Jump to ContentJump to Main Navigation
Recovering from SuccessInnovation and Technology Management in Japan$

D. Hugh Whittaker and Robert E. Cole

Print publication date: 2006

Print ISBN-13: 9780199297320

Published to Oxford Scholarship Online: September 2007

DOI: 10.1093/acprof:oso/9780199297320.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 19 October 2020

The open innovation model: Implications for innovation in Japan

The open innovation model: Implications for innovation in Japan

(p.129) 7 The open innovation model: Implications for innovation in Japan
Recovering from Success

Henry W. Chesbrough

Oxford University Press

Abstract and Keywords

This chapter explores the implications of the transition from closed to open innovation over the past two decades, which Japanese companies have been relatively late to address. Examples of the fate of similar US companies are given. On the one hand, the wide dispersion of critical knowledge for innovation is ignored at great risk. On the other, alternative paths to market for internal ideas may be ignored at great loss. Strategies for dealing with ‘false negatives’ are explored — which at first look unpromising but later turn out to be valuable — including the time-honoured Japanese practice of spin-offs.

Keywords:   Japan, R&D, knowledge, paths to market, false negatives, spin-offs


Industrial innovation is becoming more open, requiring changes in how firms manage innovation. External sources of knowledge become more prominent, while external channels to market also offer greater promise for utilizing internal knowledge. This elevates the importance of the evaluation of early stage technology projects, which often involve significant technical uncertainty and significant market uncertainty. Companies need to ‘play poker’ as well as chess, in such circumstances. Measurement errors (false positives, false negatives) are likely to arise from judgements about the commercial potential of early stage projects. Most companies' policies consciously limit ‘false positives’ in assessing a project's commercial potential, but few companies take steps to manage the risk of ‘false negatives’. New metrics may help a firm focus more upon external sources of innovation to enhance its business model, and enable the firm to salvage value from false negatives that otherwise would be lost.

Open innovation should not be understood to mean the lack of any internal mechanism to capture value from innovation. It may be better viewed as an open-closed process, where the openness helps to create value throughout the value chain of the company, its suppliers, its customers, and the ultimate end users. Once value is created, which means that the technology has been embraced by the players in the value chain, the closed-ness helps to claim a portion of that value.

Openness also has important implications for the boundary of the firm, particularly in Japan. Technologies should move more fluidly between organizations and, at times, new organizations may be the most effective means of pursuing a new technology opportunity. Spin-offs stand as a mechanism to (p.130) manage the risk of false negatives and explore new business models to create and capture value in the Japanese context, though much remains unknown about their effectiveness in practice in doing so.

The shifting process of industrial innovation

Not long ago, internal research and development was viewed as a strategic asset, and even a barrier to competitive entry in many industries. Only large companies with significant resources and long-term research programmes could compete. Research-based companies like DuPont, Merck, IBM, GE, and AT&T did the most research in their respective industries. And they earned most of the profits as well. Rivals who sought to unseat these firms had to ante up their own resources, and create their own labs, if they were to have any chance against these leaders. There were significant economies of scale in R&D, and the biggest companies generally developed the best technologies.

Vertical integration was the dominant business logic of the last century. Underlying the logic was the belief that valuable knowledge was fundamentally scarce. As a result, companies sought to develop a knowledge advantage that others could not match. This corporate world view brought with it a number of working assumptions:

  • The company which gets an innovation to market first, will win.

  • If you create the most, and the best, ideas in the industry, you will win.

  • The smart people in our field work for us: Companies competed for the best and the brightest graduates and offered these recruits the best salaries and equipment.

  • If we discover it ourselves, we will get it to market first: Internal R&D was seen as a barrier against smaller competitors.

  • To profit from R&D, we must discover it, develop it, and ship it ourselves: The rise of companies like DuPont, General Electric, General Motors, IBM, Xerox, Merck, and Procter & Gamble were all fuelled by sustained investment in internal R&D. A by-product of this emphasis was the ‘not invented here’ syndrome, where companies rejected any technology that had come from outside.

  • We should control our intellectual property so that our competitors don't profit from our ideas.

