The Innovator's Solution

by Clayton Christensen

  Because theory-building scholars struggle to define the right and relevant categorization of circumstances, they rarely can define the circumstances immediately. Early studies almost always sort researchers’ observations into categories defined by the attributes of the phenomena themselves. Their assertions about the actions or events that lead to the results at this point can only be statements about correlation between attributes and results, not about causality. This is the best they can do in early theory-building cycles.

  Consider, for illustration, the history of man’s attempts to fly. Early researchers observed strong correlations between being able to fly and having feathers and wings. Possessing these attributes had a high correlation with the ability to fly, but when humans attempted to follow the “best practices” of the most successful flyers by strapping feathered wings onto their arms, jumping off cliffs, and flapping hard, they were not successful—because as strong as the correlations were, the would-be aviators had not understood the fundamental causal mechanism that enabled certain animals to fly. It was not until Bernoulli’s study of fluid mechanics helped him articulate the mechanism through which airfoils create lift that human flight began to be possible. But understanding the mechanism itself still wasn’t enough to make the ability to fly perfectly predictable. Further research, entailing careful experimentation and measurement under various conditions, was needed to identify the circumstances in which that mechanism did and did not yield the desired result.

  When the mechanism did not result in successful flight, researchers had to carefully decipher why—what it was about the circumstances in which the unexpected result occurred that led to failure. Once categories could be stated in terms of the different types of circumstances in which aviators might find themselves, then aviators could predict the conditions in which flight was and was not possible. They could develop technologies and techniques for successfully flying in those circumstances where flight was viable. And they could teach aviators how to recognize when the circumstances were changing, so that they could change their methods appropriately. Understanding the mechanism (what causes what, and why) made flight possible; understanding the categories of circumstances made flight predictable.20

  How did aviation researchers know what the salient boundaries were between these categories of circumstance? As long as a change in conditions did not require change in the way the pilot flew the plane, the boundary between those conditions didn’t matter. The circumstance boundaries that mattered were those that mandated a fundamental change in piloting techniques in order to keep the plane flying successfully.

  Similar breakthroughs in management research increase the predictability of creating new-growth businesses. Getting beyond correlative assertions such as “Big companies are slow to innovate,” or “In our sample of successful companies, each was run by a CEO who had been promoted from within,” the breakthrough researcher first discovers the fundamental causal mechanism behind the phenomena of success. This allows those who are looking for “an answer” to get beyond the wings-and-feathers mind-set of copying the attributes of successful companies. The foundation for predictability only begins to be built when the researcher sees the same causal mechanism create a different outcome from what he or she expected—an anomaly. This prompts the researcher to define what it was about the circumstance or circumstances in which the anomaly occurred that caused the identical mechanism to result in a different outcome.

  How can we tell what the right categorization is? As in aviation, a boundary between circumstances is salient only when executives need to use fundamentally different management techniques to succeed in the different circumstances defined by that boundary. If the same statement of cause and effect leads to the same outcome in two circumstances, then the distinction between those circumstances is not meaningful for the purposes of predictability.

  To know for certain what circumstances they are in, managers also must know what circumstances they are not in. When collectively exhaustive and mutually exclusive categories of circumstances are defined, things get predictable: We can state what will cause what and why, and can predict how that statement of causality might vary by circumstance. Theories built on categories of circumstances become easy for companies to employ, because managers live and work in circumstances, not in attributes.21

  When managers ask questions such as “Does this apply to my industry?” or “Does it apply to service businesses as well as product businesses?” they really are probing to understand the circumstances. In our studies, we have observed that industry-based or product/ service-based categorization schemes almost never constitute a useful foundation for reliable theory. The Innovator’s Dilemma, for example, described how the same mechanism that enabled entrant companies to up-end the leading established firms in disk drives and computers also toppled the leading companies in mechanical excavators, steel, retailing, motorcycles, accounting software, and motor controls.22 The circumstances that mattered were not what industry you were in. Rather, there was a mechanism—the resource allocation process—that caused the established leaders to win the competitive fights when an innovation was financially attractive to their business model. The same mechanism disabled the established leaders when they were attacked by disruptive innovators—whose products, profit models, and customers were not attractive.

  We can trust a theory only when its statement of what actions will lead to success describe how this will vary as a company’s circumstances change.23 This is a major reason why the outcomes of innovation efforts have seemed quite random: Shoddy categorization has led to one-size-fits-all recommendations that in turn have led to the wrong results in many circumstances.24 It is the ability to begin thinking and acting in a circumstance-contingent way that brings predictability to our lives.

  We often admire the intuition that successful entrepreneurs seem to have for building growth businesses. When they exercise their intuition about what actions will lead to the desired results, they really are employing theories that give them a sense of the right thing to do in various circumstances. These theories were not there at birth: They were learned through a set of experiences and mentors earlier in life.

