Geographical Pattern of Online Word-of-Mouth: How Offline Environment Influences Online Sharing

Abstract

We study the geographical pattern of online word-of-mouth (WOM). Leveraging a unique research design (a 'quietly' launched WOM program), we measure online WOM of over 200,000 customers across the U.S. by tracking their WOM referrals on a national e-commerce platform. In parallel, we measure the offline environment of each customer (e.g. social environment such as neighbor interactions and mobility, and local environment such as demographics and retailer access) by combining a range of datasets from proprietary and public data sources (Facebook, Census, ESRI and SCCBS). While digital technologies have enabled customers to share at any location and to any destination, we find that the offline (social) environment of a customer still matters for the generation and movement of online WOM. First, a customer's offline social environment significantly explains online referral propensity, whereas local characteristics that are known to explain customer online behavior are not explanatory. Second, customer online referrals are either bounded locally (30% within the same zipcode) or point at destinations that are socially or demographically similar to the customer's local environment. Finally, there is a significant interaction between advertising channels and offline environment in creating WOM. Our findings on the geographical pattern of online WOM can guide firms' design of location-based interventions such as the allocation of advertising budget across geographical areas.