For most of the twentieth century, this closed innovation model worked, and worked well. We can thank it for a whole range of inventions and developments. It enabled Thomas Edison to invent the phonograph and the electric light bulb among other things. In the chemicals industry, companies like DuPont established central research labs to identify and commercialize an amazing variety of new products such as the synthetic fibres nylon, Kevlar, (p.131) and Lycra. And Bell Labs researchers discovered groundbreaking physical phenomena and harnessed those discoveries to create a host of revolutionary products, including transistors and lasers, winning a number of Nobel prizes for their research along the way.

However, this model of closed innovation ran into severe problems towards the end of the century. By way of illustration, compare Lucent, which inherited the lion's share of Bell Laboratories after the breakup of AT&T, with Cisco. Bell Labs was perhaps the premier industrial research organization of the last century. Within the old model of innovation, this heritage should have been a decisive strategic weapon for Lucent in the telecommunications equipment market. Yet, Cisco, without the deep internal R&D capabilities of Bell Labs, has consistently managed to stay abreast of Lucent, occasionally beating it to market. Today, Cisco dominates the telecommunications equipment market, while Lucent (which nearly went bankrupt at one point) is a distant follower.

How can Cisco's relative success vs. Lucent be explained? The two organizations were simply not innovating in the same manner. Lucent was a classic example of a closed innovator, devoting enormous resources to exploring the world of new materials and state of the art components and systems, seeking fundamental discoveries that could fuel future generations of products and services.

In contrast, Cisco deployed a very different, and far more open, strategy. Whatever technology the company needed, it acquired from the outside, usually by partnering or investing in promising start-ups (some, ironically, founded by ex-Lucent veterans). In this way, Cisco was able to keep up with the R&D output of perhaps the finest industrial R&D organization in the world, and without doing much internal research of its own.

The story of Lucent and Cisco is hardly an isolated instance. IBM's research prowess in computing provided little protection against Intel and Microsoft in the personal computer business. Similarly, Motorola, Siemens, and other industrial titans watched helplessly as Nokia catapulted itself to the forefront of wireless telephony in just 20 years, building on its industrial experience from earlier decades in the low-tech industries of wood pulp and rubber boots. And pharmaceutical giants like Merck and Pfizer have watched as a number of upstarts, including Genentech, Amgen, and Genzyme, have parlayed the research discoveries of others to become major players in the biotechnology industry.

These days, the former leading industrial enterprises are finding remarkably strong competition from many newer companies. These newcomers conducted little or no basic research on their own. They have been very innovative, but they have innovated with the research discoveries of others. And there is a legion of other, even newer, companies waiting to supplant these firms, if an opportunity should arise.

To make matters worse, some companies that made significant long-term investments in research found that some of the resulting output, however (p.132) brilliant, wasn't useful for them. They found ways to gracefully exit from further funding of these projects, and moved on to more promising work. Then, to their amazement, some of those abandoned projects later turned into very valuable companies. This was the experience of the Xerox Corporation, for example, with its Palo Alto Research Center. Numerous valuable computer hardware and software innovations were developed at PARC, but few of them made any money for Xerox and its shareholders.

The shift to open innovation

Over the past two decades, the management of innovation has fundamentally changed. It is still true that no company can grow and prosper without new ideas. It is also clear that the changing needs of customers, increasing competitive pressure, and the evolving abilities of suppliers necessitate continual creative thinking for a company to stay ahead of the pack.

The challenge is that the distribution of this critical knowledge has shifted from being locked up in the corporate laboratories of the biggest firms in the industry, to being dispersed among for-profit firms of all sizes, and non-profit organizations like universities and research institutes. This has important implications for how every company thinks about growth and innovation. There are many fewer economies of scale in R&D today.

The reasons behind this basic change are many and varied. In the United States one factor was the success of the GI Bill which increased college numbers in the postwar years. Other factors include the rise in the amount and quality of university research, the increased mobility of skilled personnel between companies, and the growth in venture capital and private equity that created a pool of risk capital to fund the development of new ventures. In Japan, the accession to the World Trade Organization, the liberalization of import markets, the recent acceleration in the development of legal systems to promote cooperation among academia, industry (including the conversion of national universities and research institutes into Independent Administrative Institutions), and the government personnel and institutions, the movement toward modularized product architecture and production, and the rise of the Chinese economy have all changed the innovation landscape for Japanese firms. See Probert in this volume for a description of how the landscape has changed for the Japanese pharmaceutical industry.

