Open Innovation, page 26
The freedom to conduct [economic] experiments has been the essential element accounting for the fact that industrialization has been, uniquely, anhistorical product of capitalist societies. . . . The freedom to conduct experiments is essential to any society that has a serious commitment to technological innovation. . . .
. . . There is an additional advantage to a system that encourages, or at least tolerates, multiple sources of decision-making. Not only do human agents differ considerably in their attitudes towards risk; they differ also in their skills, capabilities, and orientations, however those differences may have been acquired. This heterogeneity . . . constitutes a valuable resource that is muchmore readily enlisted into the realm of potentially useful experimentation byan organizationally decentralized environment. An economy that includessmall firms and easy entry conditions is likely to benefit from this pool of human talent far more than one dominated by centralized decision-making.
I would add that a firm that monitors and selectively accesses the decentralized environment will have an advantage over even very capable firms that do not.
12. James McGroddy, telephone interview with author,23 July 1999.
13. John Seely Brown, telephone interview with author, 6 April 1999.
14. Other research in other contexts has noted the myopia that can afflict companies’ forecasts and their resulting impact on the allocation of resources. This shortsightedness can entrench the current customers at the expense of potential new customers. See, for example, Clay Christensen and Joseph Bower, “Customer Power, Strategic Investment, andthe Failure of Leading Firms,” Strategic Management Journal17, no. 3(1996): 197–218.
15. Figure 1-2 shows Xerox’s process in 1996. For an analysis of how Xerox’s process evolved from 1979 through 1998, see Chesbrough, “Graceful Exits and Foregone Opportunities.”
16. An example of this problem within PARC was a project involving software to enable interactive collaboration over the Internet in the mid-1990s. The PARC research project manager spent several months visiting various Xerox businesses to see whether the technology could be applied internally. To his surprise, he found that the Xerox sales organization did not even use e-mail at the time, to communicate either internally or externally. As a result, these Xerox personnel were singularly unhelpful in identifying the potential value of Internet collaboration. Consequently, Xerox could not even use its own organization as a test bed for future innovation. While PARC made sure its researchers “ate their own cooking” by having them use many ofthe emerging technologies, it was far less successful in motivating other parts of the Xerox organization to try them out as well.
17. Amar Bhidé, The Origin and Evolution of New Businesses (Oxford: Oxford University Press, 2000), also contrasts corporate processes for evaluating new projects tothose of venture capitalists. He recounts many differences between these two groupsthat are consistent with my argument here, particularly the much greater depth andobjectivity of corporate evaluations, relative to those of venture capitalists. He alsonotes the high level of inertia within corporate decision processes, relative to the rapidadaptation of the VC decision processes.
18. This has a powerful implication that derives from complexity theory. When companies must take action in highly complex environments, the companies that can adapt more rapidly will fare better than companies that start in a more favorable location, but are less able to move from that location. See Giovanni Gavetti and Dan Levinthal, “Looking Forward and Looking Backward: Cognitive and Experiential Search,” Administrative Science Quarterly 45(2000): 113–137. If VC-backed companies are more adaptive to their environment than are similar ventures housed inside large corporations, the VC-backed companies may fare better in the market, even if they start with inferior technology or fewer resources.
Chapter 2
1. The work of Edith Penrose, back in 1959, provides one of the first academic explanations for why firms conduct their own internal research. “After all, the specialized firm is vulnerable. Its profitability and very survival as a firm are imperilled [sic]by adverse changes in demand for the types of products it produces and by increased competition from other producers.” Edith Penrose, The Theory of the Growth of the Firm, 2nd ed. (Oxford: Oxford University Press, 1995), 112 – 113 . In this view, internal industrial research is both an option for future growth and, even more, an insurance policy against adverse changes in the firm’s environment.
2. Rowland’s rant is quoted in David Hounshell’s entertaining account of the rise of U.S. industrial research laboratories, “The Evolution of Industrial Research in the United States,” in Engines of Innovation: U.S. Industrial Research at the End of an Era, ed. Richard Rosenbloom and William Spencer (Boston: Harvard Business School Press, 1996), 16.
3. Leading business historians such as David Hounshell and John Kenly Smith date the first establishment of industrial research laboratories to 1856, when William Perkin discovered how to synthesize a mauve dye from aniline. This was augmented bythe discovery of alizarin dyes in 1868 and the azo dyes in 1876. By 1890, the German firm Bayer had constructed a modern laboratory with books, lab equipment, and a dedicated staff. Bayer’s laboratory was soon copied throughout the chemicals industry, and many firms in the United States took their cues from Germany, which was then the acknowledged leader in industrial technology. See David Hounshell and John Kenly Smith, Science and Corporate Strategy: DuPont R&D, 1902–1980(Cambridge, England:Cambridge University Press, 1988), 4.
