willingness to *

cannibalize by (Tversky and Kahneman 1973). People judge the likelihood of events based on the ease with

which they imagine or recalled them (Barnes 1984; Folkes 1988). A future market focus causes decision

makers in a firm to become keenly aware of market related developments and their potential effects on the

firm (e.g., Hamel and Prahalad 1994). Therefore, a future market focus broadens the horizons of managers

and alerts them to new technologies, competitors and customers (see Deshpande, Farley, and Webster 1993;

Moorman 1995). It also alerts them to the potentially destructive results of radical innovations introduced by

13

competitors. Thus, managers are not overly committed to the past investments of the firm (e.g., Moorman

and Miner 1997), and are more willing to cannibalize them. So, we hypothesize:

P6: The greater the future market focus of a firm, the higher is its willingness to cannibalize.

Propositions P2, P4, Pg, and P6 suggest that specialized investments, internal markets, product

champion influence, and future market focus affect a fnm’s willingness to cannibalize. Proposition P1 argues

that willingness to cannibalize affects radical product innovation. Taken together, these propositions suggest

that willingness to cannibalize mediates the effect of the other organizational factors on radical product

innovation (see Baron and Kenny 1986).

SummaryRadical product innovation can catapult firms to positions of great success or drive once-mighty

firms to oblivion. A large body of literature has focused on firm size as a key organizational variable that

drives radical product innovation. But it ignores the rich variety of organizational forms that exist among

firms, large and small. This paper suggests an alternate model of radical product innovation that is based on

organizational and strategic factors. We propose that a key psychological factor that explains radical product

innovation is a firm’s willingness to cannibalize. Firms vary substantially in their willingness to cannibalize.

Specialized investments decrease willingness to cannibalize, and large firms may have more specialized

investments than small firms. However, internal markets, product champions, and future market focus are

organizational factors that increase willingness to cannibalize. Thus the latter three factors may be .

considered as counter forces to the effect of specialized investments.

Figure 2 illustrates these relationships. The dotted line linking size and radical product innovation

depicts the Schumpeterian hypothesis that size has a significant direct effect on radical product innovation

(e.g., Scherer 1992). The following section describes the method we used to test the propositions.

METHODOur model of radical product innovation focuses on intra-organizational factors. However, the

industry environment in the form of competitive intensity (Schumpeter 1942) or environmental turbulence

14

(Tushman and Anderson 1986) can also affect innovation. Since our focus is on organizational factors, we

controlled for the effects of the environment by restricting our sample to a few highly competitive and

turbulent high tech industries. We then collected primary data on our variables through a survey of key

managers. This section first addresses our sampling method. We then discuss the procedure we used to

administer the survey, followed by the measures. Next, we describe the model specification and analysis.

SAMPLE

We emphasized three criteria in selecting specific high tech industries for inclusion in this study.

First, the industries should exhibit frequent radical product innovations, thereby minimizing memory biases

in responses. Second, the industries selected should have sufficient variation in size, in order to test the

effects of size adequately. Third, the companies in these industries should differ sufficiently along key intra-

organizational variables. Based on information from published accounts (e.g., see Dodds 1995; Jelineck and

Schoonhoven 1990; Maremont 1995; Saxenian 1994; Utterback 1994) and interviews with managers, we

selected three high technology industries that met the above criteria: 1) computer hardware, 2) photonics, ande--T‘ .c-

3) telecommunication.

In building our database of key informants, we sought to verify that the each informant was

knowledgeable about the issues under study. We purchased a list of names and addresses of managers

involved in product strategy from CorpTech, a market research company specializing in information about

high technology companies. CorpTech collects data on the job responsibilities and titles of key managers at’

each company included in their list. This information provided us with an initial assurance that the managers

in our sample were indeed knowledgeable about relevant issues. Whenever possible, we also confirmed the

appropriateness of our key informants by phone calls made before the survey mail-out. A majority of the

managers in the sample were at the director or vice-president level. Each manager in the list was involved in

one of the three industries selected earlier.

Recall that size is a variable of considerable theoretical interest in this study. Hence we took care to

15

ensure that companies of all sizes were adequately represented in our sample. After controlling for industry

and job responsibility, we used a stratified random sampling procedure to select companies in our survey

sample (e.g., Fowler 1993). We divided the companies based on four size categories, those with: 1) less than

100 employees, 2) loo-499 employees, 3) 500-4999 employees, and 4) more than 5,000 employees. We

constructed the sample such that roughly a fourth of the sample frame consisted of companies in each of the

above four categories. Thus, about half of the companies in our sample frame fell under the Small Business

Administration’s definition of small firms as firms with less than 500 employees (see Acs and Audretsch

1988).

What is the most appropriate unit of analysis when measuring the effects of size on innovation?

Generally, the Schumpeterian literature has examined size at the corporate level. But an emergent stream of

literature has attempted to examine size at the business unit level (e.g., Scherer 1984; Scott 1984; Cohen, Levin,

and Mower-y 1987; Cohen and IUepper 1992). In a recent review of the Schumpeterian literature, Cohen (1995)

& -_T‘ : argues that the effects of size may best be measured at the business unit level. Many large firms operate

business units in several different markets and industries (e.g., see Chandler 1990). Further, the size of

business units in a firm may vary substantially. Some business units may be large (and behave like large

firms), while others may be small (and behave like small firms). The innovativeness of individual business

units can also vary substantially. Hence, we measured all variables at the SBU level]. But given the

inconsistency in the unit of analysis prior studies of firm size, we also measured size at the corporate level. If

a respondent belonged to a company with multiple SBUs, we asked him or her to focus on a specific SBU in

the company. An SBU is defined as a profit center with distinct products and markets.

SURVEY ADMINISTRATION

Surveys of senior managers at high-tech firms often involve low response rates. These managers are

generally very pressed for time, and their mail is often filtered by assistants to ensure that only important

information is forwarded to them. Further, they receive many survey requests, both from commercial and

16

non-commercial firms. Keeping these factors in mind, we used multiple approaches to increase the response

rate for the study.

First, whenever possible, we made phone calls to the respondents a few days before the actual

mailing, informing them that the survey was on its way. Second, we enclosed a crisp dollar bill in each

survey packet as a reward for the respondents. Third, we enclosed a highly personalized cover letter that

highlighted the managerial relevance of the study with the survey packet. The surveys and cover letter were

directly addressed to the respondent in question. Each cover letter also included a hand-written note to the

respondent. Fourth, we put considerable effort into producing a professional looking survey packet. Fifth, we

attempted to differentiate this research from commercial market research by highlighting the fact that the

study was dissertation research supported by universities and non-profit sponsors. Sixth, we sent two

reminder letters following the initial survey packet. The second reminder letter, a fresh copy of the

questionnaire, and a stamped return envelope were sent via priority mail.

a- --. M E A S U R E S

To develop the scales used to measure the constructs in the study, we used a multi-stage process

(e.g., Churchill 1979; DeVillis 1991). The output of the first stage was a preliminary pool of scale items. The

second stage provided the items used in the survey questionnaire. The third stage provided the final scale

items used to obtain the research results described later in this paper.

Stage 1: Preliminary Items .

We developed between lo- 16 items for each construct of interest using a combination of literature

search and in-depth interviews with knowledgeable managers. To measure radical product innovation, we

adapted items on innovativeness from Mahajan and Wind (1992), and Price and Mueller (1986). The

specialized investments items are adapted from Anderson and Weitz (1992). We operationalized firm size as

the number of full time employees in the firm, as in Acs and Audretsch (1991); Cohen, Levin and Mower-y

(1987) (see also Cohen 1995). We also included firm sales and number of full time employees in the

17

corporation as a whole, as alternate measures of size. Items modified from Aiken and Hage (1968); Davis

and Schul(l993); and Gupta (1987) provided a measure of internal autonomy. We obtained some future

market focus items by adapting items from Kohli, Jaworski, and Kumar’s (1993) MARKOR scale of market

orientation.

The face-to-face interviews involved 24 managers involved in new product development or product

strategy from 14 high technology firms nationwide. Each interview lasted between half an hour to one and a

half hours. The interviews were semi-structured; a set of prepared questions was used to guide the

interviews. We used the information from these interviews to develop further items for the constructs under

study. The interview data was also useful in refining some of the constructs, and in providing a rich

description of the process by which the independent variables in the study affected the dependent variables.