The result in both countries has been an erosion of the carefully created and nurtured knowledge monopolies inside leading industrial corporations. Instead of being retained within corporate walls, knowledge streamed out of centralized R&D to suppliers, customers, start-ups, and spin-offs. A new generation of companies arose, which innovated with ideas brought in from outside. Of course, they added to this knowledge base, and crafted innovative (p.133) business models around that knowledge. But they did little internal R&D on their own, relying instead on licensing, acquiring, and copying external technology. While many large companies in Japan remain successful within their industries, few have escaped the pressures of stronger foreign competition within Japan, combined with tremendous competition to establish market leadership in emerging economies like China and India.

We have moved from closed innovation to a new logic of innovation: open innovation. This new logic builds upon the recognition that useful knowledge is widely distributed across society, in organizations of all sizes and purposes, including nonprofits, universities, and government entities. Rather than reinvent the wheel, the new logic employs the wheel to move forward faster.

What accounts for the apparent decline in the innovation capabilities of so many leading companies? We are witnessing a ‘paradigm shift’ in how companies commercialize knowledge from‘closed innovation’ to ‘open innovation’. Closed innovation is a view that says successful innovation requires control. Companies must generate their own ideas, and then develop them, build them, market them, distribute them, service them, finance them, and support them on their own. This paradigm counsels firms to be strongly self-reliant, because one cannot be sure of the quality, availability, and capability of others' ideas.

Increasingly, however, the closed innovation approach to innovation is no longer sustainable. A paradigm of open innovation is emerging in its place (Chesbrough 2003). The open innovation paradigm assumes that firms canand should use external ideas as well as internal ideas, and internal and external paths to market, as they look to advance their technology. Open innovation assumes that internal ideas can also be taken to market through external channels, outside the current businesses of the firm, to generate additional value.

This transition to open innovation will not be easy or painless. Clair Brown's chapter in this volume explores the human resource challenges involved in implementing a more open innovation approach. People must be recruited from new places, given different assignments, provided different reward systems and new job definitions and roles. There are technical issues as well. Admitting external sources of technology into a company's innovation process increases the number of possible sources of innovation. This greater complexity places even greater burdens upon the ability to evaluate early stage technologies. It suggests that innovators must address a key concern: measurement error.

The problem of technical and market uncertainty: Measurement error

Successful commercialization of a new technology involves managing both technical and market uncertainty. The capability and performance of a fledgling technology often are poorly understood. This technical uncertainty is (p.134) compounded by market uncertainty, when early stage technology projects also address an uncertain market. How a technology might be used by customers, and what benefits it might provide to them, are far from clear. Measurement errors (both false positives and false negatives) are inevitable. Yet companies evaluate early stage R&D projects with processes that implicitly assume that the Type II (false negative) error rate is nearly zero. This is because they employ no processes to re-examine earlier negative decisions to discontinue the technology.

Evaluating the commercial potential of a new technology is less subject to measurement error when it addresses a current market with a known set of customers. Xerox had little apparent difficulty dealing with even high degrees of technical uncertainty, for example, when those projects directly addressed its copier and printer markets. The company managed to convert its entire technology base from a mechanical base in its early years, to an electromechanical base in its high growth years, to a fully electronic and digital platform in the 1990s (Chesbrough 2003: Ch. 1).

Where the innovation challenge frustrated Xerox was where the company had to apply its promising technologies outside of its current markets and customers. Here, the technical uncertainty that they had to contend with was joined to a new market uncertainty: which customers and which uses of its technology would be most valuable. The personal computer industry had to be invented, in order for these PARC technologies to become valuable.

Coping with market uncertainty greatly complicates the already difficult challenge of managing technical uncertainty, because resolving the technical uncertainty depends on which market the technology is intended to serve, and vice versa. One cannot anticipate the best path forward from the very beginning. Not only is this path unknown, it is unknowable. No amount of planning and research can reveal the facts, because they simply don't exist yet. Instead, a firm must experiment, adapt, and adjust, in response to early feedback. This is a fundamentally different process from the usual process of advancing the current business, more akin to a game of poker than to a game of chess.