4. Alfred D. Chandler, Scale and Scope: The Dynamics of Industrial Capitalism (Cambridge, MA: Harvard University Press,1990); Chandler, The Visible Hand: The Managerial Revolution in American Business (Cambridge, MA: Harvard University Press, 1977); Chandler, Strategy and Structure: Chapters in the History of American Industrial Enterprise (Cambridge, MA: MIT Press, 1962).
5. Some of the castle walls were occasionally knocked down by government enforcement of the Sherman Antitrust Act, enacted and enforced after 1898. Even this antitrust enforcement, though, further encouraged firms to invest in proprietary research tocreate new products as they sought to maintain their market positions. In hindsight, then, the Sherman Act motivated dominant firms to retain their position through horizontal mergers (instead of vertical monopolies) and by investing in new innovations. See David Mowery and Nathan Rosenberg, “The U.S. National Innovation System,” in National Innovation Systems, ed. Richard Nelson (Oxford: Oxford University Press,1993 ), 37 .
6. Mowery and Rosenberg, “The U.S. National Innovation System,” 37.
7. Franklin D. Roosevelt, quoted in Vannevar Bush, Science: The Endless Frontier (Washington, DC: U.S. Government Printing Office, 1945), xi.
8. Bush, Science: The Endless Frontier, 14.
9. Bush, 133.
10. In Henry Chesbrough and Clay Christensen, “Technology Markets, Technology Organization, and Appropriating the Returns to Research,” working paper 99–115, Harvard Business School, Boston, 1999, Christensen and I provide an extended discussion of the history of IBM’s shifting policies toward external and internal supply of critical components in its disk drives. We show that when IBM tried to use outside suppliers and found them unreliable, IBM took the supply of these components in-house. Later, external suppliers became far more capable, but IBM chosenot to utilize them even when it now could rely on them. Starting in 1993, IBM has changed this behavior, a transition I discuss in detail in chapter 5.
11. The many challenges of integrating technologies are thoroughly explored in Marco Iansiti, Technology Integration: Making Critical Choices in a Dynamic World (Boston:Harvard Business School Press, 1998).
12. Suppose, for example, that a Merck research scientist developed a chemical compound whose molecular structure became quite valuable. If that scientist chose to go to a new start-up company, the compound would incontestably remain Merck’s property, and Merck could appropriate its value without fear of leakage. However, in other industries in the information technology sector, similar employee defections have been associated with substantial diffusion of valuable knowledge without any compensation received by the former employer.
13. Henry Chesbrough, “Environmental Influences upon Firm Entry into New Sub-Markets,” Research Policy (in press), shows that firms with former IBM executives were associated with higher rates of entry into new market segments in the hard-disk-drive industry, compared to U.S. firms that did not have these former executives.
14. This is not the end of the story in many industries. Research by Jae Yong Song, Paul Almeida, and Geraldine Wu, “Mobility of Engineers and Cross Border Knowledge Building: The Technological Catching-Up Case of Korean and Taiwanese Semiconductor Firms,” Comparative Studies of Technological Evolution, vol. 7, Research on Technological Innovation, Management, and Policy (Oxford: Elsevier Science, 2001), 59–84, shows that many of these foreign national professionals eventually return to their home countries and bring substantial knowledge and know-how with them when they return. Using patent data in the semiconductor industry, Song and his coworkers show how Korean and Taiwanese semiconductor firms were able to catch up to the leading-edge U.S. and Japanese firms through these “returning brains.”
15. National Science Foundation, Science Resource Studies: Survey of Graduate Students and Postdocs (Washington, DC: National Science Foundation, 1998),
16. For a well-done, accessible introduction to VC and its role in financing innovation, see Paul Gompers and Josh Lerner, The Money of Invention: How Venture Capital Creates New Wealth (Boston: Harvard Business School Press, 2001).
Chapter 3
1. J. Thursby and S. Kemp, “Growth and Productive Efficiency in University Intellectual Property Licensing,” Research Policy 3, no 1(2002): 109–124, concluded that U.S. universities were becoming more commercially productive with their research. They report that university patents have risen from 250 in 1980 to over 1,500 annually in 2000, and provide interesting evidence that universities are getting more output, as measured by the number of licenses they receive for these patents, per unit of“input.
2. Samuel Kortum and Josh Lerner, “What Is Behind the Recent Surge in Patenting?” Research Policy 28(January 1999): 1–22.
3. In the earlier era, large companies also looked down on the quality of R&D activity conducted by smaller companies, but no more. Today, the quality of technical personnel in start-up firms can be surprisingly high. Managers at corporate research centers such as PARC report that their biggest competition in hiring brilliant new researchers out of leading university Ph.D. programs is not other research centers, such as IBM’s Watson Research Center, Lucent’s Bell Labs, or even a federal lab. It is start-up firms and universities. When these groups can lure the best and the brightest to their organizations, away from the large company laboratories, the perceived historic superiority of large-firm R&D can no longer be taken for granted.