  If some people have learned the theories that we call intuition, then it is our hope that these theories also can be taught to others. This is our aspiration for this book. We hope to help managers who are trying to create new-growth businesses use the best research we have been able to assemble to learn how to match their actions to the circumstances in order to get the results they need. As our readers use these ways of thinking over and over, we hope that the thought processes inherent in these theories can become part of their intuition as well.

  We have written this book from the perspective of senior managers in established companies who have been charged to maintain the health and vitality of their firms. We believe, however, that our ideas will be just as valuable to independent entrepreneurs, start-up companies, and venture capital investors. Simply for purposes of brevity, we will use the term product in this book when we describe what a company makes or provides. We mean, however, for this to encompass product and service businesses, because the concepts in the book apply just as readily to both.

  The Outline of This Book

  The Innovator’s Dilemma summarized a theory that explains how, under certain circumstances, the mechanism of profit-maximizing resource allocation causes well-run companies to get killed. The Innovator’s Solution, in contrast, summarizes a set of theories that can guide managers who need to grow new businesses with predictable success—to become the disruptors rather than the disruptees—and ultimately kill the well-run, established competitors. To succeed predictably, disruptors must be good theorists. As they shape their growth business to be disruptive, they must align every critical process and decision to fit the disruptive circumstance.

  Because building successful growth businesses is such a vast topic, this book focuses on nine of the most important decisions that all manag
ers must make in creating growth—decisions that represent key actions that drive success inside the black box of innovation. Each chapter offers a specific theory that managers can use to make one of these decisions in a way that greatly improves their probability of success. Some of this theory has emerged from our own studies, but we are indebted to many other scholars for much of what follows. Those whose work we draw upon have contributed to improving the predictability of business building because their assertions of causality have been built upon circumstance-based categories. It is because of their careful work that we believe that managers can begin using these theories explicitly as they make these decisions, trusting that their predictions will be applicable and reliable, given the circumstances that they are in.

  The following list summarizes the questions we address.

  Chapter 2: How can we beat our most powerful competitors? What strategies will result in the competitors killing us, and what courses of action could actually give us the upper hand?

  Chapter 3: What products should we develop? Which improvements over previous products will customers enthusiastically reward with premium prices, and which will they greet with indifference?

  Chapter 4: Which initial customers will constitute the most viable foundation upon which to build a successful business?

  Chapter 5: Which activities required to design, produce, sell, and distribute our product should our company do internally, and which should we rely upon our partners and suppliers to provide?

  Chapter 6: How can we be sure that we maintain strong competitive advantages that yield attractive profits? How can we tell when commoditization is going to occur, and what can we do to keep earning attractive returns?

  Chapter 7: What is the best organizational structure for this venture? What organizational unit(s) and which managers should contribute to and be responsible for its success?

  Chapter 8: How do we get the details of a winning strategy right? When is flexibility important, and when will flexibility cause us to fail?

  Chapter 9: Whose investment capital will help us succeed, and whose capital might be the kiss of death? What sources of money will help us most at different stages of our development?

  Chapter 10: What role should the CEO play in sustaining the growth of the business? When should CEOs keep their hands off the new business, and when should they become involved?

  The issues that we tackle in these chapters are critical, but they cannot constitute an exhaustive list of the questions that should be relevant to launching a new-growth business. We can simply hope that we have addressed the most important ones, so that although we cannot make the creation of new-growth businesses perfectly risk free, we can help managers take major steps in that direction.

  Notes

  1. Although we have not performed a true meta-analysis, there are four recently published studies that seem to converge on this estimate that roughly one company in ten succeeds at sustaining growth. Chris Zook and James Allen found in their 2001 study Profit from the Core (Boston: Harvard Business School Press) that only 13 percent of their sample of 1,854 companies were able to grow consistently over a ten-year period. Richard Foster and Sarah Kaplan published a study that same year, Creative Destruction (New York: Currency/Doubleday), in which they followed 1,008 companies from 1962 to 1998. They learned that only 160, or about 16 percent of these firms, were able merely to survive this time frame, and concluded that the perennially outperforming company is a chimera, something that has never existed at all. Jim Collins also published his Good to Great (New York: HarperBusiness) in 2001, in which he examined a universe of 1,435 companies over thirty years (1965–1995). Collins found only 126, or about 9 percent, that had managed to outperform equity market averages for a decade or more. The Corporate Strategy Board’s findings in Stall Points (Washington, DC: Corporate Strategy Board, 1988), which are summarized in detail in the text, show that 5 percent of companies in the Fortune 50 successfully maintained their growth, and another 4 percent were able to reignite some degree of growth after they had stalled. The studies all support our assertion that a 10 percent probability of succeeding in a quest for sustained growth is, if anything, a generous estimate.