Playing poker: The management of false negatives

A large number of false negatives have emerged over the years, where projects that looked initially very unpromising turned out later to be commercially quite valuable. When Intel first obtained its design win for the 8088 microprocessor for the IBM PC, it did not regard this as even ranking among the top 50 prospects for the chip (Moore 1996). IBM almost abandoned a software project (the XML parser) that today forms the centrepiece of its WebSphere Internet services strategy (Chesbrough 2000a). The compound UK-92480 that was under development as a treatment for hypertension within Pfizer did not (p.135) achieve sufficiently positive clinical results to warrant further development. Due to a rather unusual side effect, however, UK-92480 gave rise to one of Pfizer's most profitable compounds today, Viagra. Similarly, Thalidomide, which was driven from the market in the 1960s due to the large number of birth defects encountered by pregnant women taking the drug, has re-entered the market successfully in the late 1990s as the preferred treatment for myeloma, a fatal form of cancer in bone marrow. In this volume, Yamaguchi shows how, one after another of the large Japanese electronic firms and NTT gave up on gallium nitride crystals as the path to creating a blue LED only to find out later that researchers at a small firm outside the mainstream had successfully commercialized the blue LED based on the gallium nitride solution.

How can firms manage these false negatives? By their very nature, false negatives are projects that seem unpromising inside a company due to the lack of fit with the business model of that company. As a result, these projects receive no further support. This is as it should be. One cannot continue to support unpromising initiatives or else nothing would get out into the market. How then can one determine whether or not an unpromising project truly lacks value?

In these situations, a company must develop a second process for managing innovation, a process for playing poker. The analogy comes from Jim McGroddy, the former head of IBM's TJ Watson Research Center:

When you're targeting your technology to your current business, it's like a chess game. You know the pieces, you know what they can and cannot do. You know what your competition is going to do, and you know what your customer needs from you in order to win the game. You can think out many moves in advance, and in fact you have to, if you're going to win.

In a new market, you have to plan your technology entirely differently. You're not playing chess any more, now you're playing poker. You don't know all the information in advance. Instead, you have to decide whether to spend additional money to stay in the game to see the next card.

The metaphor of poker is well suited to situations of high technical and market uncertainty. Not all the information is yet known in these situations, yet companies often manage these situations as though they were just like situations in the main business, where they are playing chess. Xerox was actually very good at chess, at finding technologies to advance its copier and printer business. However, it was a poor poker player, unable to explore the potential options of computing technologies in new markets (Chesbrough 2002, 2003).

To play poker, companies need to meter their capital carefully, and to stage their investments in projects upon the receipt of new information. Projects still have to have funding terminated. But now the company (p.136) must observe what happens after that decision. How are the researchers responding to the decision to terminate further support? Have they moved onto the next project or are they still committing time to the terminated one? If the latter is the case, have they found any external customers for the project?

A second process to play poker is to expose the ‘failures’ to outsiders, to gain their perspective on the potential of these projects. (After all, once you have decided to discontinue their funding there is little at risk for you.) When IBM placed its XML Parser software on its external AlphaWorks website back in 1998, it had discontinued internal funding for the project. However, the number of people who downloaded this particular code from the website was ten times the usual number. To IBM's credit, they took note of this high interest level and began to probe the technology more closely. They reconsidered their earlier decision, and today the XML Parser is a core element of IBM's WebSphere Internet services initiative (Chesbrough 2000a).

A third approach is to out-license the rejected project, which allows another firm to utilize the ideas and see if they are valuable. This not only provides additional funds to the licensing firm, it can allow that licensor to watch and learn from the experience of the licensee. When Intel originally invented the microprocessor, it did so under a contract from Busicom in Japan. As Intel saw what Busicom was doing, it realized that the microprocessor had great potential, and bought back the licence (Moore 1996).

Spin-offs: A Japanese mechanism to manage false negatives?

Organizations can also respond to a potential false negative by creating a new venture to pursue the technology, without being constrained by the current business model. Forming an external spin-off venture allows the technology to develop further outside the originating firm than it would if kept bottled up within. Having an external venture spin-off enables new learning to occur. Moreover, if the venture becomes profitable, the equity owned by the originating firm may become valuable.

This organizational strategy attempts to achieve greater decentralization, higher incentives, and greater focus while preserving coordination with the parent firm. Such endeavours have had a checkered past in the US (Burgelman and Sayles 1986; Block and MacMillan 1993; Chesbrough 2000b, 2003), but are commonly done in Japan (Odagiri 1992; Odagiri and Goto 1993). There are indications that forming new subsidiaries is becoming even more prevalent in Japan (Sako 1997). Companies like Fujitsu are themselves the end product of a series of ‘hivings off’, with the Furukawa group partnering with Siemens to form Fuji Electric, and then Fujitsu spinning off from Fuji Electric in the 1930s. (Fanuc would later spin off from Fujitsu in the 1960s).

