Winter 2008 / No. 74
Unintentional Damage
By Gil Lowerre
Market research is a delicate art. Raw data from customers must be tempered first by interpretation and then by contextual knowledge. Only after both steps does research become information.
As an industry, we are the beneficiaries of active research programs by consulting houses (such as ours), research trade associations, and carriers. And most studies include the author’s interpretations and conclusions. But authors do not always provide the full context that enables readers to use the research as action-oriented information. A classic example includes “intentional” data, a notoriously unreliable category of market research. Intentional data is created when consumers are asked under what conditions they would change their behavior. “Under which of the following conditions would you be more likely to buy...?” Consumers usually have no problem answering these types of questions, but far too often, actual behavior has no relationship to their stated intentions. Where there is no alternative (in purchasing a product that has not yet been manufactured, for example), intentional data is the best that can be developed, and the researcher’s responsibility is to simply warn the reader that intent often is different than behavior. But where empirical data is available, intent is not only a bad substitute, it can be downright misleading. Don’t ask whether consumers would prefer buying a product in a red or blue box. Conduct a controlled experiment offering each and see which they actually pick.
A recent article quoting a research trade association study said, “When choosing voluntary benefits, 65% of businesses prefer to deal with carriers that are providing at least some of their existing, employer-paid benefits. Only 15% chose voluntary benefit carriers they were not already dealing with for company-paid benefits.” The key word here is “prefer,” although the article did not explain that this is intentional data and may or may not have any relationship to actual behavior. Those statements are not information; they are dangerous data.
In this case, there is no relationship between the data and behavior. The fact is that carriers that do not offer employer-paid benefits made well over 50 percent of the new voluntary sales last year. And among brokers selling employer-paid benefits, less than 20 percent felt that it was very or extremely important to use the same carrier that provided group benefits.
If they had been asked, employers would probably have preferred to buy their voluntary coverages through a close relative, for free, or from a carrier that never makes mistakes. But that’s not information either. Until researchers do a better job of providing appropriate context to help their audiences understand (or ignore) their findings, bad information will continue to lead to bad decision-making. Caveat emptor.
Check out a sample of Eastbridge’s published research on our website at www.eastbridge.com.
|
