Subscriptions, churn and the right to cancel
Customer retention and churn appear on 33 recent papers and subscription pricing and models on 10, across the AI, operations research and finance journals. Subscriptions now cover software, cars, media and razors, and regulators are forcing easy cancellation. How much of subscription revenue depends on people not cancelling, and what happens when they can?
What prompted this#
Since October 2025, 52 papers carry a subscription, churn or customer-retention keyword, spread across AI (13), operations research (12), finance (9), economics (9) and management (7): "customer retention" on 18, "customer churn" on 8, "customer churn prediction" on 7, "subscription pricing" on 6, "subscription models" on 4. Most churn papers are prediction exercises; the economics of inertia is rarely their subject.
Questions#
- When "click to cancel" rules took effect, how much did cancellations rise, and did firms respond with lower prices or more retention offers?
- How much of subscription revenue comes from subscribers who no longer use the service, by category?
- Do firms that make cancelling hard lose customers' later purchases and referrals?
Data that could help#
Transaction panels with recurring charges and usage, the dates of cancellation rules by jurisdiction, and firm disclosures of churn rates.
Next step#
Use bank transaction data to estimate the share of recurring charges with no matching usage and compare cancellation rates before and after a click-to-cancel rule.