(p.137) Spin-off formation may be particularly helpful in the Japanese institutional environment, where labour markets are relatively rigid and start-up venture capital is relatively scarce. These conditions make the creation of de novo start-up firms problematic, and comparatively advantage spin-off companies that are ‘endowed’ by the parent firm with human capital and financial resources.

These subsidiary firms are managed differently from the parent firm in terms of labour policies, pay packages, and sometimes even the union relationship (Sako 1997). Generally, these subsidiaries do not enjoy the status of the parent firm, and must recruit from second- and third-tier universities for their personnel. As such, observers believe that such firms provide technology that is ancillary to the core technology of the parent firm (Odagiri 1992; Okimoto and Nishi 1994). Because of the pressures on lifetime employment in Japan, observers also feel that these subsidiaries are increasingly important to the Japanese employment system, and that ‘lifetime employment’ isnow promised within the enterprise group, rather than within the parent firm (Sako 1997; Kusunoki and Numagami 1998). The work of James Lincoln and his colleagues reported in this volume shows that Japanese firms often utilize an organizational process that will incubate a new technology inside the parent corporation, and then spin it off as a separate entity to commercialize the technology.

There are risks to consider with this approach, however. As Ritschev and Cole (2003) report in their research on ‘internal venture capitalism’ at companies like Sony, large companies may constrain the operations of new spinoffs in ways that reduce their chances for success. One type of constraint is to limit the markets for the venture to those deemed ‘strategic’ by the corporate parent. Another different constraint is, as Ritschev and Cole (2003: 145) put it, ‘Many large Japanese manufacturing firms cannot resist the temptation of solving their problem of redundant fifty-year old engineers through new spin-offs’. These transferred personnel may lack the skills needed for success in the new venture, while the young company lacks the time required to retrain them for the new skill requirements.

Spin-off formation has important benefits for innovation. The new venture's activities reveals new information about the potential for a technology that might otherwise remain latent. When Lucent's New Ventures Group formed Lucent Digital Video as a separate company, it judged that digital video was far from being ready for the market. Once LDV got going, though, it became clear that the market was closer-and bigger-than Lucent originally judged. Lucent found that it was selling hundreds of millions of dollars of telecommunications equipment to the Chinese market, bundled with the digital video encoders from this tiny start-up company. Lucent ended up reacquiring the rest of the venture and hastened its own entry into digital video (Chesbrough and Socolof 2000). Had Lucent not formed the spin-off, it may never have realized the market potential of this technology.

(p.138) Metrics for managing open innovation

IBM also has embraced the idea of enabling others to utilize their own technology. The company reported royalties of US$1.7 billion in 2001, about 15 percent of its operating income that year. It received these royalties in payment for licensing its technology for other companies to use in their businesses. Procter & Gamble (P&G) similarly has set a policy in place that, if a patented technology had not been picked up by at least one P&G business within three years, that technology would be made available to outsiders-even competitors. P&G rightly assumes that its technology is perishable, and that keeping it on the shelf only dissipates any potential value from the technology. If P&G is not going to use it, it is better to let others do so and profit thereby.

P&G is also an active participant in the marketplace for externally generated ideas. It determined that, in 2001, about 10 percent of its pipeline of new products came from external sources. It decided that in order to meet its growth objectives, the percentage of external ideas should rise to 50 percent over the next five years.

If the context of industrial innovation is shifting from closed to open, and if there is latent value in managing false negatives, companies will need to alter their usual metrics for managing innovation. These metrics will help companies play poker as well as chess.

This was the subject of a workshop held at the Industrial Research Institute's Spring Meeting in May 2003. Many large Japanese firms were among the attendees, including representatives from Fujitsu, Hitachi, NEC, and Toshiba. In response to the challenges of managing innovation within an open system, and to monitor the opportunities offered by that system, a number of metrics were identified across multiple small groups within the workshop (reported in Chesbrough 2004). These metrics included:

  1. 1 What %of your sales of products and services last year came from externally licensed technologies? Is this%increasing or decreasing from 2–3 years ago?

  2. 2 What % of your net income last year came from technology licensed out to other companies? Is this % increasing or decreasing from 2–3 years ago?

