Stop Listening to Your Customers

Every business wants to know what influences and persuades their customers. To get this information they ask. But there is a problem with asking people what they want: most people don't know. Only they won't tell you they don't know. They'll give you an answer. Yet the answers won't likely be any good. Steve Martin, co-author of Yes!: 50 Scientifically Proven Ways to Be Persuasive, explains why.

The great majority of the decisions we make in our information-overloaded, distraction-heavy lives are made outside our conscious awareness, driven more by contexts than cognitions. As a result, asking someone to pinpoint what will influence them in the future is a bit like saying, “tell me how you will behave in the future when you are not thinking about what I have just asked you about?”

Behavioural scientists Wes Schultz and Robert Cialdini provide compelling evidence of why asking people to predict what will influence their future decisions and behaviors is so often ineffective. In one set of studies, they asked several hundred homeowners in California to predict which of four messages would be most successful at persuading them to take steps to conserve energy and reduce their overall consumption. The four messages were 1) conserving energy helps the environment; 2) conserving energy protects future societies; 3) conserving energy saves you money; 4) many of your neighbors are already conserving energy.

Those shown the message about what their neighbors were doing rated it as the least likely to influence their behaviours. Yet when meter readings were taken, the researchers discovered that this was the most effective message when it came to changing behavior even though this same message was rejected by most as having any sway. Even though most will deny its effect, our desire to keep up with the Joneses is both universal and automatic. For example, recent studies have shown that compared to the usual approach of threatening those who fail to pay their taxes on time with fines, it is far more effective to inform them that the majority of people in their neighborhood already have paid. By doing so, governments can realize many more millions in revenues.

Not only are we pretty poor at recognizing what will influence our future behavior, we're not that great at recognizing what persuaded us after the event either. In one well-known study conducted at a busy New York City subway station, after counting the percentage of commuters who donated to a street musician as they walked past him, researchers made one small change to the situation: Immediately before an approaching commuter reached the musician, another person (who was in on the act) would drop a few coins into the musician's hat. The result? An eight-fold increase in donations. When interviewed afterwards, those who donated universally failed to attribute their actions to the fact they had seen someone else give money first, preferring instead to provide alternate (and incorrect) justification for their actions. “I liked the song he was playing”; “I'm a generous person”; and “I felt sorry for the guy.”

The message: ask fewer questions and instead run small experiments and observe the results. Who, for example, would have guessed that alerting people to the fact that many previous guests reused their towels and linens would significantly increase reuse. Stop asking and start experimenting.