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

For your marketing

Possible founder application

Where product use is collaborative, consider whether a departing or inactive central user affects teammates. Validate network effects and choose retention targets using both direct and downstream value.

Reported result

What the study found

Churn by a high-centrality member caused a stronger negative cascade in peers' usage and subscriptions than churn by a low- or medium-centrality member. Short-term targeting may prioritize lower-centrality churners, while longer-term retention can favor high-centrality users because of network spillovers.

Research question

What the authors investigated

How does one user's churn affect usage and retention among connected peers, and whom should firms target to prevent network cascades?

Design

The authors use data from an upscale gym, including exogenous customer churn, and estimate a structural model of plan choice, usage and dynamic peer effects.

Sample and setting

Members of one upscale gym with observed social-network ties, plan choices and usage.

Scope and transfer

Limitations to keep in view

One gym and its observed social network; peer effects may differ in software teams, buyer organizations and products with less visible usage. The targeting tradeoff depends on campaign and relationship costs.

Research sources

Marketing Science publisher abstract and issue metadata; setting, model and centrality results are reported by the journal.