Stage 2: Survey Items

This stage involved refinement of the item pool from Stage 1 by successive reviews by academic

experts, and two rounds of pretests using managerial samples. A panel of marketing academics who were-c- . .

knowledgeable about the phenomena being studied reviewed the item pool, and rated each item on its

relevance to the construct as defined by us. In addition, the experts made comments on individual items as

they saw fit. Further, they evaluated the clarity and conciseness of each item. Based on the feedback from the

panel, we eliminated or modified individual items.

We then put together the revised item pool in the form of a questionnaire. Five managers in the -

industries which formed the empirical context for the study then completed the questionnaire. These

managers were actively involved in new product development activities in their firms. They completed the

questionnaire in the presence of one of the researchers, and made comments about individual items as they

worked through it. Later, we modified the questionnaire based on the comments of these managers.

Using a randomly selected sub-sample of managers from the actual survey sample frame, we then

conducted a second, fax based, pretest. The purpose of this pretest was twofold: 1) to provide a final

18

validation of the scale items before the large scale survey (e.g., Allreck and Settle 1985), and 2) to obtain an

approximate idea of the response rate that could be expected from the actual survey. As much as possible,

we tried to ensure that this pretest simulated the actual survey. However, we chose the fax as the mode of

administration in order to shorten the response turnaround time (Dickson and MacLachlan 1996). A reminder

letter was sent to persons who had not responded at the end of one week after the first fax.

Of the 25 questionnaires sent via fax, we received 6 completed responses (24% response rate) at the

end of two weeks after the first mailout. To increase the response rate in the final survey, we deleted some

items from the scales, thus shortening the questionnaire. After examining the responses to the pretest, we

chose not modify the remaining items substantially.

Stage 3: Scale Items

We conducted this stage of scale refinement after the responses to the final, large sample survey had

been collected and coded. We used an iterative process, involving calculation of coefficient a and factor

analysis. First, we examined the internal consistency of the sets of items by computing coefficient a values-7’ .

for each scale. We dropped items with low item to total correlations. We also deleted items which produced a

substantial or sudden drop in the item to total correlations (see Churchill 1979).

We then performed an exploratory factor analysis using the remaining items, and deleted all items

that did not load heavily on the primary factor. Next, we conducted a confirmatory factor analysis to test for

convergent and discriminant validity (Anderson and Gerbing 1988; and Bagozzi and Phillips 1982). For all.

the constructs, parameter coefficients of scale items were significant (~2.00), suggesting adequate

convergent validity. To assess discriminant validity, we performed a series of confirmatory factor analyses

using two-factor subsets of the variables. As Bagozzi and Phillips (1982) recommend, we ran each two-factor

confirmatory model twice - first by constraining the correlation between the constructs to 1 .O, and then by

freeing this parameter. We then computed the difference in overall model ~2 values between the constrained

and unconstrained models. For all the factor pairs considered, the ~2 values were significantly lower for the

19

- -.

unconstrained models, suggesting adequate discriminant validity among the measures?

We next recalculated coefficient a scores. The appendix provides the a values for each scale. The a

values range from .63 to .93, and all scales except one have reliabilities over .7 (see Nunnally 1978). We

finally constructed additive, equally weighted indices for each of the constructs.

The Internal Markets variable is a second order factor consisting of the internal autonomy and

internal competition variables (e.g., Gorsuch 1974). To compute the Radical Product Innovation index, we

standardized and summed the scale items. Before responding to these items, respondents first completed a

two page exercise with specific examples to clearly define the term and clarify its meaning to them.

To measure Specialized Investments, we first asked respondents to name and describe the most

radical product innovation introduced in their industry in the last three years3. We then asked them to

indicate the established technology that preceded this radical product innovation. Later, we presented them

with the specific questions that measured specialized investments. Hence, our scale of specialized

investments measures the level of specialized investments owned by the SBU in the established technology

that preceded the most radical innovation introduced in their industry.

The appendix provides descriptive statistics for the variables. Mean responses are generally close to

the center of each scale, providing some evidence to suggest that respondents did not inflate or deflate their

responses very much. Further, there exists sufficient variation in each scale, suggesting that the SBUs in our

*sample differ over the variables under study. Table 2 provides the pair-wise correlations for all of the

variables in the study. Few exogenous variables are correlated at high level, providing further evidence of

discriminant validity.

MODEL SPECIFICATIONOur theory and hypotheses suggest the model of radical product innovation shown in Figure 2, and

represented mathematically below:

Radical Product Innovation = p,, W illingness to Cannibalize + y 12 Size + EI (1)

20

Willingness to Cannibalize = PI1 Specialized Investments + y zf Internal Markets+ y 21 Influence of Product Champion + y 23 Future Market Focus

+ E2 (2)

Specialized Investments = yJI Size + E3 (3)

We tested this model using path analysis (Johnson and Wichem 1992). Note that path diagram in

Figure 2 forms a recursive system, i.e., there are only one-way causal flows in the system. Recursive models,

with the assumption of independent errors, fulfill the rank and order conditions for identification with no

additional exclusion restrictions4 (see Land 1973, p. 3 1 for a formal proof). Hence, we can obtain consistent

estimates of the parameters in each equation (see Maddala 1977).

To test alternate models, we compare the goodness of fit of the models after accounting for the

sample size and the number of restrictions imposed on the model (see Dillon and Goldstein 1984; Specht

1975). The goodness of fit statistic for recursive models with uncorrelated residuals is given by Q = (l-

MI)/( l-Ml), where A41 and M2 are the squared multiple correlations for each of the alternate models. The

values of Ml and A42 are obtained from the general formula:

M= 1 - (l-&2)( l-@)...( I-Rp2), (4)

where R$ is the ordinary squared multiple correlation coefficient of the ith equation in thesystem of equations that forms the recursive model, and

p is the number of equations in the model.

We can test the significance of the fit of the model by forming the test statistic IV below (Specht 1975). For

large samples, IV is ~2 distributed with d degrees of freedom.

IV=-(N-6)1@ (5)

where N is the sample size and d is the number of over-identifying restrictions.

Based on Proposition 1 we expect l31 I, the path coefficient that links Willingness to Cannibalize to Radical

Product Innovation in Figure 1, to be positive and significant. Proposition P2 suggests that p2I, the

coefficient between Specialized Investments and Willingness to Cannibalize will be negative and significant.

21

P3 through P6 lead us to expect that the coefficients 731, ~21, ~22, and ~23 will be positive and significant. The

Schumpeterian hypothesis suggests that ~12, the coefficient between Size and Radical Product Innovation

will be positive and significant.

RESULTSThis section covers the response rate from the survey and the results pertaining to each proposition

outlined earlier. It then examines if the effects of specialized investments on willingness to cannibalize are

overcome by internal markets, strong product champion roles, and future market focus.

RESPONSE RATE

Surveys were mailed out to 504 Strategic Business Units. Of these, 21 surveys were returned because

the address was wrong, the addressee had left the company, or the company had gone out of business. These

figures provide a potential sample size of 483 Strategic Business Units. At the end of four weeks after the

first mailing we had received 194 completed surveys 5. This figure provides us with an effective response rate

of 40.17%. This response rate is relatively high, when compared to other surveys using similar samples of

senior managers at high-tech firms. The high response rate may be explained by the many efforts we took to

increase cooperation from the respondents.

_ ESTIMATION RESULTS

This section describes the tests of each proposition discussed before. The results described in Table 4

correspond to equations 1,2, and 3 noted earlier. A formal test of industry effects indicates that the specific

industries (among the three we studied) that the respondents were affiliated with did not have statistically

significant effects on any of our key variables (ANOVA of industry effects provide p values greater than .lO

in all cases). Hence we do not include any industry-specific variables in the equations in Table 4.

How does Firm Size affect Radical Product Innovation? Is the shape of the relationship positive,

negative, U-shaped, or bell-shaped, as various segments of the literature suggest? Our results indicate that

Firm Size has no effect on Radical Product Innovation. As Table 4 indicates, the coefficient of Firm Size on

is small and non-significant (y12=-.02, p>.76)6. When the square of Firm Size is included as an additional

22

variable in the equation, its coefficient is also small and non significant (standardized coefficient value = .l,

px5).

An analysis of alternative model specifications reveals that an unrestricted model that includes Firm

Size does not provide a better fit with the data than a restricted model without Firm Size (W<X~J~Y= 5.99).

Overall, these results suggest that Firm Size has no significant effect on Radical Product Innovation. The

results contradict much of the Schumpeterian literature. Ironically, in the Schumpeterian literature, the

controversy is about the sign of the relationship, not whether the relationship exists at all.

If Not Size, Then What?

