Exploring Peer Effects Associated with User Churn in a Socially Connected Business

Abstract

In a socially connected business, the churn of a user can lead to reduced consumption of her peers, which can potentially spread to other users in the network. We investigate these negative peer effects associated with churn of a user, and how they vary over time and with respect to the position of the churner in the network. We use data from an upscale gym with information on exogenous churns of users that allows us to identify peer effects. We estimate a structural model of users’ plan choice and usage decisions, that allows for bidirectionality and asymmetry in peer effects, as well as direct and indirect peer effects. Our key results are: (i) Unlike the churn of a low and a medium centrality user, the churn of a high centrality user has a strong negative impact on usage and subscriptions of other users in the network which increases over time. This suggests that if the negative effect from a high centrality churner is not stopped early on, it can spread throughout the network. (ii) To stop the negative impact of churn, a firm should target promotions to other users who have a high degree of centrality. This suggests that the conventional wisdom that firms should target promotions to users with a highest risk of churn, or highest upliftment in purchase probability/CLV may not work in the context of socially connected business with strong peer effects.

Keywords: Peer effect, Social Network, Customer Relationship Management