  3. 3 How long does it take for patented ideas inside the company to be put into use via a company's own products or services (i.e., taken to market via a new product or service)? Has this time interval changed in the past five years? In what direction?

  4. 4 What % of your internal ideas are offered for external licence? How much time elapsed between the patenting of ideas and their external licensing?

  5. 5 How many projects were terminated in the past year? How many were reviewed at a later date? How many subsequently were offered to external parties for further development?

  6. (p.139)
  7. 6 Of the projects tracked in #5, are any of them developing faster technically, and/or growing faster in the market than expected? Are any projects able to raise external capital for further development? Have they signed any major customers?

Metrics 1 and 2 focus management attention on the outputs of the open innovation process, whether that be growth in product sales or growth in licensing activity. Participants in the workshop felt that the senior leadership within their own companies needed their R&D organization's metrics to connect directly to corporate sales and profit measures.

Metrics 3 and 4 focused on a second ‘currency’ for R&D, namely time to market for new products and services, either internally as in metric 3 or externally as is metric 4. Shortening the time required for products and services to get to market was viewed as important, as this increased the rate of learning from R&D for the company, and increased the productivity and effectiveness of R&D as well. A more subtle benefit is that the prediction horizon of the marketing organizations in these organizations was shorter than the usual time it took for the R&D cycle to run its course. Reducing the time to market for new technologies increased the chances that the innovation output was still desired by the market (and that the market hadn't shifted in the meantime).

Participants felt that metrics for managing ‘false negatives’ were at an early stage of understanding. No participants reported any internal tracking system that actively monitored the occurrences of false negatives. The typical pattern was that, once a decision was taken to terminate funding support for a given project, no further tracking of that project was done.

Initial metrics to manage false negative projects in metric 5, therefore, should focus on recording their incidence and build a tracking system to follow them after the initial decision to terminate further support. Metrics like those in metric 6 should evaluate any further progress of potentially false negative projects against the expectations of the company that terminated further funding support. Most projects will likely cease at this point.

When a project continues and makes further progress that significantly exceeds expectations, a re-assessment of the project's technical and/or market potential is warranted. The ability of a project to raise external capital or to sign a major customer, should act as a strong signal that a false negative may exist. A poker playing company may reverse itself at this stage and find a way to get back into the game.

For Japanese companies that are endowed with strong R&D portfolios, it is important that new ways be found to unlock the potential value in these portfolios. Spin-offs are not the only means to do this, but they may be an effective means to explore situations where new business models are needed to commercialize the technologies in new markets. If an established business (p.140) model already exists, then the company would do well to license the technology. But many new R&D programmes lack a clear path to market. These situations are where spin-offs can be most useful. At the moment, venture capital markets and entrepreneurship remain under-developed in Japan. Spinoffs may provide the most effective near-term mechanism to rejuvenate the Japanese innovation system, even as longer term initiatives in higher education and the financial and labour markets begin to bear fruit.

It should be noted in passing that implementing these concepts of open innovation in Japan will also require more proactive management of intellectual property (IP). Japan is actively exploring how to become a leader in IP management, and many of its largest companies have rich portfolios of patents that may enable more open, proactive utilization of Japanese technologies in a wide variety of industrial contexts.

Open up to a point: Open-closed innovation

Mr Hajime Sasaki, chairman of the NEC Corporation, had an interesting and important analysis of open innovation. He argued in an address1 to the Japanese International Intellectual Property Society in Tokyo that the term ‘open innovation’ was a little misleading. He stated that, understood properly, it should be viewed as open-closed innovation. In a forward that he has graciously contributed to the Japanese language version of Open Innovation (Sanno Daigaku Publishers, 2004), Mr Sasaki reminds us that openness is necessary to create value for customers in the innovation process, and to enable a value chain to deliver that value profitably. A certain amount of closed-ness is needed, however, to make a profit from innovation and to be able to continue to innovate in the future. According to Mr Sasaki, at NEC they regard open innovation as an ‘open-closed’ process.

Intel also exemplifies the open-closed approach. Much of the internal R&D it undertakes is done to connect the company to external research in its supply chain (through its Components Research Lab) or to its customers and developers (through its Intel Architecture Labs). Intel also spends more than US$100 million annually in funding university research, seeking new ideas that it can bring into its business. Intel does not own these ideas; it does, however, gain early access to them. So Intel is open in these regards.

