The Innovator's Solution

by Clayton Christensen

  Digital camera owners use their cameras for jobs they already had been trying to get done. For example, most of us use such a camera to verify on the spot that the image is good, and if it isn’t, we delete it and try again—the same job as taking multiple shots on film of the same pose. And we send digital images much less expensively and conveniently to far more people over the Internet than we ever had been able to do when we ordered double prints. (Interestingly, have you noticed what we do after we’ve looked at an image that has been e-mailed to us? We click “close,” putting it back in some “envelope” on our hard drive.) The things we prioritize in our lives are remarkably stable.

  Another example: Hundreds of millions have been spent to apply new technologies—the Internet and e-book displays, specifically—to reshape the college textbook industry. Innovators have attempted to develop and sell tablets that can display downloaded e-books. And with many textbooks, you can click on a URL to obtain far more information about the topic than could possibly be included within the limits of a book. Would we expect these investments to generate significant growth? Our guess is that they will not. Although we would like to believe that all undergraduate students are rigorous seekers of knowledge, the job that many college students are really trying to get done, from our observation, is to pass their courses without having to read the textbook at all.

  These companies have spent a lot of money helping students to do more easily something that they have been trying not to do. It would probably take far less money to create from the same technology a service called “Cram.com”—a utility that would make it easier and cheaper for students to cram more effectively for their exams. This would likely work because cramming is something that students already are trying to do, but with marginal efficacy. There are a lot of textbook-avoiders on campuses—a huge market of nonconsumption.

  After logging on, Cram.com would ask subscribers what course they need to cram for—say, College Algebra. Then it would ask which of this list of textbooks the professor expected them to have read by now. It would ask them to click on the type of problem that they are having trouble with, and it would walk them through a tutorial.

  The next year, Cram.com would need to offer a new and improved service, one that made it even easier and faster to cram better—inching up from the least-conscientious to the sporadically diligent tiers of the student population. After a few years, two students might be overheard in the college bookstore anguishing over the exorbitant price of a textbook: “You know, my brother took that course last year. He’s a good student, but he never even bought the book. He just used Cram.com from the beginning of the semester, and he did great.” Bingo. A new-market disruption that helped customers achieve what they already had been trying to do.

  Identifying disruptive footholds means connecting with specific jobs that people—your future customers—are trying to get done in their lives. The problem is that in an attempt to build convincing business cases for new products, managers are compelled to quantify the opportunities they perceive, and the data available to do this are typically cast in terms of product attributes or the demographic and psychographic profiles of a given population of potential consumers. This mismatch between the true needs of consumers and the data that shape most product development efforts leads most companies to aim their innovations at nonexistent targets. The importance of identifying these jobs to be done goes beyond simply finding a foothold. Only by staying connected with a given job as improvements are made, and by creating a purpose brand so that customers know what to hire, can a disruptive product stay on its growth trajectory.

  Notes

  1. See, for example, chapter 7 in Dorothy Leonard, Wellsprings of Knowledge (Boston: Harvard Business School Press, 1996).

  2. Some researchers (for example, Joe Pine, in his classic work Mass Customization [Boston: Harvard Business School Press, 1992]) argues that ultimately segmentation may be unimportant because individual customers’ needs might be addressed individually. Although this is conceivable, getting there will take some time. We will show in chapters 5 and 6 that in many circumstances it is not possible. Segmentation, in other words, will always be important.

  3. We are deeply indebted to two of our colleagues who originally introduced us to this way of thinking about the structure of markets. The first is Richard Pedi, CEO of Gage Foods in Bensenville, Illinois. Rick coined for us the language “jobs to be done.” Independently, Anthony Ulwick of Lansana, Florida–based Strategyn, Inc., has developed and used a very similar concept in his consulting work, using the phrase “outcomes that customers are seeking.” Tony has published a number of pieces on these concepts, including “Turn Customer Input into Innovation,” Harvard Business Review , January 2002, 91–98. Tony uses these concepts to help his firm’s clients develop products that connect with what their customers are trying to get done. We are also indebted to David Sundahl, who as Professor Christensen’s research associate helped formulate many of the initial ideas upon which this chapter was built.

  4. Many of the details in this account have been changed to protect the proprietary interests of the company while preserving the fundamental character of the study and its conclusions.

  5. The language in this paragraph reveals a nested system. Within the overarching job to be done are many unique outcomes that need to be achieved in order for the job to be done perfectly. Hence, when we use the term outcome in our work on segmentation, we refer to the individual things that need to be done right, such as lasting a long time, not creating a mess, and so on, in order for the job to get done right.