The results lend support to the proposition (PI) that the key predictor of an organization’s propensity

for Radical Product Innovation is its Willingness to Cannibalize. As Table 4 indicates, the coefficient of

Willingness to Cannibalize is positive and statistically significant at the ~~001 level. The standardized

coefficient for Willingness to Cannibalize is substantially larger than the effect of Firm Size on Radical

Product Innovation. A formal test comparing the l3 values of Willingness to Cannibalize and Firm Size

indicates that the difference between the coefficients is significantly different from zero (pc.001). These

results support our thesis that Willingness to Cannibalize provides a better indicator of Radical Product

Innovation than Firm Size.

Do Specialized Investments Cause Firms to be Less Willing to Cannibalize?

The coefficient of Specialized Investments is negative as hypothesized, but significant only at p<. 10.

Thus, proposition P2 is weakly supported. This result suggests that as specialized investments in a particular

technology increase, firms do become less willing to engage in actions that could reduce the value of these

investments. This finding, combined with the results on PI, supports the view that firms with large amounts

of specialized investments are likely to be less radically innovative, since they are generally less willing to

cannibalize than firms with few specialized investments.

Do Large Firms Have More Specialized Investments Than Smaller Firms?The effect of Firm Size on Specialized Investments is positive and significant (p<.Ol), as

23

hypothesized by proposition P3. Thus, larger firms do have more specialized investments than other firms.

Moreover, this result indicates that Firm Size does have an indirect effect on Radical Product Innovation

through its effect on Specialized Investments, even though it has no direct effects on Radical Product

11 x P21 x Pl 1 = .Ol),Innovation. But we find that this indirect effect is small (standardized indirect effect = y3

though statistically significant at the p<. 10 level (see Sobel 1982).

What Are the Key Drivers of Willingness to Cannibalize?

Propositions P4 through P6 focus on variables that help firms overcome the negative effects of

specialized investments. P4 proposes that the presence of active internal markets increases an organization’s

willingness to cannibalize. As Table 4 indicates, the coefficient of Internal Markets is positive and

statistically significant (pc.01). P5 suggests that firms with strong product champion roles would be more

willing to cannibalize than other firms. The results support this hypothesis (pc.01). P6 suggests that firms

that focus more on future markets relative to current markets would be more willing to cannibalize than

others. The results confirm the hypothesis (pc.01).- --_r‘ .C’

Overall, the results indicate that the above three organizational factors cause firms to be more

radically innovative by increasing their Willingness to Cannibalize. Willingness to Cannibalize has a strong

positive effect on Radical Product Innovation. Contrary to the Schumpeterian hypothesis, Firm Size has no

effect on Radical Product Innovation. All the propositions relating these three variables to Willingness to

Cannibalize are supported by the results.

T E S T O F M E D I A T I N G E F F E C T S

Does Willingness to Cannibalize mediate the effect of our four exogenous variables: Specialized

Investments, Internal Markets, Product Champion Influence, and Future Market Focus on Radical Product

Innovation? To formally test the mediating effects of Willingness to Cannibalize, we used the procedure

suggested by Baron and Kenny (1986).

These authors note three conditions that must be fulfilled to demonstrate mediating effects (see also

24

Robinson 1996). First, the exogenous variables and proposed mediator must each be significantly related to

the dependent variable (in our case, Radical Product Innovation), when considered separately. By examining

the correlation matrix in Table 2, and the regression equations la, 2a, 3a, and 4a in Table 5, we find that this

condition holds true for all our variables. Second, the exogenous variables should be significantly related to

the mediator. The pairwise correlations in Table 2 and equation 2 in Table 4 indicate that Willingness to

Cannibalize fulfills this condition. The third condition stipulates that the relationship between the exogenous

variables and the dependent variable should be weaker or non-significant when the proposed mediator is in

the regression equation relative to when the proposed mediator is not in the equation. Comparing the

coefficients of the exogenous variables in equations la, 2a, 3a, and 4a with their coefficients in equations 1 b,

2b, 3b, and 4b indicates that this third condition holds too.

Based on the above analysis, we conclude that the effects of Specialized Investments and Internal

Markets on Radical Product Innovation are fully mediated by Wilhngness to Cannibalize, since the direct

a---.-. : effects disappear when the mediator is controlled for. The effects of Product Champion Influence and Future

Market Focus on Radical Product Innovation are partially mediated by Willingness to Cannibalize since the

direct effects are reduced when the mediator is controlled for.

TESTS OF ALTERNATE MODELSNote from the results above that the effects of Product Champion Influence and Future Market Focus

do not completely disappear when Willingness to Cannibalize is taken into account. Thus, our exogenous

variables may also have statistically significant direct effects on Radical Product Innovation. A comparison

of alternate model specifications reveals that a model that includes the direct effects of all the organizational

variables in Figure 2 does provide a better fit with the data than a model that includes only the indirect

effects (W= 28.97 >xJdd~= 9.49). Results from the corresponding equation are portrayed in Table 6. But this

full model does not provide a better tit with the data than a reduced model that excludes the direct effects of

Specialized Investments and Internal Markets (W = 1.03 <x&f= 5.99).

25

These results indicate that most of the variables play the role we expected of them, except for one.

Namely, Willingness To Cannibalize is not a complete mediator of the effects of two of the four exogenous

variables on Radical Product Innovation: Product Champion and Future Market Orientation. We were unable

to determine if this is because of measurement error or because the exogenous variables have independent

direct effects on Radical Product Innovation.

ADDITIONAL ANALYSIS

Although the propositions suggested in this paper were generally supported by the results of the

survey, we found an unexpectedly weak empirical relationship between Specialized Investments and

Willingness to Cannibalize. One reason for these surprising results may be that Specialized Investments have

the strongest effects on incumbent firms; i.e., firms that actually participated in the technology preceding the

radical product innovation, and are an not important issue for non-incumbents. To explore this issue, we

conducted some additional analyses. From our sample, we isolated all SBUs that reported having participated

-- --.in the previous technology generation (n=130). We then re-estimated the path analysis described earlier.

Table 7 provides the results using the revised sample. The results remain generally the same as in the

previous sample, except for one key difference. The effect of Specialized Investments on Willingness to

Cannibalize is now large (p21= -. 28) and significant at p< .Ol . This result suggests that Specialized

Investments strongly decrease the Willingness to Cannibalize of incumbent firms.

The finding that Specialized Investments reduce Willingness to Cannibalize, while Internal Markets,

Product Champions Roles, and Future Market Focus increase it, prompts the following question: Can

organizational factors helpJirms compensate for the negative eflects of specialized investments? If so, these

organizational factors would be a means for managers to break the trap of specialized investments. If not,

managers of firms with many specialized investments may be bound to become victims of their past

investments.

To answer this question, we compared the standardized coefficient of Specialized Investments with

26

that of the three other organizational variables in Figure 2 (see Table 7). Standardization allows us to

compare different independent variables, because it puts all the variables on the “same” scale (Kim and

Ferree 198 1, p.23). The arithmetic sum of the three organizational variable coefficients is substantially

higher than that of Specialized Investments (i.e., l321 < ~21 + ~22 + ~23, ~~05) in both the full sample of

firms, and in the sample of incumbent firms. This result suggests that active internal markets, strong product

champion roles, and a focus on future markets can more than compensate for the negative effect of

specialized investments on willingness to cannibalize.

DISCUSSIONThis section first discusses some implications of the findings. It then points to some limitations of

the research, and highlights some promising areas for further research.

Many authors suggest that radical product innovations can have powerful effects on a firm’s

performance, and perhaps its very survival. Much current research has been focused around the

Schumpeterian view that size is an important determinant of radical product innovation. We find that the key

driver of radical product innovation is an organization’s willingness to cannibalize. The presence of

” investments specialized to an existing technology reduces a firm’s willingness to cannibalize. However,

internal markets, product champions, and a future market focus overcome the negative effects of specialized

investments. These results raise questions about some commonly held views about technology and product

market strategy.

Is Size An Important Driver of Radical Product Innovation?

The results on size raise doubts about a key hypothesis from Joseph Schumpeter’s classic and highly

cited work - Capitalism, Socialism, and Democracy (1942). In contrast to the Schumpeterian view, our

results suggest that firms of all sizes can be radical innovators, if they are willing to cannibalize investments

specialized to a particular technology. Hence the key issue to consider is not which is the largerfirm, but

which firm is more willing to cannibalize its own specialized investments. Ignoring this key variable is likely

27

to lead to contradictory, and perhaps dubious findings. More broadly, public policy makers have always been

wary of large firms. In the area of innovation this fear is unfounded. Large firms that are willing to

cannibalize will be innovative (e.g., Hewlett-Packard - see Wall Street Journal 1994). Those that are not

(e.g., typewriter firms such as Remington and Underwood - see Engler 1970), will be threatened or overrun

by firms that are more innovative. In either case, customers benefit with a stream of radically innovative

products.

