- Stay Connected
- Alumni Benefits
- Career Management
- Support the School
- Make a Gift
- Where to Give
- Annual Giving
- Major Giving
- Planned Giving
- Corporate Giving
- How to Give
- Report to Investors
- Contact Us
- Alumni Clubs
Have you ever taken a friend's recommendation for a book, CD or restaurant, only to regret it? That's because we vastly overestimate how well our friends know us, according to new research by Professor Gita Johar of the Marketing Division.
Johar, working with Andrew Gershoff of the University of Michigan, conducted studies in which participants were asked about their favorite types of movies. They were also asked if their friends would be able to name these types. When their friends were put to the test, their responses were often wildly off base, Johar found. In fact, when it comes to knowledge of our tastes and preferences, close friends are no better than mere acquaintances.
However, most of us tend to give our friends the benefit of the doubt, particularly in close relationships. The reason? The belief that our friends know us well protects our trust in the friendship and helps maintain our sense of our own identity. “We want to believe that we are important to others, particularly to others we care about,” the researchers say.
These findings are important to companies that are trying to understand how customers rely on “recommendation agents,” or people who make decisions on their behalf. It is an area of research that is receiving more attention in the world of online commerce, since many Web sites use recommendations as a way to encourage purchases. Typically, sites base their suggestions on searches by other customers, product reviews by users of the site or a customer’s purchasing history. Most sites use relatively simple algorithms—for example, if a customer bought Johnny Cash’s album At Folsom Prison, the site may suggest the soundtrack to the film Walk the Line. “There are a lot of recommendations that companies and Web sites give you,” says Johar. “The question is really to what extent do people trust recommendations given by others.”
The answer, she has found, is that people want specific, personalized suggestions from friends whom they believe, correctly or not, know them well. Therefore, she suggests companies create networks that allow friends to exchange recommendations. “I would assume a person within my network knows me really well, and would act on that recommendation more than I’d act on a recommendation from a stranger,” Johar says.
Some companies have taken initial steps in that direction. Amazon.com’s Profile option allows customers to make their recent purchases, wish lists and favorites viewable to their friends. Others are considering ways to build word-of-mouth through social networking Web sites like MySpace, Friendster or Facebook. In the off-line world, understanding this aspect of consumer behavior could benefit almost any business in which relationships are formed between customers and service providers.
Taking into account how little people know about their friends’ likes and dislikes, some consumers may wonder if they should seek recommendations elsewhere. “From a normative standpoint, you’d be better off not trusting your friends as much, because they don’t know you as well as you think,” Johar says. But the nature of friendship motivates us to like our friends’ suggestions, even when they prove wrong, and this, Johar concludes, keeps us coming back for more.