The Innovator's Dilemma

The claim

Clayton Christensen was a professor at Harvard Business School who had worked as a consultant and run a manufacturing company before turning to research. His doctoral work studied the computer disk drive industry, where technology changed so fast that, as he put it, it was like studying fruit flies in genetics: many generations could be observed in a short time. The Innovator’s Dilemma, published in 1997, grew out of that research.

Its claim is a paradox. Well-run companies, which do everything management textbooks recommend, listening to customers, investing in the most profitable products, are the ones most likely to be overturned by a particular kind of technological change. Christensen distinguishes two kinds of innovation. Sustaining innovations improve products along the dimensions that existing customers value, and incumbents nearly always win them. Disruptive innovations offer a different set of attributes, typically cheaper, simpler and more convenient but worse on the measures that mainstream customers care about. They take root in small or new markets, improve over time and eventually become good enough to displace the incumbents.

The dilemma is that the rational decision, at every step, is to ignore the disruption. By the time it is clearly a threat, it is usually too late.

How the case is built

The book’s core evidence is the history of disk drives. Each new generation was physically smaller: from fourteen-inch drives for mainframes to eight-inch drives for minicomputers, five-and-a-quarter-inch drives for desktop computers, and three-and-a-half-inch drives for portable machines. Each smaller drive initially offered less capacity than mainstream customers needed, so leading manufacturers, listening to those customers, did not pursue it. New entrants did, selling to new markets that valued the smaller size. As capacity improved, the new drives became good enough for the mainstream, and most of the leaders of each generation failed to make the transition.

Christensen then explains why. Companies operate within value networks: their customers, suppliers and cost structures shape what they can profitably do. Resources are allocated by processes that favour projects with large markets and high margins, so a small, low-margin disruptive market cannot win the internal competition for investment. A large company needs large new markets to keep growing, and disruptive markets start small. And because the technology at first performs worse, existing customers tell the company not to bother.

He extends the analysis to other industries: mechanical excavators, where hydraulic machines started in small markets before overtaking cable-operated ones; steel, where minimills began with low-quality products and moved up; and retailing.

His prescription is to give responsibility for a disruptive technology to an organisation whose customers want it, often a separate unit small enough to get excited about small wins, with a cost structure that can profit from low margins, and to plan for learning and discovery rather than precise forecasts.

What it takes for granted

The theory assumes that one can identify a disruptive innovation in advance, or at least early, by its characteristics. Christensen and his co-authors later warned that the term was being used too loosely, applied to any new entrant that succeeded, and they clarified in 2015 that disruption describes a specific process: entry at the low end or in a new market, followed by movement upmarket.

It also assumes that the disk drive pattern generalises. In 2014 the historian Jill Lepore argued in The New Yorker that many of the book’s cases were chosen selectively and that several supposedly disrupted companies had in fact prospered. Christensen responded that her reading misunderstood the theory. Later empirical studies have found that some cases fit the theory well and others less so. The debate centred less on whether disruption happens than on how often, and how predictably.

Since it was published

The book became one of the most influential business books ever published. Its vocabulary passed into everyday language, especially in technology, where “disruption” became a goal rather than a description. Christensen extended his ideas in The Innovator’s Solution (2003), on how to create disruptive growth, and in work on healthcare and education. His later theory of jobs to be done, set out in Competing Against Luck (2016), asked what customers hire a product to do. He died in January 2020. The site’s idea page Why Good Management Fails develops the theory and its debates, and essays such as Software You Never Own and The Number That Became a Person show it at work.

What it asks of someone deciding

The book asks what your best customers are telling you to ignore. The products they do not want may be the ones that will replace yours.

It asks whether your processes can fund something small. If every investment must promise a large market and high margins, the next disruptive technology will lose the internal competition every time. A separate unit, with its own customers and cost structure, may be the only way.

And it asks whether you can tolerate being wrong about forecasts. Disruptive markets cannot be predicted precisely, because they do not yet exist. Planning for discovery, with small bets and fast learning, suits them better than detailed business plans.