Is Cannibalization an “Error?”

Marketers generally view cannibalization as an outcome of improper product positioning and

management. A casual review of widely used marketing principles textbooks suggests that this view is

commonly shared. For example, these texts refer to cannibalization as “a form of negative synergy,” an

“unfavorable consequence of . . . . new product introduction,” a “risk,” and a problem to “avoid.” Another

text categorically states that “if there is a great deal of cannibalization, sales shift from one product to the

new product, with little overaIl gain for thefirm” (Bearden, Ingram and LaForge 1995, p. 267, emphasis-0 X’ :.

added). Indeed, among the eleven marketing textbooks we examined, all except two textbooks that discuss

the phenomenon of cannibalization frame it in a negative context. Only Kerin and Peterson (1997) and

Dickson (1997) provide relatively balanced views of the topic. Even the research literature tends to view

cannibalization negatively (e.g., Copulsky 1976; Moor-thy 1984; Moorthy and Png 1992).

One reason cannibalization is generally viewed as a problem is that it is mostly considered in the

context of incrementally new products or line extensions. The results pertaining to willingness to cannibalize

in this paper indicate that in the context of radicalproduct innovation such views may be misleading or

harmful. Viewing cannibalization as something to be avoided may cause managers to persist in courses of

action even when it is no longer appropriate to do so. To our knowledge, this study is the first to empirically

validate the positive effects of willingness to cannibalize.

28

Does a Focus On Core Competence Guarantee Success?

Authors in recent years have urged f%-ms to focus on their “core competencies” (e.g., Day 1994;

Marino 1996; Prahalad and Hamel 1990; Rastogi 1995). Frequently, these competencies are portrayed in the

form of expertise in a particular technology. For example, Prahalad and Hamel (1990, pp. 82-84) note Sony

and JVC’s competencies in videotape, and Honda’s competencies in internal combustion engines. They

argue that firms should focus on building and exploiting these core competencies in related markets (for

example, Honda is able to parlay its competencies in engines into strong positions in automobiles,

motorbikes, lawn mowers, and power generators).

We suggest that there is a flip side to this argument. Today’s core competencies can easily become

the source of tomorrow’s “core rigidities” (Leonard-Barton 1990; see also Moor-man and Miner 1997).

Strictly following the prescriptions from the core competency literature would cause firms to focus on

specific technologies, and on exploiting these across product categories. They would build many investments

that are specialized to these technologies. The results of this study indicate that, other things being equal,__) -- L.

they would then become unwilling to cannibalize these investments, leading to their downfall when the next

radical product innovation strikes the industry. Consider Honda and JVC’s plight if they were to focus

exclusively on their core competencies in gasoline engines and videotape, and the technologies of the future

turn out to be electric motors and digital video devices!

Even though focus on core competencies may sometimes lead firms to develop synergies with other

competencies in the firm, these are most likely to result in incremental innovations or market breakthroughs.

Radical product innovations arise from a substantially different technology as indicated by our discussion of

S-curves. We suggest that firms look beyond their core competencies in specific technologies, and focus on

the evolving opportunities and threats in future markets.

Should Firms Always Stay Close to Their Customers?

Frequently, customers of the future have different needs and wants from current customers (e.g., von

Hippel 1986). Similarly, competitors of the future may be very different from current competitors. For

29

instance, Sony is typically not considered a photography company. However, it was the first to

commercialize electronic imaging technology (Television Digest 1988). If electronic imaging takes off,

traditional photography companies may well find Sony to be one of their biggest competitors (e.g., see

Bylinsky 1989).

The strong direct and indirect effects of future market orientation suggest that radically innovative

firms tend to focus on future customers and competitors who could enter their markets, more than on those

they currentZy deal with. This finding fits well with an emerging stream of literature in the technology

management area that indicates that listening closely to current customers may actually mislead firms and

lead them to their downfall (Bower and Christensen 1995; Christensen and Rosenbloom 1995). Given their

information processing limitations, a strong focus on current customers may cause managers to focus less on

future customers (see Barnes 1984).

Most current research in the area of market orientation does not differentiate between current and

future markets. Our results clearly point to the need to differentiate between these two types of markets.a---

Further research on methods to identify likely future customers and competitors would provide valuable

insights.

Should Firms Emphasize Synergies Among Business Units?

Lately, many authors have advocated a push for “synergies” among business units in firms (e.g.,

Vizjak 1994; see also Davis and Thomas 1993). Such a push, they argue, allows firms to exploit inter-

relationships among business units.

Recall that internal markets involve high degrees of autonomy for business units relative to the

corporate office. Active internal markets promote willingness to cannibalize because individual business

units only consider the specialized investments that they possess, which may be less than the specialized

investments owned by the firm as a whole. Further, the threat of internal competition cause managers to

become more willing to cannibalize than otherwise.

30

An emphasis on synergy results in a decrease in both internal autonomy and competition among

business units. In the product development context, the corporate office takes the lead in product strategy

formulation and implementation. If the firm has many specialized investments, the corporate office is likely

to emphasize incremental products that build on these investments, rather than radical product innovations,

that could potentially destroy them. While good arguments exist for exploiting economies of scale and scope

among business units, our study suggest that such efforts should not be made at the expense of reducing the

autonomy of business units. Business units should themselves judge when and with whom to cooperate or

compete.

L IMITATIONS AND FUTURE RESEARCH

The phenomenon of radical product innovation is too complex to be exhaustively described in any

one study. Resource constraints required us to focus our research on the drivers of radical product innovation

that we present here. Future research should identify other factors involved in radical product innovation. We

outline further study limitations and avenues for research below.- -- :

First, the current study is based on only three high technology industries. Research needs to

determine whether it holds equally well for low technology markets. Such a quest may not be futile. For

example, the demise of the typewriter, slide rule, and camera suggests that the today’s stable, low tech

industries can very quickly be transformed into the tomorrow’s fast moving high tech industries. However,

firms in relatively stable industries may have the luxury of longer response times to a technological

breakthrough.

Second, this paper did not analyze firm performance in terms of profits and market share. This issue

certainly warrants further study. Indeed, the relation between radical product innovation and firm

performance may not be straightforward. In the short run, such innovations appear to lead to higher profits

and market share for the firms that pioneer them (Geroski, Machin and Van Reenen 1993; Kleinschmidt and

Cooper 1991). Certainly, failure to embrace a radical product innovation has caused the downfall of many an

31

established firm (e.g., Foster 1986; Webster 1997). But few pioneers retain market share leadership over the

long term (Golder and Tellis 1993). Firms need to do more than simply introduce radical product innovations

and wait for the rewards to flow in (Tellis and Golder 1996).

Third, the scales used to measure some of the variables studied in this paper could be improved. In

particular, our scale for internal competition is not broad enough, and may contain some ambiguity. Our

scale for internal autonomy relates to general strategy rather than to specific decisions. This limitation may

account for the generally weaker results of internal markets. Future research should examine specific

decisions on which the SBU enjoys autonomy.

Fourth, some findings of this study suggest that many large, incumbent organizations are now able to

overcome inertia and commitment to prior technology that beset large firms in the past. In many older

industries, however, the dominant incumbents appear to have succumbed to such inertial pressures (e.g.,

Foster 1986; Utterback 1994). These differences raise the question: are the radically innovators of today

different from those of yore? One way to answer the question would be a historical analysis of radical

product innovations over time using archival sources.

Finally, firms do not live by radical innovation alone. It is only one of four types of innovations a

firm may undertake. Future research should examine how firms can effectively manage the mix of

incremental innovations, market breakthroughs, technological breakthroughs, and radical innovations (e.g.,

see Gatignon and Xuereb 1997; Tushman 1996).

CONCLUSIONWhat really causes some firms to be radically innovative over Iong periods of time, while many

others ossify and perish? We suggest that the answer lies in the extent to which the firms are prepared to give

up the old and embrace the new. Firms must break out of the “natural human trait” (Maddock, p. 367) that

propels them to use yesterday’s bag of tools to solve tomorrow’s problems. They must do so today, while

they stilI have options, not tomorrow, when they wilI have nothing Ieft save a useless bag of tools. They must

32

be willing to cannibalize before there is nothing of value left to cannibalize.

