How We Mispredict Preferences
Barasz, Kate; Kim, Tami; John, Leslie K.
Publisher: American Marketing Association
Original document: The Role of (Dis)similarity in (Mis)predicting Others' Preferences
Do you like both classical music and ska? How about greasy pizza and Michelin-starred cuisine? You probably recognize that you yourself can appreciate two very different things simultaneously... but you'd be hard-pressed to find a company trying to sell you both. Few marketing professionals seem to acknowledge that such dissimilar tastes could exist in one target customer.
Even in our daily life, studies show that people expect others to like products that are similar and to dislike dissimilar things. Yet, human beings do like dissimilar things, and ignoring this means missing out on the chance to offer customers a greater variety of choices.
Listen in as professor Kate Barasz, who joined IESE's faculty in September 2016, briefly describes her research into the role similarity -- and dissimilarity -- plays in (mis)predicting others' preferences.
Methodology, Very Briefly
The research, published in the Journal of Marketing Research, is based on a set of five experiments carried out by Barasz, then a doctoral candidate at Harvard Business School, in collaboration with fellow doctoral candidate Tami Kim and Harvard professor Leslie K. John. Each experiment evaluated how people inferred information about others' preferences based on their choices.
In the first study, for example, the co-authors asked 205 participants to guess what kind of vacation a hypothetical Facebook friend liked after reading a status update. Those who read "headed to a lake!" were much more likely to guess that the "friend" didn't like city vacations than those who didn't know where he was going. The other studies looked at dissimilar movie genres, musical genres and products. Even when a monetary reward was at stake (as in one of the experiments, which looked at highly-rated dissimilar movies), participants tended to acknowledge that they could choose dissimilar things, yet failed to predict these preferences in others.