Marketing Papers

Document Type

Technical Report

Date of this Version


Publication Source

Marketing Science





Start Page


Last Page





Social learning can occur when information is transferred from existing customers to potential customers. It is especially important when the information that is conveyed pertains to experience attributes, i.e., attributes of products that cannot be fully verified prior to the first purchase. Experience attributes are prevalent and salient when consumers shop through catalogs, on home shopping networks, and over the Internet. Firms therefore employ creative and sometimes costly methods to help consumers resolve uncertainty; we argue that uncertainty can be partially resolved through social learning processes that occur naturally and emanate from local neighborhood characteristics. Using data from Bonobos, a leading U.S. online fashion retailer, we find not only that local social learning facilitates customer trial but also that the effect is economically important because about half of all trials were partially attributable to it. Merging data from the Social Capital Community Benchmark Survey, we find that neighborhood social capital, i.e., the propensity for neighbors to trust each other and communicate with each other, enhances the social learning process and makes it more efficient. Social capital does not operate on trials directly; rather, it improves the learning process and therefore indirectly drives sales when what is communicated is favorable.

Copyright/Permission Statement

Originally published in Marketing Science © 2013 INFORMS

This is a pre-publication version. The final version is available at


Bayesian learning, experience attributes, Poisson model, social capital, social learning



Date Posted: 15 June 2018

This document has been peer reviewed.