Clearly, cannibalization is a difficult and painful thing to do. It requires the firm to swim against the

tide of organizational inertia. Our results suggest that firms that are able to do so are specially designed for

the purpose. They possess unique organizational features that help them overcome the natural instinct to

preserve, rather than destroy past investments. Thus, radical innovation may not be just the province of the

small (or the large or the medium-sized). Firms of any size could play this game, if they have the appropriate

organizational setup.

33

FIGURE 1: A DYNAMIC VIEW OF PRODUCT INNOVATION

Benefits/$

- Actual path- - - * * - * Theoretical path

Market Breakthrough

New Technology (Tj

Time

FIGURE 2: HYPOTHESIZED MODEL OF RADICAL PRODUCTINNOVATION

FirmSize

WillingnessRadical

Pto Cannibalize

‘I w ProductInnovation

34

TABLE 1: TYPES OF PRODUCT INNOVATIONS

Customer Need Ful f i l lment /$

L o w H i g h

L o w I n c r e m e n t a l I n n o v a t i o n M a r k e t B r e a k t h r o u g h

N e w n e s s o fT e c h n o l o g y

H i g h T e c h n o l o g i c a l B r e a k t h r o u g h R a d i c a l I n n o v a t i o n

TABLE 2: CORRELATION MATRIX

VARIABLE Radical Willingness Specialized Internal Influence Future FirmProduct to Investments Markets of Product Market Size

-m --.=‘ .-. Innovation Cannibalize Champions Focus

Radical ProductInnovation

Willingness toCannibalize

SpecializedInvestments

Internal Markets

Influence of ProductChampion

Future MarketFocus

Firm Size

1.00

.54*** 1 .oo

-.12* -.17** 1.00

.13* .31*** -.08 1 .oo

.46*** .46*** -.02 .21*** 1 .oo

.49*** .51** -.16* .16** .43*** 1.00

-.04 -.04 .20*** .04 -.06 -.03 1 .oo

* p<.l; **p<.o5; ***p<.o1

35

TABLE 4: PATH COEFFICIENTS OF THE HYPOTHESZED MODEL(STANDARDIZED)

(Including all SBUs in the data set - n=192)IndeDendent Variable

Equation Dependent Willingness Specialized Internal Influence of Future FirmNo. Variable to Investments Markets Product Market Size R2

Cannibalize Chamnion Focus

1 RadicalProduct

Innovation

.54*** -.02 .29

2 Willingness - -.10* .20*** .27* * * .34*** .38to

Cannibalize

3 Specialized - .20*** .04Investments

TABLE 5: TEST OF MEDIATING EFFECTS(Using dataporn all SBUs in the data set - n=I92)

_.q T. : Dependent Variable: Radical Product InnovationIndependent Variable

Equation Willingness to Specialized Internal Influence of Future MarketNo. Cannibalize Investments Markets Product Champion Focus R2

la -.12* .02

lb .53*** -.02 .29

2a

2b .55***

3a

3b .41***

.13*

-.04

.46*** -22

.27***

.02

.29

4a .49*** .24

4b .39*** .29*** .35

Note: All coefficients are standardized values.* pc.1; **p<.o5; ***p<.o1

36

TABLE 6: TEST OF DIRECT EFFECTS(Including all SBUs in the data set - n=192)

Independent Variable

Dependent Willingness toVariable Cannibalize

SpecializedInvestments

InternalMarkets

Influenceof ProductChampion

FutureMarketFocus

FiITnSize R2

RadicalProduct

Innovation

.33*** -.02 -.06 .22*** .23*** .oo .39

TABLE 7: RESULTS OF EXTENDED ANALYSIS(Using datafiom incumbent SBUs only - n= 130)

IndeDendent Variable

-- --

Equation Dependent Willingness Specialized Internal Influence of F u t u r e F i r mNo. Variable to Investments Markets Product Market Size R2

Cannibalize Champion Focus

1 Radical .37*** .oo -.09 .ll .30*** -.Ol .39Product

Innovation

2 Willingness - -.2s*** .23*** .27*** .31*** SOto

Cannibalize

3 Specialized - .19** .04Investments

Note: All coefficients are standardized values.* pc.1; **p<.o5; ***p<.o1

3 7

REFERENCESAcs, Zoltan J. and David B. Audretsch (1991), Innovation and Technological Change: An International Comparison,Ann Arbor, MI: University of Michigan Press.

_B__ ---- (1988), “Innovation in Small and Large Firms: An Empirical Analysis,” American Economic Review, 74(Septdmber), 678-9 1.

Aiken, Michael and Jerald Hage (1968), “Organizational Independence and Intraorganizational Structure,” AmericanSociological Review, 33,912-30.

Ali, A. (1994), “Pioneering Versus Incremental Innovation: Review and Research Propositions,” Journal of ProductInnovation Management, 11 (January), 46-56.

Allreck, Pamela L., and Robert B. Settle (1985), The Survey Research Handbook, Homewood, IL: Irwin.

Anderson, Erin and Barton Weitz (1992), “The Use of Pledges to Build and Sustain Commitment in DistributionChannels,” Journal of Marketing Research, 29 (February), 18-34.

Anderson, James C. and David W. Gerbing (1988), “Structural Equation Modeling in Practice: A Review andRecommended Two-Step Approach,” Psychological Bulletin, 103 (May), 4 1 l-423.

Anderson, Philip and Michael L. Tushman (1990), “Technological Discontinuities and Dominant Designs: A CyclicalModel of Technological Change,” Administrative Science Quarter&, 35 (December), 604-633.

Aronson, E. (1968), The Social Animal, San Francisco: Freeman.

& TW .-- Arrow, Kenneth (1962), “Economic Welfare and the Allocation of Resources for Invention, in The Rate and Direction ofEconomic Activity, Universities - National Bureau Committee for Economic Research, Princeton: Princeton UniversityPress, 609-25.

Bagozzi, Richard P. and Lynn W. Phillips (1982), “Representing and Testing Organizational Theories: A Holistic. Construal,” Administrative Science Quarter+, 27 (September), 459-489.

Baldwin, William L. and John T. Scott (1987), Market Structure and Technological Change, New York: HarwoodAcademic Publishers.

Baron, R., and D. Kenny (1986), “The Moderator-Mediator Distinction in Social Psychological Research: Conceptual,Strategic, and Statistical Considerations,” Journal of Personality and Social Psychology, 5 1 (December), 1173-82.

Barnes, James H. (1984), “Cognitive Biases and Their Impact on Strategic Planning,” Strategic Management Journal, 5(April-June), 129-137.

Bearden, William O., Thomas N. Ingram, and Raymond W. LaForge (1995), Marketing: Principles and Perspectives,Chicago: Richard D. Irwin.

Bennett, Roger and Robert Cooper (198 l), “Beyond the Marketing Concept,” Business Horizons, 22 (June), 76-83.

Blau, Peter M. (1970), “A Formal Theory of Differentiation in Organizations, American Sociological Review, 35 (April),201-18.

____ and Richard M. Schoenherr (197 l), The Structure of Organizations, New York: Basic Books.

Bonoma, Thomas V. (1988), “Bridge Builders Needed on Strategy Street,” Marketing News, 22 (Nov 21), 8.3 8

----( 1991), “How Are We Doing? Not Too Bad Overall,” Marketing News, 25 (September 30), 17.

Booz, Allen & Hamilton ( 1982), New Product Management for the 1980 ‘s, NY: Booz, Allen & Hamilton.

Boulding, William, Ruskin Morgan, and Richard Staelin (1997), “Pulling the Plug to Stop the New Product Drain,”Journal of Marketing Research, 34 (February), 164-76.

Bower, Joseph L. and Clayton M. Christensen (1995), “Disruptive Technologies: Catching the Wave,” Harvard BusinessReview, 73 (January-February), 43-54.

Brockner, Joel and Jeffrey Z. Rubin (1985), Entrapment in Escalating Conflicts: A Social Psychological Analysis, NewYork: Springer-Verlag.

Burgelman, Robert A. and Modesto A. Maidique (1988), Strategic Management of Technology and Innovation,Homewood, Ill: Irwin.

Bylinsky, Gene (1989), ‘Where Japan Will Strike Next,” Fortune, 120,7 (September 25), 42-5 1.

Caldwell, D. F. and O’Reilly, C. A (1982), “Responses to Failures: The Effects of Choices and Responsibility onImpression Management,” Academy of Management Journal, 25 (March), 121-136.

