Introducing a new product is essentially an exercise in persuading people to change their behavior. hbr_130x130Many companies try to tackle this challenge by making the functional benefits of the new seem so much more compelling than the old. But this approach rarely works. After all, how many of us as children enjoyed eating our vegetables just because our moms said they were better for us than desert?

But how much quicker would your attitude have adjusted if your best friend had dared you to eat them? Or if eating broccoli had suddenly become the newest craze in your fifth-grade class?

A new form of social data that harnesses the power of peer pressure is emerging as a potentially powerful way to change behavior and spur the growth of new categories of products. It works because peer pressure data goes beyond demonstrating the functions of a product to satisfy deeply powerful emotional or social needs we may not even realize we have.

Many people are aware by now, for instance, that utility companies across the U.S. have been taking advantage of peer pressure to reduce energy consumption by including charts in electricity bills showing how energy efficient you are compared to your neighbors. Companies such as Opower and My Energy have developed these data systems, and they can now point to studies that show the combination of data and social pressure reduces home energy use.

Presented in the right way, peer data can also be effective in changing consumer financial habits, such as encouraging a higher savings rate. Voya Financial (formerly ING U.S.) has one such application, called CompareMe, that promises to increase retirement savings rates by drawing on survey data to compare your rate to that of people with similar incomes, and to the average in your state. Putnam Investments has developed a “How Do I Compare?” feature for its 401K clients. Academic studies are showing that the peer effect works in motivating people to save more.

Read the rest at Harvard Business Review

Robyn M. Bolton is a partner at Innosight.