Acquisition · Hidden Markov model applied to two empirical settings
Some Customers Would Rather Leave Without Saying Goodbye
Marketing Science · 2018
Summary
Overt churners were often more active before leaving, while many silent churners disengaged early and rarely canceled formally.
For your marketing
Possible founder application
Separate cancellation from quiet nonuse when measuring churn. Build retention reporting that includes product inactivity and communication response rather than treating account status as the only signal.
Reported result
What the study found
Overt churners were often more active before leaving, while many silent churners disengaged early and rarely canceled formally. The two churn groups responded differently to communications, so one retention strategy may not suit both.
Research question
What the authors investigated
How should firms distinguish customers who formally cancel from those who silently disengage?
Design
The authors develop a hidden Markov model for overt and silent churn and apply it to a daily-deal website and a performing-arts organization.
Sample and setting
Two hybrid customer-relationship settings where customers can formally terminate or simply stop interacting with the firm.
Scope and transfer
Limitations to keep in view
The two applications are specific to their services and datasets. The model's churn-state assumptions and communication responses should be validated for a SaaS product's billing and activity data.
Research sources
Marketing Science abstract and author-hosted paper; both applications and the distinction between overt and silent churn are reported in the paper.