Capon, Noel and Rashi Glazer (1987), “Marketing and Technology: A Strategic Coalignment,” Journal of Marketing,51 (July), 1-14.

Carpenter, Gregory S. and Dominique M. Hanssens (1994), “Market Expansion, Cannibalization, and InternationalAirline Pricing Strategy,” International Journal of Forecasting, 10 (September), 3 13-326.

& _- :Chandler, Alfred D. (1990), Scale and Scope: The Dynamics of Industrial Capitalism, Cambridge, MA: Belknap.

Chandy, Rajesh K. and Anne Stringfellow (1995), “Organizational Innovation for Effective New Product Development,”Marketing Science Institute Report No. 95- 120.

Chakrabarti, Alok K. (1974), “The Role of the Champion in Product Innovation,” California Management Review, 17(Winter), 58-62.

Christensen, Clayton M. and Richard S. Rosenbloom (1995), “Explaining the Attacker’s Advantage: TechnologicalParadigms, Organizational Dynamics, and the Value Network,” Research Policy, 24 (March), 233-58.

Churchill, Gilbert A. (1979), “A Paradigm for Developing Better Measures of Marketing Constructs,” Journal ofMarketing Research, 16 (February), 64-73.

Coe, Brian (1989), “The Roll Film Revolution,” in The Story of Photography, Colin Ford, ed. London: Century.

Cohen, Wesley (1995), “Empirical Studies of Innovative Activity,” in Handbook of the Economics of Innovation andTechnological Change, Paul Stoneman , ed. Cambridge, MA: Blackwell.

____ and Steven Klepper (1992), “The Anatomy of Industry R&D Distributions,” American Economic Review, 82(September), 773-88.

____ and Richard C. Levin (1989), “Empirical Studies of Innovation and Market Structure,” in Handbook of IndustrialOrganization, Volume II, R. Schmalensee and R. D. Willig , eds. New York: Elsevier Science Publishers.

3 9

____ ____ and D. C. Mower-y (1987), “Firm Size and R&D Intensity: A Re-examination,” Journal of IndustrialEcdnom& 35 (June), 543-63.

Cooper, R. (1994), Winning at New Products, Reading, MA: Addison-Wesley.

Copulsky, William (1976), “Cannibalism in the Marketplace,” Journal of Marketing, 40 (October), 103-5.

Crawford, C. Merle (1994), New Products Management, Burr Ridge, IL: Irwin

Davis, Peter S. and Patrick L. Schul(1993), “Addressing the Contingent Effects of Business Unit Strategic Orientationon Relationships Between Organizational Context and Business Unit Performance,” Journal of Business Research, 27(July), 183-20 1.

Davis, Rachel, and L. G. Thomas (1993), “Direct Estimation of Synergy: A New Approach to the Diversity-PerformanceDebate,” Management Science, 39 (November), 1334-46.

Day, George S. (1994), “The Capabilities of Market-Driven Organizations,” Journal of Marketing, 58 (October), 37-54.

----( 1996), “Using the Past as a Guide to the Future: Reflections on the History of the Journal of Marketing,” JournaZ OfMarketing, 60 (Winter), 14- 17.

Deshpande, Rohit, and Frederick E. Webster, Jr. (1989), “Organizational Culture and Marketing: Defining the ResearchAgenda,” Journal of Marketing, 53 (January), 3-5.

____ John U. Farley, and Frederick E. Webster (1993), “Corporate Culture, Customer Orientation, and Innovativeness inJapanese Firms: A Quadrad Analysis,” Journal of Marketing, 52 (January), 23-36.

-3 c. : DeVillis, Robert F. (199 1), Scale Development: Theory and Applications, Newbury Park: Sage Publications.

Dickson, Peter (1997), Marketing Management, Fort Worth, TX: Dryden Press

Dickson, John P. and Douglas L. MacLachlan (1996), “Fax Surveys: Return Patterns and Comparison with Mail* Surveys,” Journal of Marketing Research, 33 (February), 108- 13.

Dillman, Don A. (1978), Mail and Telephone Surveys: The Total Design Method, New York: Wiley.

Dillon, William and Matthew Goldstein (1984), Multivariate Analysis: Metho& and Applications, New York: Wiley.

Discount Store News (1991), “Photo Biz Bucks the Recession,” 30 (February 18), 29.

Dodds, Philip V. W. (1995), Digital Multimedia Cross-Industry Guide, Boston: Focal Press.

Ellul, Jacque (1964), The Technological Society, New York: Vintage Books.

Engler, G. N. (1970), “The Typewriter Industry: The Impact of a Significant Technological Innovation,” doctoraldissertation, UCLA.

Ettlie, J. E., W. P. Bridges, and R. D. O’Keefe (1984), “Organization Strategy and Structural Differences for RadicalVersus Incremental Innovation,” Management Science, 30 (June), 682-95.

-___ and A. H. Rubenstein (1987), “Firm Size and Product Innovation,” Journal of Product Innovation Management, 4,89-108.

Festinger, Leon (1957), A Theory of Cognitive Dissonance, Evanston, IL: Row Peterson.4 0

Fischer, William A., Willard Hamilton, Curtis P. McLaughlin, and Robert W. Zmud (1986), “The Elusive ProductChampion,” Research Management, 15 (May-June), 13- 16.

Fisher, F. M. and P. Temin (1973), “Returns to Scale in Research and Development: What Does the SchumpeterianHypothesis Imply?’ Journal of Political Economy, 8 1, 56-70.

Folkes, V. (1988), “The Availability Heuristic and Perceived Risk,” Journal of Consumer Research, 15 (June), 13-23.

Forrester, Jay W. (1965), “A New Corporate Design,” Sloan Management Review, 7 (Fall), 5-17.

Foster, R. (1986), Innovation: The Attacker’s Advantage, New York: Summit Books.

Fowler, Floyd J., Jr (1993), Survey Research Methods, Newbury Park, CA: Sage Publications.

Galbraith, J. K. (1952), American Capitalism, Houghton Mifflin.

---- (1968), The New Industrial State, Hamish Hamilton.

Gatignon, H. and J-M. Xuereb (1997), “Strategic Orientation of the Firm and New Product Performance,” Journal ofMarketing Research, 34 (February), 77-90.

Geroski, Paul, Steve Machin and John Van Reenen (1993), “The Profitability of Innovating Firms,” RAND Journal ofEconomics, 24 (Summer), 198-212.

Ghemawat, Pankaj (1991a), Commitment: The Dynamic of Strategy, New York: The Free Press.

-c) _T’ :’ ---- (199 1 b), “Market Incumbency and Technological Inertia,” Marketing Science, 10 (Spring), 16 l-7 1.

Glazer, Rashi (1991), “Marketing in an Information-Intensive Environment: Strategic Implications of Knowledge as anAsset” Journal of Marketing, 55 (October), I-20.

- Golder, Peter N. and Gerard J. Tellis (1993), “Pioneer Advantage: Marketing Logic or Marketing Legend?” Journal ofMarketing Research, 30 (May), 158-70.

Goldman, Kevin (1995), “Cable-TV Ads Fight Satellite-Dish Treatment,”The Wall Street Journal, (February 6), B8.

Gorsuch, Richard L. (1974), Factor Analysis, Philadelphia: W. B. Saunders Company.

Gupta, Anil K. (1987), “SBU Strategies, Corporate-SBU Relations, and SBU Effectiveness in Strategy Implementation,”Academy of Management Journal, 30 (September), 477-50 1.

Hage, Jerald and Michael Aiken (1967), “Program Change and Organizational Properties,” American Journal ofSociological, 72 (March) 503- 19.

Halal, William E., Ali Geranmayeh, and John Pourdehnad (1993), Internal Markets, New York: Wiley.

Hamel, Gary and C. K. Prahalad (1994), Competing for the Future, Boston: Harvard Business School Press.

Hannan, M. T. and J. Freeman (1977), “The Population Ecology of Organizations,” American Journal of Sociology, 82(5), 929-964.

Henderson, Rebecca and Kim B. Clark (1990), “Architectural Innovation: The Reconfiguration of Existing ProductTechnologies and the Failure of Established Firms,” Administrative Science Quarterly, 35 (January), 9-30. 41

---- (1993), “Underinvestment and Incompetence as Responses to Radical Innovation: Evidence from thePhotolithographic Alignment Equipment Industry,” RAND Journal of Economics, 24,2 (Summer), 248-7 1.

Howell, Jane M. And Christopher A. Higgins (1990), “Champions of Technological Innovation,” Administrative ScienceQuarterly, 3 5 (June) 3 17-4 1.