4. National Science Foundation, Science and Engineering Indicators, NSF/Scientific Resource Study (Washington, DC: National Science Foundation, 1998),
5. The overall length of tenure remains at three and one-half years, from 1983 to 2000, but this understates the mobility of the work force because of its aging in those years (older workers are less mobile than younger workers). Within age groups, the length of tenure has declined for all age groups, between 1983 and 2000. See
6. Even now, after the VC bubble has popped and investing has returned to 1998 levels, VC remains a powerful force to be reckoned with, relative to what companies arespending overall in their R&D. The VC world invested $48 billion in the United States in 1999(Venture Economics Web site,
7. I heard many variations of these concerns at the annual meeting of the Industrial Research Institute in Williamsburg in 1999,
8. Merck & Co., Annual Report (2000), 8.
9. Ibid.
10. See the Venture Economics Web site,
11. In Henry Chesbrough, “Making Sense of Corporate Venture Capital,” Harvard Business Review, March 2002, 90 – 99, I explore how companies can utilize corporate venture investments to advance their own strategic goals.
12. The use of the ecosystem as a metaphor for how businesses compete and survive was well employed earlier by James Moore, The Death of Competition: Leadership and Strategy in the Age of Business Ecosystems (New York: HarperBusiness, 1996). The point I am making here is how venture capitalists play a very positive role in creating, shaping, and developing the ecosystem.
13. See Eric von Hippel, The Sources of Innovation (New York: Oxford University Press, 1988), for a superb account of the powerful role that lead users can play in the innovation process.
14. Internal competition should not be avoided, but it will need to be managed. For a useful approach to managing such competition, see Julian Birkinshaw, “Strategies for Managing Internal Competition,” California Management Review 44, no. 1(2002): 21–38. Internal competition cuts both ways. Internal technology groups may move faster to respond to the needs of their marketing and sales divisions when marketing and sales have recourse to external technology sources as well. Internal technology groups ignore their downstream division’s needs, or are late to respond, at their own peril. If the downstream business can access an alternative technology outside, it chastens the internal upstream group while protecting the overall firm from being late in the market. A better mousetrap tomorrow may not be as valuable as a good mousetrap available today.
15. The argument about the relationship of technical complexity to organizational integration is developed at length in Henry Chesbrough and Clay Christensen,“Technology Organization, Technology Markets, and the Returns to Research,” working paper 99–115, Harvard Business School, Boston, 1999. In that paper, we also show that modularity need not be the end state of a technology’s evolution. There can be cycling between vertical integration and modularity, followed by a return to integration. The role of internal R&D in resolving complex technological interdependencies was also discussed in Henry Chesbrough and Ken Kusunoki, “The Modularity Trap: Innovation, Technology Phase Shifts and the Resulting Limits of Virtual Organizations,” in Managing Industrial Knowledge,ed. I. Nonaka and D. Teece (London: Sage Press, 2001), which discussed the Japanese hard-disk-drive industry.
16. See also Michael Cusumano and Annabelle Gawer, Platform Leadership: How Intel, Microsoft, and Cisco Drive Industry Innovation (Boston: Harvard Business School Press, 2002), for other examples of how companies coordinate the architecture of a system without making each of its parts. The book does a good job of conveying the need for coordination within a platform, which specifies the relationship between the interdependent parts, as well as the need to evolve the platform, so that its performance does not stagnate over time.
17. Henry Chesbrough and David Teece, “When Is Virtual Virtuous? Organizing for Innovation,” Harvard Business Review, January–February 1996, recounted the experience of IBM in the PC industry. IBM behaved very virtually for a company ofits size, creating an independent business unit to develop an open architecture for its IBM PC. As part of its drive to move fast and remain flexible, the company outsourced the microprocessor from Intel and the operating system from Microsoft. However,IBM subsequently lost control of its architecture, and today the profits from the PC architecture that IBM created flow through to Intel and Microsoft. Chesbrough and Kusunoki (“The Modularity Trap”) explore how companies need to shift organizational modes as an industry becomes modular. They explain that companies must nonetheless retain enough systems knowledge to shift back to a more integrated mode when an architecture reaches its performance limit and a new generation of architecture must be created.
Chapter 4
This chapter draws heavily from Henry Chesbrough and Richard Rosenbloom, “The Role of the Business Model in Capturing Value from Innovation,” Industrial and Corporate Change 11, no. 3 (2002): 529 – 556 . My colleague Dick Rosenbloom has been a student of managing innovation for many decades and has had the opportunity to work with Xerox and PARC for many years. I found it extremely valuable to compare the insights I had gained from studying the commercialization practices of “the ones that got away” with Dick’s deep knowledge of Xerox’s processes for commercializing technology in its own businesses.
1. See Chesbrough and Rosenbloom, “Role of the Business Model,” for a more academic treatment of this definition and its roots in the earlier business strategy literature. That paper also points out the importance of the cognitive element of the business model, which is absent from most definitions of the topic. I will discuss that aspect later in this chapter, in the context of Xerox’s evaluation of its PARC research technologies.
