The Impact of User-Generated Content on Sales: A Randomized Field Experiment

This study examines the causal relationship between popularity information and purchasing behavior in an online store. In a randomized field experiment we exogenously manipulated the visibility of user-generated content similar to Google’s +1s or Facebook’s Likes. Displaying the number of people who “Like” a product caused a +12.97% sales increase (13,883.74 EUR in the treatment group; 12,289.46 EUR in the control group). We find that popularity information influences shopping behavior significantly if it is displayed in the consumers’ leisure time. This result is consistent with observational learning. For well-planned and goal-oriented purchases, knowing the preferences of others is of little importance. This information is more valuable on not so goal-oriented and, hence, more time-consuming shopping trips where consumers are searching for interesting, new products. The results also suggest that Likes have a significant monetary value, but without orthogonal variation, the valuation of Likes can easily be overestimated (by a factor of 2.26 in our sample).