Huff, James O., Anne S. Huff, and Howard Thomas (1992), “Strategic Renewal and the Interaction of Cumulative Stressand Inertia,” Strategic Management Journal, 13 (Summer), 55-76.

Jaworski, Bernard J. (1988), “Toward a Theory of Marketing Control: Environmental Context, Control Types, andConsequences,” Journal of Marketing, 52 (July), 23-39.

Jelinek, Mariann, and Claudia Bird Schoonhoven (1990), The Innovation Marathon, Cambridge, MA: Basil Blackwell.

Johnson, R. A. and D. W. Wichem (1992), Applied Multivariate Statistical Analysis, Englewood Cliffs: Prentice Hall.

Kahneman, D. and A. Tversky (1979), “Prospect Theory: An Analysis of Decision Making Under Risk,” Econometrica,47,263-9 1.

Kaldor, A. G. (197 I), “Imbricative Marketing,” Journal of Marketing, 35 (April), 19-25.

Kerin, Roger A, Michael G. Harvey, and James T. Rothe (1978), “Cannibalism and New Product Development,”Business Horizons, (October), 25-3 1.

---- and Peterson, Robert A. (1997), Strategic Marketing Problems: Cases and Comments, Englewood Cliffs, NJ:Prentice Hall.

m-7‘Kim, Jae-On and G. Donald Ferree, Jr. (198 l), “Standardization in Causal Analysis,” in Linear Models in SocialResearch, Peter V. Marsden , ed. Beverly Hills: Sage Publications.

Kimberly, John (1976), “Organizational Size and the Structuralist Perspective: A Review, Critique, and Proposal,”- Administrative Science Quarterly, 2 1 (December), 57 l-97.

Klatsky, S. R. (1970), “Relationship of Organizational Size to Complexity and Coordination,” Administrative ScienceQuarterly, 15 (December), 428-38.

Kleinschmidt, Elko J. and Robert G. Cooper (1991), “The Impact of Product Innovativeness on Performance,” Journalof Product Innovation Management, 8 (December), 240-5 1.

Kohli, Ajay K., Bernard J. Jaworski, and Ajith Kumar (1993), “MARKOR: A Measure of Market Orientation,” Journalof Marketing Research, 30,4 (November), 467-78.

_-__ and ---- (1990), “Market Orientation: The Construct, Research Propositions, and Managerial Implications, Journal ofMarketing, 54 (21, 1- 18.

Land, Kenneth (1973), “Identification, Parameter Estimation, and Hypothesis Testing in Recursive SociologicalModels,” in Structural Equation Models in the Social Sciences, Arthur S. Goldberger and Otis D. Duncan , eds. NewYork: Seminar Press, Inc., 19-48.

Lehmann, Donald R. (1997), “A Different Game: Setting the Stage,” presentation at MSI’s Conference on Really NewProducts, Evolving Markets, and Responsive Organizations, May 5-6, Boston, MA.

4 2

Leonard-Barton, D. (1992), “Core Capabilities and Core Rigidities: A Paradox in Managing New ProductDevelopment,” Strategic A4anagement Journal, 13 (Summer), 11 l-25.

Maddala, G. S. (1977), Econometrics, New York: McGraw-Hill.

Maddock, Sir Ieuan (1982), “Why Industry Must Learn To Forget,” New Scientist, (February 1 l), 367-69 .

Mahajan, Vijay and Yoram Wind (1992), “New Product Models: Practice, Shortcomings, and Desired Improvements,”Journal of Product Innovation Management, 9 (June), 128-39.

Maremont, Mark (1995), “Kodak’s New Focus,” Business Week, 3409 (January 30), 62-69.

Marino, Kenneth E. (1996), “Developing Consensus on Firm Competencies and Capabilities,” Academy of ManagementExecutive, 10 (August), 40-5 1.

Martin, Justin (1995), “Ignore Your Customer,” Fortune, (May l), 123-26.

Mason, Charlotte H. and George R. Milne (1994), “An Approach for Identifying Cannibalization within Product LineExtensions and Multi-Brand Strategies,” Journal of Business Research, 3 1 (October-November), 163-70.

Menon, Anil, Sundar G. Bharadwaj, and Roy Howell (1996), “The Quality and Effectiveness of Marketing Strategy:Effects of Functional and Dysfunctional Conflict in Intraorganizational Relationships, Journal of the Academy ofMarketing Science, 24 (Fall), 299-3 13.

Mitchell, W. (1989), “Whether and When? Probability and Timing of Incumbents’ Entry into Emerging IndustrialSubfields,” Administrative Science Quarterly, 34 (June), 208-30.

---- (199 l), “Dual Clocks: Entry Order Influences on Industry Incumbent and Newcomer Market Share and SurvivalWhen Specialized Investments Retain their Value,” Strategic Management Journal, 12 (February), 85- 100.

____ and K. Singh (1993) “Death of the Lethargic: Effects of Expansion into New Technical Subfields on Performance ina Firm’s Base Business,” Organizational Science, 4, 152-80._

Moorthy, Sridhar K. (1984), “Market Segmentation, Self-Selection, and Product Line Design,” Marketing Science, 3(Fall), 288-307.

____ and I. P. L. Png (1992), “Market Segmentation, Cannibalization, and the Timing of Product Introductions,”Management Science, 38 (March) 345-59.

Moot-man, Christine (1995), “Organizational Market Information Processes: Cultural Antecedents and New ProductPerformance,” Journal of Marketing Research, 32 (August), 3 18-35.

____ and Anne Miner (1997), “The Impact of Organizational Memory on New Product Performance and Creativity,”Journal of Marketing Research, 34 (February), 91- 106.

Nakata, Cheryl and K. Sivakumar (1996), “National Culture and New Product Development: An Integrative Review,”Journal Of Marketing, 60 (January), 6 l-73

Nason, Stephen (1995), Learning Disabilities in Organizations, Unpublished Doctoral Dissertation, University ofSouthern California, Los Angeles, CA.

Olson, Eric M., Orville C. Walker, and Robert W. Ruekert (1995), “Organizing for Effective New ProductDevelopment: The Moderating Role of Product Innovativeness,” Journal of Marketing, 59 (January), 48-62.

43

Pavitt, K. (1990), “What We Know About the Strategic Management of Technology,” California Management Review,32 (Spring) 17-26.

Peters, Thomas J. and Robert H. Waterman (1982), In Search of Excellence: Lessonsfiom America’s Best RunCompanies, New York: Warner Books.

Prahalad, C. K. and Gary Hamel (1990), “The Core Competence of the Corporation,” Harvard Business Review, 68(May-June), 79-92.

Price, James L. and Charles W. Mueller (1986), Handbook of Organizational Measurement, Marshfield, MA: Pitman.

Rastogi, P. N. (1995), Management of Technology and Innovation: Competing Through Technological Excellence, NewDelhi: Sage.

Robinson, Sandra L. (1996) “Trust and Breach of the Psychological Contract,” Administrative Science Quarterly, 41(December), 574-99.

Rothwell, Roy (1977), “The Characteristics of Successful Innovators and Technically Progressive Firms,” R&DManagement, 7, 19 l-206.

Rumelt, Richard (1995), “Inertia and Transformation,” in Cynthia Montgomery, ed. Resource-Based and EvolutionaryTheories of the Firm: Towards a Synthesis, Boston: Kluwer Academic Press.

Sahal, Devendra (1985), “Technological Guideposts and Innovation Avenues,” Research Policy, 14 (June), 6 l-82.

Saxenian, AnnaLee (1994), Regional Advantage: Culture and Competition in Silicon Valley and Route 128, Cambridge,MA: Harvard University Press.

- C’ .L-Schon, Donald (1963), “Champions for Radical New Inventions,” Harvard Business Review, 4 1 (March-April), 77-86.

Scherer, F. M. (1992), “Schumpeter and Plausible Capitalism,” Journal of Economic Literature, 30 (Sept), 14 16-33.

- ---- (199 l), “Changing Perspectives on the Firm Size Problem,” in Innovation and Technological Change, Eds: ZoltanAcs and David Audretsch, NY: Harvester Wheatsheaf, 1991, pp. 24-38.

---- (1984), Innovation and Growth: Schumpeterian Perspectives, Cambridge, MA: MIT Press.

Schumpeter, J. A. (19 12), The Theory of Economic Development, Cambridge: Harvard University Press.

---- ( 1942), Capitalism, Socialism, and Democracy, NY: Harper.