  2. Because all of these transactions included stock, “true” measures of the value of the different deals are ambiguous. Although when a deal actually closes, a definitive value can be fixed, the implied value of the transaction at the time a deal is announced can be useful: It signals what the relevant parties were willing to pay and accept at a point in time. Stock price changes subsequent to the deal’s announcement are often a function of other, exogenous events having little to do with the deal itself. Where possible, we have used the value of the deals at announcement, rather than upon closing. Sources of data on these various transactions include the following:

  NCR

  “Fatal Attraction (AT&T’s Failed Merger with NCR),” The Economist, 23 March 1996.

  “NCR Spinoff Completes AT&T Restructure Plan,” Bloomberg Business News, 1 January 1997.

  McCaw and AT&T Wireless Sale

  The Wall Street Journal, 21 September 1994.

  “AT&T Splits Off AT&T Wireless,” AT&T news release, 9 July 2001.

  AT&T, TCI, and MediaOne

  “AT&T Plans Mailing to Sell TCI Customers Phone, Web Services,” The Wall Street Journal, 10 March 1999.

  “The AT&T-Mediaone Deal: What the FCC Missed,” Business Week, 19 June 2000.

  “AT&T Broadband to Merge with Comcast Corporation in $72 Billion Transaction,” AT&T news release, 19 December 2001.

  “Consumer Groups Still Questioning Comcast-AT&T Cable Merger,” Associated Press Newswires, 21 October 2002.

  3. Cabot’s stock price outperformed the market between 1991 and 1995 as it refocused on its core business, for two reasons. On one side of the equation, demand for carbon black increased in Asia and North America as car sales surged, thereby increasing the demand for tires. On the supply side, two other American-based producers of carbon black exited the industry because they were unwilling to make the requisite investment in environmental controls, thereby increasing Cabot’s pricing power. Increased demand and reduced supply translated into a tremendous increase in the profitability of Cabot’s traditional carbon black operations, which was reflected in the company’s stock price. Between 1996 and 2000, however, its stock price deteriorated again, reflecting the dearth of growth prospects.

  4. An important study of companies’ tendency to make investments that fail to create growth was done by Professor Michael C. Jensen: “The Modern Industrial Revolution, Exit, and the Failure of Internal Control Systems,” Journal of Finance (July 1993): 831–880. Professor Jensen also delivered this paper as his presidential address to the American Finance Association. Interestingly, many of the firms that Jensen cites as having productively reaped growth from their investments were disruptive innovators—a key concept in this book.

  Our unit of analysis in this book, as in Jensen’s work, is the individual firm, not the larger system of growth creation made manifest in a free market, capitalist economy. Works such as Joseph Schumpeter’s Theory of Economic Development (Cambridge, MA: Harvard University Press, 1934) and Capitalism, Socialism, and Democracy (New York: London, Harper & Brothers, 1942) are seminal, landmark works that address the environment in which firms function. Our assertion here is that whatever the track record of free market economies in generating growth at the macro level, the track record of individual firms is quite poor. It is the performance of firms within a competitive market to which we hope to contribute.

  5. This simple story is complicated somewhat by the market’s apparent incorporation of an expected “fade” in any company’s growth rate. Empirical analysis suggests that the market does not expect any company to grow, or even survive, forever. It therefore seems to incorporate into current prices a foreseen decline in growth rates from current levels and the eventual dissolution of the firm. This is the reason for the importance of terminal val
ues in most valuation models. This fade period is estimated using regression analysis, and estimates vary widely. So, strictly speaking, if a company is expected to grow at 5 percent with a fade period of forty years, and five years into that forty-year period it is still growing at 5 percent, the stock price would rise at rates that generated economic returns for shareholders, because the forty-year fade period would start over. However, because this qualification applies to companies growing at 5 percent as well as those growing at 25 percent, it does not change the point we wish to make; that is, that the market is a harsh taskmaster, and merely meeting expectations does not generate meaningful reward.

  6. On average over their long histories, of course, faster-growing firms yield higher returns. However, the faster-growing firm will have produced higher returns than the slower-growing firm only for investors in the past. If markets discount efficiently, then the investors who reap above-average returns are those who were fortunate enough to have bought shares in the past when the future growth rate had not been fully discounted into the price of the stock. Those who bought when the future growth potential already had been discounted into the share price would not receive an above-market return. An excellent reference for this argument can be found in Alfred Rappaport and Michael J. Mauboussin, Expectations Investing: Reading Stock Prices for Better Returns (Boston: Harvard Business School Press, 2001). Rappaport and Mauboussin guide investors in methods to detect when a market’s expectations for a company’s growth might be incorrect.

  7. These were the closing market prices for these companies’ common shares on August 21, 2002. There is no significance to that particular date: It is simply the time when the analysis was done. HOLT Associates, a unit of Credit Suisse First Boston (CSFB), performed these calculations using proprietary methodology applied to publicly available financial data. The percent future is a measure of how much a company’s current stock price can be attributed to current cash flows and how much is due to investors’ expectations of future growth and performance. As CSFB/HOLT defines it,