Jae Lee, Ph.D. Student in Marketing, The Wharton School, and David Bell, Xinmei Zhang and Yongge Dai Professor of Marketing, The Wharton School
Abstract: Internet retail sales are impeded when consumers have difficulty acquiring information about non-digital product attributes. Firms therefore employ creative (and sometimes costly) methods such as two-way free shipping to help consumers resolve the problem of incomplete consumer knowledge on non-digital attributes. We argue that consumers and firms can benefit from social learning processes that occur “naturally” and emanate from neighborhood characteristics. We use trial data from Bonobos.com, a leading US online retailer for men’s fashion apparel, to demonstrate that social learning occurs. We further show that geographic variation in “neighborhood social capital” moderates this process and explains geographic variation in online sales. We find that social learning reduces pre-trial consumer disutility by reducing bias in consumers’ initial evaluation of non-digital attributes. The estimates imply that up to 50% of all trials in the first three and a half years of Bonobos’ sales were influenced by social learning. In addition, social learning is more efficacious in neighborhoods with more social capital. Our estimates imply that a one-standard deviation reduction in the social capital stock of all zip codes would slow trials and reduce them by about 4%.