Scott, John T. (1984), “Firm Versus Industry Variability in Technical Change,” in R&D Patents and Productivity, ZviGliriches , ed. Ann Arbor, MI: University of Michigan Press.

Smircich, Linda (1983), “Concepts of Culture and Organizational Analysis,” Administrative Science Quarterly, 28(September), 339-58.

Sobel, M. E. (1982), “Asymptotic Confidence Intervals For Indirect Effects in Structural Equations Models,” inSociological Methodology, S. Leinhart, ed. San Francisco, Jossey-Bass.

Specht, David (1975), “On the Evaluation of Causal Models,” Social Science Research, 4, 113-33.

Staw, Barry M. (198 l), “The Escalation of Commitment to a Course of Action,” Academy of Management Review, 6(October), 577-87.

4 4

Tang, Ming-Je (1988), “An Economic Perspective on Escalating Commitment,”Strategic Management Journal, 9(Summer), 79-92.

Tauber, Edward M. (1974), “How Marketing Discourages Major Innovation,”Business Horizons, 17 (June), 22-6.

Teece, David J. (1987), “Profiting from Technological Innovation: Implications for Integration, Collaboration,Licensing, and Public Policy,” in The Competitive Challenge: Strategies for Industrial Innovation and Renewal, David J.Teece, ed. New York: Harper & Row.

Television Digest (1988), “Consumer Mavica Debuts,” 28, 5 1 (December 19), 14.

Tellis, Gerard J. and Peter N. Golder (1996), “First to Market, First to Fail? Real Causes of Enduring MarketLeadership,” Sloan Management Review, 37 (Winter), 65-75.

Tomatzky, Louis G. and M. Fleischer (1990), The Process of Technological Innovation, Lexington, MA: Lexington.

Tushman, M. L. and P. Anderson (1986), “Technological Discontinuities and Organizational Environments,”Administrative Science Quarterly, 3 1 (September), 439-65

----and Charles A. O’Reilly (1996), “The Ambidextrous Organization: Managing Evolutionary and RevolutionaryChange,” California Business Review, 3 8 (Summer), l-23.

Tversky, Amos and Daniel Kahneman (1973), “Availability: A Heuristic for Judging Frequency and Probability,”Cognitive Psychology, 5 (September), 207-32.

Urban, Glen L., Bruce D. Weinberg, and John R. Hauser (1996), “Premarket Forecasting of Really-New Products,”-4 --_T- .-. Journal of Marketing, 60, 1 (January), 47-62.

Utterback, J. and W. J. Abernathy (1975), “A Dynamic Model of Process and Product Innovation,” Omega, 3,639-56.

---- (1994), Mastering the Dynamics of Innovation, Boston, MA: Harvard Business School Press.

Veblen, Thuston (19 14), The Instinct of Workmanship, New York: Norton.

Vizjak, Andrej (1994), “Exploiting Your Synergy Potential: Promoting Collaboration Between Business Unit,” LongRange Planning, 27 (February), 25-35.

von Hippel, Eric (1986), “Lead users: A Source of Novel Product Concepts,” Management Science, 32 (July), 791-807.

Wall Street Journal (1994), “Shoving Back: How H-P Used Tactics of the Japanese to Beat Them at Their Game,”September 8, Al.

Webster, Frederick E. (1997), “The Future Role of Marketing in the Organization,” in Reflections on the Futures ofMarketing, Donald R. Lehmann and Katherine E. Jocz, eds. Boston, MA: Marketing Science Institute.

Wind, Jerry and Vijay Mahajan (1997), “Issues and Opportunities in New Product Development: An Introduction to theSpecial Issue,” Journal of Marketing Research, 34 (February), l- 12.

45

APPENDIX: SCALE ITEMSAll Likert type items are measured using 7point (Strongly Agree - Strongly Disagree) scales. Items marked with anasterisk are reverse coded.

RADICAL PRODUCT INNOVATION (a = .72)

Likert Type Scale (Mean = 8.68; S.D. =3.09):This SBU rarely introduces products that are radically different from existing products*.This SBU lags behind in introducing radical product innovations*.

Numeric Scale (Mean = I. 83; S. D. = 1.68):Please indicate the number of radical product innovations introduced by your SBU in the last three years.

Ordinal Scale (Mean = 3.48; S.D.=I.$l):% of total sales from Radical Product Innovations introduced in the last three years.(Under 1%; l%-5%; 5%-10%; lo%-20%; 20%-30%; Over 30%)

WILLINGNESS TO CANNIBALIZE (a = .85)(Mean =34.99; S. D. = 8. I5)

This SBU can easily change its organizational scheme to lit the needs of a new product.

This SBU supports projects even if they could potentially take away sales from existing products.

We find it difficult to change established procedures to cater to the needs of a new product*.

-3 c- :. This SBU easily replaces one set of abilities with a different set of abilities to adopt a new technology.

We tend to oppose new technologies that cause our manufacturing facilities to become obsolete*.

We are very willing to sacrifice sales of existing products in order to improve sales of our new products.

We can easily change the manner in which we carry out tasks to fit the needs of a new product.

This SBU will not aggressively pursue a new technology that causes existing investments to lose value*.

SPECIALIZED INVESTMENTS (a = .93)(Mean =26.77; S.D.=15.21)

Switching to the new technology involves losing a lot of the investment in the established technology.

It is difficult for us to apply investments made in the established technology to the new technology.

Many of our manufacturing skills cannot be applied to the new technology.

Our current marketing abilities are not very useful in marketing products based on the new technology.

We have a substantial investment in facilities dedicated to the established technology.

We possess a large amount of assets that will have little value outside the established technology.

Much of our technical expertise cannot be applied to the new technology.

46

Many of our current operating procedures cannot be applied in the new technology.

To be successful in the new technology, we need to substantially change the manner in which we carry outmany of our tasks.

We have to significantly reinvent this SBU to operate successfully in the new technology.

We have to retrain or lay off employees to operate successfully in the new technology.

INTERNAL AUTONOMY (cc = 88)(Mean = I 7.66; S. D. =6.3 7)

The corporate office has much more influence than the SBU managers in formulating the SBU’s productstrategy*.

Few strategic actions can be taken in the SBU until the corporate office approves the decisions*.

Even small product strategy issues have to be referred to someone in the corporate office for a final answer*.

All new product decisions need the corporate office’s approval*.

INTERNAL COMPETITION (a = .63)(Mean =8.44; S. D. =2.39)

The SBU is free to enter markets served by other SBUs in the firm.

When making strategic decisions, the SBU has to make sure it does not take away customers from other SBUs*.

PRODUCT CHAMPION INFLUENCE (a = .83)(Mean =25. I4; S. D. =5.79)

Product champions play an important role in this SBU.

Senior managers in this SBU strongly support champions of radical product ideas.

Activities of product champions have a clear impact on product development in this SBU.

Top managers in this SBU are frequently the most ardent champions of radical product ideas.

Product champions wield considerable clout in this SBU

FUTURE MARKET FOCUS (a = .72)(Mean =16.91; S.D.=4.35)

The SBU gives more emphasis to customers of the future, relative to current customers.

Market research efforts in the SBU are aimed at obtaining information about customers’ needs in the future,relative to their current needs.

We are slow to detect fundamental shifts in our industry (e.g., competition, technology, regulation)*.

The SBU is oriented more toward the future than the present.

4 7

---

1 For ease of exposition, we use the terms “firm” and “SBU” interchangeably in the remainder of this paper. We use theterm “company” to refer to the corporation as a whole.2 The differences exceeded the critical value (~2 Idf > 3.84) in all cases.3 If the respondents felt that no radical product innovations had been introduced in their industry during the last threeyears, they were asked to note this by checking a box, and then describe the product that came closest to being a radicalproduct innovation. Given the nature of the industries we studied, relatively few persons suggested that no radically newproducts had been introduced in their industry.4 An analysis of the residuals from each of the equations estimated using the data in this study reveals that the residualsare independent of each other. Thus, the independent errors assumption is appropriate in this context.5 Two outliers were deleted from the sample, providing us with a final sample size of 192 firms.6 We also regressed firm size alone on radical product innovation. Its coefficient is also small @=-.04) and notsignificantly different from zero (p>S). Analyses using alternate operationalizations of the size variable - Corporate Size(number of employees in the firm as a whole), SBU Sales (dollar sales of the SBU in the last year), and natural logarithms ofCorporate Size and SBU Sales, respectively - yielded similar results. Unfortunately, we did not have data on CorporateRevenue for the set of firms we studied; we were therefore unable to study the effects of this variable on Radical ProductInnovation.

48