  6. One can see this problem even in the recent marketing trend toward so-called markets of one. Markets of one drive companies to provide customization options that meet all the needs of individual customers. But customization comes at a price. What is more, it often does not provide an understanding of the underlying outcomes-driven logic of customer purchasing decisions. Because market research tools as sophisticated as geocoding pay attention to the attributes of people, they cannot yield market segmentation schemes that make sense to customers—each of whom has many jobs that he or she is trying to get done. There actually is a lot of commonality in jobs to be done within a population of people and companies, suggesting that targeting markets of one may often not be a viable or desirable marketing objective.

  7. The observation that customers search across product categories to find ways to achieve needed outcomes is grounded in psychological research, which demonstrates that our perceptual systems are geared toward understanding what we can use objects to do and whether they are optimal for such purposes. For example, psychologist James J. Gibson, widely respected for his research on theories of perception, has written about “affordances,” a concept that mirrors what we term “jobs” or “outcomes.” According to Gibson, “The affordances of the environment are what it offers . . . , what it provides or furnishes, either for good or ill.” Gibson asserts that we see the world not in terms of primary qualities, like being yellow or being twenty-four ounces by volume, but in terms of outcomes: “What we perceive when we look at objects are their [outcomes], not their qualities. We can discriminate the dimensions of difference if required to do so in an experiment, but what the object affords us is what we normally pay attention to.” What matters about the ground, for example, is that it provides us a platform on which to stand, walk, build, and so forth. We don’t “hire” the ground for its color or moisture content per se. The affordances of products, in Gibson’s terms, are the outcomes that those products enable their users to achieve. See James J. Gibson, The Ecological Approach to Visual Perception (Boston: Houghton Mifflin, 1979), 127.

  8. Finding a “killer app” has been a holy grail of innovators ever since Larry Downes and Chunka Mui popularized the term in Unleashing the Killer App (Boston: Harvard Business School Press, 1998). Unfortunately, much of what has been written on this search has simply comprised accounts of historically successful killer apps. We think that
a rigorous study of such applications would show that they were killers because the product or service was squarely positioned on a job that a lot of people already were trying to get done—the innovation in question simply helped them get it done better, and more conveniently.

  9. The firm headed by Mr. Ulwick that we mentioned in note 3 has proprietary methods for categorizing job-defined markets and measuring their size.

  10. This information was recounted to us in a July 2000 interview with Mickey Schulhoff, who worked for over twenty years as CEO of Sony America and served for much of this time as a member of Sony Corporation’s board of directors.

  11. We must emphasize here that we have absolutely no inside information about any of the companies or products mentioned in this section, nor have we conducted any formal market research on these products or jobs. Rather, we have written this material simply to illustrate how theories that are constructed on circumstance-based categories about what products will connect with customers can bring clarity and predictability to what historically has been a hit-and-miss task in innovation. It may very well be, for example, that given RIM’s strategy of emphasizing sales to enterprises rather than individual customers, it is the corporate CIO manager who has the job to do: being sure that the firm’s knowledge workers are able to communicate and be contacted on a real-time, no-excuses basis. The same exercise would be useful if applied to this job.

  12. As this book was being written, in fact, RIM and Nokia announced a partnership through which Nokia will license RIM’s software to enable wireless e-mail on Nokia’s phones—a deal that makes sense for both firms because in many ways their products are hired to do the same job. Whether one would prefer to produce the BlackBerry that ultimately will compete against wireless phones to do this job, or whether it would be better to provide the software inside others’ wireless phones, as the new Nokia-RIM arrangement provides, is a question that the theory in chapters 5 and 6 will address.

  13. We have gone out on the end of a very long limb in making these statements, because the future has not yet happened. We have presented this analysis provocatively in order to illustrate the fundamental principle. In all probability, the makers of wireless hand-held devices will engage in a headlong rush to incorporate every competitor’s latest features on their products, leading the industry very prematurely to a situation in which products are un-differentiated, commoditized, one-size-fits-all solutions. When this happens, we urge our readers not to conclude that “Christensen and Raynor were wrong.” We would assert that although some blurring and copying of features will inevitably occur, the longer each manufacturer focuses on incorporating those features and functions that do a unique job well and the longer they position their marketing message on that unique job, the faster the suppliers of these devices will grow because they will gain share not against each other, but against other products and services that get hired to do those jobs. We would also argue that these firms will preserve their differentiability and profitability longer if they focus their improvement trajectory on a unique job. The fact that they are unlikely to do this does not disprove the principle.

  14. See, for example, Leonard, Wellsprings of Knowledge ; Eric von Hippel, The Sources of Innovation (New York: Oxford University Press, 1988); and Stefan Thomke, Experimentation Matters: Unlocking the Potential of New Technologies for Innovation (Boston: Harvard Business School Press, 2003).