To capture value from these ideas, however, Intel uses its internal labs. Most of Intel's internal research is concentrated in its Microprocessor Research Lab, which focuses on new generation Pentium technologies and architectures. It is very closed about the activities in this part of its business and it seldom outlicenses any of its work in this lab to other companies.

So Mr Sasaki's point is well taken. To go further, open innovation concepts are not equally applicable to all industries. For example, the nuclear reactor (p.141) industry depends mainly on internal ideas and has low labour mobility, little venture capital, few (and weak) start-ups, and relatively little research being conducted at universities. Whether this industry will ever migrate towards open innovation is questionable.

At the other extreme, some industries have been open innovators for some time. Consider Hollywood, which replaced the studio system (which in its heyday was highly vertically integrated, and rather closed) with a far more open model. Since at least the 1960s the industry has innovated through a network of partnerships and alliances between production studios, directors, talent agencies, actors, scriptwriters, independent producers, and specialized subcontractors such as the suppliers of special effects. And the mobility of this workforce is legendary: every waitress is a budding actress, every parking attendant has a screenplay he is working on. And everyone has an agent.

Many industries-including those of copiers, computers, disk drives, semiconductors, telecommunications equipment, pharmaceuticals, biotechnology, and even military weapons and communications systems2-are currently undergoing a transition from closed to open. For such businesses, a number of critically important innovations have emerged from seemingly unlikely sources. Indeed, the locus of innovation in these industries has migrated past the confines of the central R&D laboratories of the largest companies and is now situated among various start-ups, universities, research consortia, and other outsiders. And the trend goes well beyond high technology. Other industries such as automotive, health care, banking, insurance, and consumer package goods have also been moving toward open innovation.

Issues for further researcha

While open innovation suggests a greater external focus to industrial R&D, there may be many paths by which to get there. In the area of spin-offs, for example, it is important to contrast the use of voluntary spin-offs in Japan with the US pattern of spin-offs, most of which are involuntary (from the perspective of the originating firm). Involuntary spin-offs result from engineers and managers moving from one company to a competing company without the permission of the first company. This occurs frequently in the US and is almost legendary in places such as Silicon Valley. For the originating company, this flow can be quite disruptive to the continuity of internal research and development activities (Okimoto and Nishi 1994).

While these involuntary spin-offs may be disruptive, they have been quite prolific in many key high tech industries in the US. One can construct a genealogy of disk drive firms from the diaspora of engineers emanating from IBM, Memorex, Control Data, and a few other early entrants into the drive (p.142) industry. A similar genealogy of semiconductor firms also could be developed from firms that emerged out of AT&T, Fairchild, and Texas Instruments.3

Involuntary spin-offs face an exciting, but Darwinian, world of high risk and high reward. When individual engineers and managers perceive new opportunities arising from innovation, they can opt to form an involuntary spin-off. If things go well, the new firm will raise capital, begin product shipments, and perhaps achieve an initial public offering (IPO) or be acquired at an attractive profit. If, however, subsequent events prove unfavourable for the venture, its financial backers will shut it down.

Voluntary spin-offs, which are much more common in Japan, face different prospects. They can help a large firm focus upon a new market opportunity without creating the disruption that might ensue if that opportunity were pursued inside the firm. If the spin-off fails, there may be some possibility for employees to return to the original company. Initial capital provided by the firm reduces the financial difficulty of raising initial start-up capital. And the spin-off generates new knowledge about the market and the technology. This reduces the risks to established firms. As a result they may enter later, but they can enter with greater confidence that they will be able to protect their investments upon entry.

However, creating new spin-off organizations may introduce tensions between the new entity and the parent organization. How these tensions can be managed is an important research question that remains to be answered. While Lincoln's work is encouraging (Chapter 12, this volume), Ritschev and Cole (2003) offer a more mixed assessment. The effectiveness of voluntary spin-offs is not yet well established, and the impact of forming such spin-offs upon the performance of the parent organization after the spin-off has occurred is similarly unexplored. Clair Brown's research in Chapter 8 of this volume also examines the costs and benefits of incorporating more open approaches within the company's core human resource practices.

With these different pathways to open innovation, there may also be different metrics required to track the progress of innovation systems. The processes for managing false positives and false negatives are poorly understood at this point. There may be an analogy to the early days of the quality movement, when Juran and Deming advanced their concepts of management responsibility for quality and statistical process control respectively to a recalcitrant US audience, only to find their ideas enthusiastically embraced in Japan. Quality used to be inspected at the very end of the process, until Juran and Deming's concepts entered into Japanese manufacturing practice. False positives and false negatives also may be identified and managed throughout the innovation process in the future, rather than being identified at the end of the process.

