DisruptiveInnovation.Org - Latest Discussions

Anomalies: Improving Good Theory

David Sundahl, 9/8/2016


There is another important consequence of the power of anomaly: all good theories must be susceptible to anomalies, otherwise there is no progress. Correlation is like a sliding scale of confidence--it’s akin to an analog signal that we receive with varying degrees of fidelity. Consequently, we can seldom state definitively if some event is an anomaly or merely “noise.” On the other hand, a causal model is more like a digital signal—it gives us discrete “yes” or “no” indicators for anomalies. As we mention in the first chapter, the causal approach to manufacturing created clear go/no-go signals that either kept work moving or initiated an investigation into a defect. Without the clear binary signals, sub-par work would inevitably sneak through.

In building theories in management—and other social science disciplines—the causal model enables a clear yes/no indication for anomalies. Anomalies, as we discovered, are the mechanism that drives improvement in theories. Disruption has benefited greatly from anomalies. But what happens when good work stops at the level of correlation?

There are few more influential gurus than Jim Collins whose books Built to Last and Good to Great have dominated best-seller lists for years. Yet his research is a lightning rod for criticism – from some very smart people. This is our sense of why.

In Good to Great, Collins outlines seven characteristics that he and his team found common to all firms that had a particularly successful and durable performance. One of the characteristics of the CEOs in each of the good-to-great companies is that they were home-grown, humble and self-deprecating. He asserts that if such men and women were positioned atop other companies, they too would go from good to great. But how about Steve Jobs? He was not humble or self-deprecating. Yet Apple prospered under his leadership and has proven to be quite resilient without him at the helm.

Can the theory account for this? If we accept that Apple at least appears to be an anomaly, what can be done?

Collins’ theory is at risk of being skewered on the horns of a dilemma. On the one hand, he can ensure that the criteria for selecting good-to-great firms excludes Apple. But excluding a firm that everyone would agree has not only done well, but has come out of a decade-long slump to have the largest market cap in history, seems like a failure of the theory. On the other hand, if Collins admits that Apple is an anomaly, he’s got a serious problem. Are there exceptions to the theory? Or are some elements more important than other? Collins seems to think not:

No single finding by itself makes a great organization; you need to have them all working together as an integrated set. Then work sequentially through the framework, starting with “first who” and moving through all the major components. Meanwhile, work continuously on your own development toward Level 5 leadership. I have laid out this book in a sequence consistent with what we observed in the companies; the very structure of the book is a road map.

Collins’ theory is induced from all of the firms that were good to great at the time his study was conducted. His study concluded that all and only the firms that went from good to great had all of the components together. It has not fared well either in predicting the future of these firms or in predicting which firms will go from good to great.

Our purpose is not to take pot shots at an incredibly smart guy like Jim Collins. What we wish to show is that research projects that stop at correlation cannot turn into longer programs of research into different areas and applications. When we stop short of the hard work of understanding the underlying causality of the phenomena we study, we put ourselves into a one-and-done situation. There is no clear way for research that stops short of causality to advance beyond “the literature.”

The hard work of theory-building does not stop with the articulation of a model. A causal model, however, makes it possible to accommodate anomalies that clarify the boundaries and conditional application of a theory—whether to apply the “invisible hand” or the “visible hand,” for instance. The improvability of processes in the quality revolution in manufacturing was essential. Likewise, if a revolution in the marketplace of ideas is to happen, ideas must be amenable to improvement through anomalies.