  15. In concept, of course, being able to carry one small device that does everything in a briefcase or purse is something that all customers would say they want. But it is rare that there are no technological trade-offs to adding diverse functionality to a product. Software makes it less expensive to tailor a single physical platform to do a range of focused jobs. Our proposition, however, is that even in this situation, a company would do better by using one single hardware platform to market different software-defined, optimized products that are positioned on different jobs. It is likely that for a long time electronic devices that combine such a wide range of functionality in the interests of doing many jobs simultaneously—organize me, connect me, help me have fun, and so forth—are likely to end up more like a Swiss army knife: a pretty good knife, terrible scissors, a marginal bottle opener, and a crummy screwdriver. As long as the jobs that customers need to get done arise at independent points in time and space, we would expect that most customers will continue to carry multiple devices until a one-size-fitsall omnibus device can do all jobs as well as its focused competitors.

  16. The experience that Intuit had in disrupting the small business accounting software market with its QuickBooks product typifies this situation. Until the early 1990s the only available small business software had been written by accountants for accountants. Because they defined their market in terms of the product, they framed their competitors as other makers of accounting software. The vision that this framing gave them about how to get ahead of their competitors, therefore, was to engage in an arms race of sorts: Be faster adding features and functionality in the form of new reports and analyses that could be run. The industry gradually converged upon undifferentiated, one-size-fits-all products, into which everybody had appended everybody else’s features.

  Intuit’s marketers were wont to watch what jobs the customers of Intuit’s Quicken personal financial management software were trying to get done for themselves when using the product. In the course of doing this, they observed to their surprise that a large proportion of Quicken users were employing it to keep track of their small business’s finances. The job, they learned, was basically to keep track of cash. These small business owners had their fingers in every dimension of their business and did not need all of the financial reports and analyses that the prevailing software providers had cobbled into their products. Intuit launched QuickBooks at this job that small business owners needed to get done—“Just help me be sure I don’t run out of cash”—and succeeded spectacularly. Within two years the company had seized 85 percent of the market with a disruptive product that lacked most of the functionality of the competing products.

  17. Theodore Levitt has been a leading proponent of this view among those who research and write about issues in marketing. Christensen remembers that when he was an M.B.A. student he heard Ted Levitt declare, “People don’t want to buy a quarter-inch drill. They want a quarter-inch hole.” In our words, they have a job to do, and they hire something to do the job. Levitt’s best-known explanation of these principles is found in Theodore Levitt, “Marketing Myopia,” Harvard Business Review, September 1975, reprint 75507.

  18. For suggestions on how the magnitude of job-defined market segments can be measured, see Anthony W. Ulwick, “Turn Customer Input into Innovation,” Harvard Business Review, January 2002, 91–98.

  19. We are grateful to Mike Collins, founder and CEO of the Big Idea Group, for his comments that led to many of the ideas in this section. Mike reviewed an early draft of this chapter, and his thoughts were extraordinarily helpful.

  20. One reason that some (but not all) “category killer” retail formats—companies such as Home Depot and Lowe’s—have been able to disrupt established retailers so successfully is that they are organized around jobs to be done.

  21. Because many marketers inadvertently and over time tend to segment their markets along attribute-based categorizations of products and people, it is unfortunate, but not surprising, that they often do to their brands the same thing that they have done to their products. Brands often have become omnibus words that don’t do well any of the jobs that customers need to get done when they hire the brand. Because most advertisers want a brand’s meaning to be flexible enough for a range of products to be housed under its umbrella, many brands have lost their association with a job. When this happens, customers remain confused about what product to buy to get the job done when they find themselves in a particular circumstance.

  CHAPTER FOUR

  WHO ARE THE BEST CUSTOMERS

  FOR OUR PRODUCTS?

/>   Which customers should we target? Which customer base will be the most valuable foundation for future growth? Is our growth potential greatest if we pursue the largest markets? How can we predict which competitors will target which sets of customers? What sales and distribution channels will most capably embrace our product and devote the resources required to grow the market as fast as possible?

  The message of chapter 2 was that although sustaining innovations are critical to the growth of existing businesses, a disruptive strategy offers a much higher probability of success in building new-growth businesses. chapter 3’s message was that managers often segment markets along the lines for which data are available, rather than in ways that reflect the things that customers are trying to get done. Using flawed segmentation schemes, they often introduce products that customers don’t want, because they aim at a target that is irrelevant to what customers are trying to get done. This chapter addresses two questions that are closely tied to the last: Which initial customers are most likely to become the solid foundation upon which we can build a successful growth business? And how should we reach them?

  It’s relatively straightforward to find the ideal customers for a lowend disruption. They are current users of a mainstream product who seem disinterested in offers to sell them improved-performance products. They may be willing to accept improved products, but they are unwilling to pay premium prices to get them.1 The key to success with low-end disruptions is to devise a business model that can earn attractive returns at the discount prices required to win business at the low end.