In sum, there are clear changes underway in the industrial innovation system, in the US, in Japan, and throughout the leading industrial economies. (p.143) There is an increasing appreciation that early stage technologies intended for nascent markets suffer from high degrees of measurement error. Both Type I errors (false positives) and Type II errors (false negatives) can arise in the evaluation of these technologies. Companies have designed their R&D evaluation systems to manage the Type I errors, but typically lack any system to manage the risk of Type II errors. Spin-offs stand as one mechanism that offers a means to manage these latter measurement errors, though we have much to learn about their performance in practice.


Bibliography references:

Block, Z. and I. Macmillan (1993). Corporate Venturing: Creating new businesses within the firm, Boston, MA: Harvard Business School Press.

Burgelman, R. and L. Sayles (1986). Inside Corporate Innovation, New York: Free Press.

Chesbrough, H. (2000a). alphaWorks: IBM's technology talent agents, Harvard Business School case #9–601–001.

——(2000b). ‘Designing Corporate Ventures in the Shadow of Private Venture Capital’, California Management Review, Spring, 42 (3): 31–49.

——(2002). ‘Graceful Exits and Foregone Opportunities: Xerox's management of its technology spinoff organizations’, Business History Review, winter, 76(4): 803-38.

——(2003). Open Innovation: The new imperative for creating and profiting from technology, Boston, MA: Harvard Business School Press.

——(2004). ‘Managing Open Innovation: Chess vs. poker’, Research-Technology Management, Jan.-Feb. 47: 13–16.

(p.144) Chesbrough, H. with S. Socolof (2000). ‘Commercializing New Ventures from Bell Labs Technology: The design and experience of Lucent's new ventures group’, Research-Technology Management, March, 43: 1–11.

Kusunoki, K. and T. Numagami (1998). ‘Interfunctional Transfers of Engineers in Japan: Empirical findings and implications for cross-functional integration’, IEEE Transactions.

Moore, G. (1996). ‘Some Personal Perspectives on Research in the Semiconductor Industry’, in R. Rosenbloom and S. William (eds.) Engines of Innovation: Industrial research at the end of an era, Boston, MA: Harvard Business School Press.

Odagiri, H. (1992). Growth Through Competition, Competition Through Growth, Oxford: Clarendon Press.

——and A. Goto (1993). ‘The Japanese System of Innovation: Past, present and future’, in R. Nelson (ed.) National Innovation Systems: A comparative analysis, Oxford: Oxford University Press.

Okimoto, D. and Y. Nishi (1994). ‘R&D Organization in Japanese and American Semiconductor Firms’, in M. Aoki and R. Dore (eds.) The Japanese Firm: Sources of Competitive Strength, Oxford: Oxford University Press.

Ritschev, D. and R. Cole (2003). ‘Social and Structural Barriers to the IT Revolution in High-Tech Industries’, in J. Bachnik (ed.) Roadblocks on the Information Highway: The IT revolution in Japanese education, New York: Lexington Books.

Sako, M. (1997). ‘Forces for Homogeneity and Diversity in the Japanese Industrial Relations Systems’, in M. Sako and H. Sato (eds.) Japanese Labour and Management in Transition, London: Routledge.


(1.) Keynote Address at the Japan International Patent Licensing Seminar, meeting of the Japan International Intellectual Property Society, Royal Park Hotel, Tokyo, 27 January 2004.

(2.) In 2000, the Central Intelligence Agency financed a venture capital firm, In-Q-Tel, intended to assist the intelligence agency in identifying promising technologies from start-up companies. The reason for forming this unusual organization was that the defence procurement process is so onerous that most start-up companies avoid selling to the government. Because important new technologies are emerging from start-ups in areas such as software and cryptography, to take two, the CIA decided it needed a new process to access this technology.

(3.) An important, but seldom recognized, dependency emerges here. Start-up firms that raid the talent of established firms are highly dependent upon the presence of successful, established firms to supply the management and technical talent they require. Start-up firms, and the venture capital that funds these firms, have no interest in paying for training for their people. Without an ample supply of qualified people to hire from these start-ups would be greatly impaired in their ability to grow.