Last week, Harvard historian Jill Lepore wrote a scathing critique of Harvard Business School Professor Clayton Christensen's theory of disruptive innovation. Just a few days ago, Christensen responded to the critique in an interview conducted with Business Week writer Drake Bennett. Many people have asked me about my views on this debate in the past few days, given that I spent a great deal of my career at Harvard as a student and faculty member. Therefore, here are a few thoughts.
First, let me acknowledge that I know Clay Christensen very well, and I have never met Jill Lepore. I took a doctoral seminar with Clay when I was a graduate student, and I co-taught a doctoral seminar with Clay when I was on the faculty at HBS. I have known him for twenty years. In the Business Week article, Bennett writes that colleagues and students generally think of Clay as a "generous and thoughtful and upbeat" person. I would agree. He has integrity and cares about others a great deal. Thus, I'm biased. I must admit to being taken aback by the scathing nature of the critique in this New Yorker piece. It's one thing to identify the flaws in another person's work, but the tone of the criticism caught me off-guard... particularly given that both individuals are colleagues at the same academic institution (though Harvard is a big place, and the two have apparently not met).
Does the New Yorker piece raise a good point though? Let me start by acknowledging that the term "disruptive innovation" has become a widely overused cliche. People seemingly apply to the term to every failure of large incumbent firms. They also fail to examine the multiple causes of such failures. They fixate solely on the dynamics explained in Christensen's model. We should be careful not to blame Christensen for all the mistakes of his readers though. On the other hand, all of us in academia fall into the trap, at times, of taking our favorite hammer (our preferred frameworks and models) and beginning to see every situation in the world as a nail (a problem or situation that can be explained by our theory). Christensen has applied his framework quite broadly, and perhaps at times, he has made mistakes by extending it to situations in which it does not quite fit (the iPhone perhaps).
I see three other challenges with the model of disruptive innovation. First, I think it works beautifully as a descriptive model helping us explain past situations. It can be more problematic when applied as a predictive framework. Many academic frameworks have this weakness. Markets and organizations are fundamentally complex, and we would be hard-pressed to find frameworks that have high predictive accuracy in a wide range of contexts. However, many people crave simplicity. We demand that academics provide findings that can be generalized across many settings. We don't like when academics say that the answer "depends" on the situation. Second, incumbent firms can certainly make crucial mistakes if they fail to react in a timely and appropriate fashion to a disruptive innovation. On the other hand, many firms these days have made huge bets as they have tried to counter what they viewed as disruptive threats. Many big errors have been made in this regard. In short, there are two types of errors: we probably should pay as much attention to the bad bets made in over-reacting to apparent disruption as we do to the incumbent firms who fail to react swiftly to disruption and get toppled from their perch atop an industry. Finally, we have to be careful about the prescription that incumbent firms should separate units that are aiming to cope with a disruptive threat. It may indeed be the right short term strategy, so that the core business unit does not "eat its young" so to speak. However, in the long run, a company has to think about how the whole is worth more than the sum of the parts. Competitive advantage comes from the alignment or fit with an integrated system of activities. If the new unit is "too separate," then the firm may find itself with two parts that are not mutually enhancing. In fact, there may be fundamental contradictions that make it quite difficult for the two units to co-exist in the same organization.
