Retention cohort recipes
The Retention agent slices your subscribers into cohorts and runs recipes against them - reusable patterns that surface where money is leaking and what to do about it. Here are the cohorts it builds and the recipes worth knowing.
Cohorts AppApex builds
- Trial-to-paid, split by trial length and paywall variant
- Day 1, day 7, and day 30 retention
- Retention by acquisition channel
- Retention by price point and billing period
Recipes worth running
- 1Trial length test
Compares conversion and downstream retention across trial lengths to find the length that maximizes paid LTV, not just trial starts.
- 2Onboarding drop-off
Finds the step where trial users go quiet before ever hitting the paywall, then recommends where to move the paywall or add value.
- 3Win-back for lapsed
Identifies recently churned cohorts with high prior engagement and proposes a targeted win-back offer.
- 4Billing retry and grace period
Surfaces involuntary churn from failed payments and recommends grace-period and retry settings to recover it.
Wiring a recipe to action
Recipe findings arrive in your Inbox as recommendations with the cohort, the estimated recoverable revenue, and the proposed change. Eligible action classes can be handed to Autopilot so the agent ships them inside the guardrails you set.
Recipes need enough conversions per cohort to clear the significance gate. Small apps may see a cohort marked low-confidence until volume builds - that is the agent refusing to recommend on noise.