The Pitfalls of Superstitious Learning
Posted on Harvard Business Review: July 19, 2011 12:12 PM
In business research, we often build up theories about what works and what doesn’t work based on evidence that’s correlated, but not causal. We pick an outcome—successful performance, say—and incorrectly identifying what caused that outcome. This is a major flaw in our work, as my colleague Phil Rosenzweig has pointed out in his work on the “halo effect.”
In organizations, there’s a similar phenomenon, which I’ll call “superstitious learning.” Superstitious learning takes place when the connection between the cause of an action and the outcomes experienced aren’t clear, or are misattributed. For example, consider a manager in a company that fortuitously entered a growing market just at the right moment. This manager appears successful and is rewarded with several promotions into the senior ranks. Obviously, the guy must know what he’s doing, because he has always experienced success, right? Actually, no—one of the least fair realities of modern business is that it is entirely possible to have good outcomes without being particularly skillful (why else would Dilbert be so popular?). Often, the only antidote to everyone thinking the person is golden is to have some kind of setback take place. Let this manager encounter a problem, and his or her true abilities will emerge to be tested.
For instance, a major retailer prided itself on the steady, profitable growth of its music CD business over a long period of time. It was taking share from other players and had become the go-to destination for many people interested in popular music. What was the problem? No one was looking at CD sales changes relative to total music sales changes. What was happening was that digital downloads were exploding in their own growth; meanwhile revenues for the whole industry were shrinking . The real story was not one of success. Rather, it was one of grabbing share from a market that was in deep decline. Indeed, Forrester Research predicts that the industry as a whole (not just the physical CD part of it) will be in slow decline for some time to come. It was only when managers started to look at the music category as a whole that they realized their “major success” was actually a dead end.
The same principle applies to career development: be wary of those who have never experienced a failure, as they are likely to be unprepared when the unexpected happens.
A better way to attribute success to the right causes is to lay out your “theory of the case” and then test it against reality, before outcomes are known. Then, if you can, test the theory under different situations—what the academics call boundary conditions— to see if it holds across those. When you have a practice that seems to produce replicable results, that is worth beginning to believe in.
Related Harvard Business Review Links:
Subscribe to Harvard Business Review
Sign up for Management Tip of the Day free email newsletter
Copyright © 2012 Harvard Business School Publishing. All rights reserved. Harvard Business Publishing is an affiliate of Harvard Business School.