-
-
Marketplace
-
Channel Resources
Articles from this Site
-
One brand, one Web site! DM Review is now the home of all the content you're used to at BIReview.com and much more. If you are registered at BIReview.com, you're already registered at DM Review. If not, take just a moment to sign up for all the free services we have for you at the new DMReview.com.
Where's My Halo?
Back in college, my statistics and probability professor explained the difference between correlation and causation: "Anyone who eats pickles will surely die," he said, while merrily munching on a gherkin. Professor Phil Rosenzweig, who teaches at the famous IMD business school in Switzerland, has written a book called "The Halo Effect" that similarly reminds us to beware of seductive correlative and causational thinking.The "Halo Effect" is a term coined around 1920 by American psychologist Edward Thorndike, who found that most people are born with cognitive bias; they tend to infer specific traits on the basis of general impressions. A person who wears glasses must be studious (from reading so much). Automobile companies count on this kind of bias when marketing cars. When was the last time you saw someone who didn't look beautiful or handsome in a car ad? The frenzy over the iPhone is another example, given the number of people who seem to believe that owning one will make them appear "with it."
Rosenzweig saw the Halo Effect at work while reading management books such as Jim Collins' "Good to Great: Why Some Companies Make the Leap... and Others Don't," and Tom Peters' "In Search of Excellence." We're all familiar with the school of writing that associates certain leadership traits with corporate success. By that standard, a successful company is one with strong leadership, and a poorly performing business is one with weak leadership. Rosenzweig asked aloud if this meant that all companies that perform poorly have weak leaders; and conversely that no successful company has ever had weak leadership. How would this explain Chainsaw Al Dunlap? Dunlap had earned a reputation for strong, aggressive leadership, and was applauded by stockholders for his appointment as CEO at the Sunbeam-Oster Corporation. Dunlap, in large part by reputation, pushed Sunbeam's share price from $12 to $53, but was unceremoniously fired four months later when the stock dropped to $11. The chainsaw had never left his side, so something else must have happened.
Rosenzweig says linking success say with strong leadership creates "introspective attribution," that makes it very difficult to measure leadership independently of an eventual outcome. To avoid this problem, he says, business must start with a fundamental question: Can a company take a practice that has been associated with high corporate performance, apply it elsewhere, and get the same results? As he points out, this is a highly unlikely outcome, if for no other reason that the context of a successful practice is very difficult to find duplicated in the real world.
I saw this myself a few years ago when I was involved in assessing large Department of Defense programs for systemic sources of program risk. The lack of communication between government officials and contractors - and poor communications between independent contractors on the same team - have long been seen as fundamental problems. To remedy the situation, the DoD instituted an integrated product team (IPT) policy, based on highly-publicized reports that commercial companies such as Boeing, and some DoD programs like the F-18 Super Hornet project, had reported that IPTs helped reduce communication problems.
Yet when we did our assessments a few years after the policy was mandated, IPTs kept coming up as a major source of the problems afflicting defense programs. As one senior manager told us, "IPTs are a great way to take a long time to make a bad decision." Some programs had gone wild with IPTs and created dozens of them. Other programs created IPTs that turned into debating societies. Still other programs weren't given any decision authority, and became faux IPTs. What the DoD had forgotten was that a policy that was successful for one program wasn't likely be successful for another unless the basic operating contexts were very, very similar. Our report wound up listing IPTs as a source of communication risk.
Rosenzweig sees other types of cognitive bias at work, such as "Survivor Bias." For example, a very rich man wants to change his stockbroker although his investments are making good returns. The investor believes he would have become even richer had he used one of the top-three stockbrokers listed in a famous business magazine. The problem is, of course, that someone has to be the "best" stock picker in any given year. Did that person get there by being good or by being lucky? The inside joke on Wall Street is that there are always three successful stock pickers - two who got there by luck, and one who knew what they were doing. The trouble is always figuring out who is who.
Rosenzweig cautions that we must be very, very careful when we read about any "secret sauces" for corporate success. If you read a story about a corporation using business intelligence to gain a competitive advantage, was it because of BI or was it really something else? Perhaps a better question would be, given all companies that have seen their performance go up or down in the past couple of years, how many have used business intelligence? Next, of that sample pool, how many had the same level of experience in implementing BI, and were they using the same type of application system? Were they all competing in the same market space? Until you can separate the independent and dependent variables of corporate performance, you are going to have a hard time linking improved performance to the use of BI.
Another delusion common to people in the BI world seems to be that more data means better data. It is a very tempting trap to fall into. Rosenzweig suggests something that few companies want to hear, because compared to diet and exercise, taking a pill to lose weight sounds like a lot more fun. The mistake is to make attributions of success (or failure) based solely on outcomes. Invest in your company's critical thinking and decision processes. Work to understand the risks and uncertainties that drive your business. Finally, be realistic. Success formulas don't exist, so quit looking.
Robert Charette is president of risk management consultancy ITABHI Corp. and director of the Cutler Consortium's ERM & Governance practice. He can be reached at charette@itabhi.com.
For more information on related topics, visit the following channels:
