Mobile Retention
Part of Mobile subscription and upgrade journeys
Comparing trial starts with retained paying users
Follow a mature trial cohort through first payment and later paid renewal without confusing starts, subscribers and paying users.
Track the same eligible trial cohort from trial start through its first successful paid period and a specified later paid renewal. Show the count at each stage. More trial starts can coexist with fewer later paying subscribers; a recently started trial cannot yet be judged on renewal.
Define the starting cohort
Choose the subscription product or offer, storefront, country of purchase and trial-start period. State whether the unit is a person, account or subscription.
One person may hold more than one subscription, and a plan change may appear differently across report views. Keep first-time and returning subscribers separate if that distinction matters to the decision.
Record when each trial begins and when its first paid period is due under the offer. Keep trials still running or awaiting a payment outcome in a pending group. Do not divide paid conversions this week by trial starts this week unless they belong to the same completed cohort.
Follow distinct outcomes
| Stage | Count for the starting cohort |
|---|---|
| Trial start | Eligible subscriptions that began a trial. |
| First paid period | Cohort members with a verified first paid subscription outcome. |
| Later paid renewal | Those same members with a verified renewal at the selected later paid period. |
For a custom trial-to-paid rate, divide verified first paid outcomes by trial starters whose trial and defined payment-resolution window have both finished.
For a paid-renewal rate, divide verified later renewals by members who reached the first paid period and had enough time to reach the selected renewal. Also show later paying members as a count out of the original mature trial cohort. Put the denominator and observation window beside every percentage.
Decide whether “paying” means a successful charge, a paid subscription period or paid entitlement on a stated date. These are different outcomes.
A cancelled subscription may retain access until the paid period ends, while a billing issue may remain unresolved during recovery. Keep such states separate rather than counting access alone as a later payment.
Tracking a Trial-to-Paid Cohort: Key Stages
- Trial startEligible subscriptions that began a trial (defined by product, storefront, country, and trial-start period)
- First paid periodCohort members with a verified first paid subscription outcome
- Later paid renewalSame cohort members with a verified renewal at the selected later paid period
Trial Starts vs. Retained Paying Users: Cohort Metrics
- Trial starts (cohort size)Total eligible trial starts within defined period
- First paid conversion rateVerified first paid outcomes ÷ completed trial cohort
- Paid renewal rateVerified renewals ÷ members who reached first paid period and had sufficient time to renew
Key Cohort Metrics for Subscription Retention
- Trial starts (Australia)Track per app store (Apple App Store / Google Play) using local storefront data
- First paid period success rateMeasured after full payment-resolution window closes
- Renewal rate (60–90 days post-first payment)Based on verified renewal events in App Store Connect or Google Play Console
Read store reports within their definitions
App Store Connect's subscriptions dashboard includes metrics such as Plan Starts, Conversion to Paid, Churn and Paid to Offer. Its subscription events include Free Trial Start. Check the definitions and populations for each metric before comparing it with a cohort calculation.
For Google Play, check the definitions and populations in the relevant reports before using them in a cohort comparison. Do not pool the stores' displayed percentages without checking their populations, events and report views. If the app sells subscriptions through another route, record that route separately until compatible records can be joined. An app analytics event labelled as a completed paywall does not verify a payment.
Use the sequence to investigate
If trial starts rise but first payments do not, inspect eligibility, the offer promise, the trial experience and payment outcomes. If first payments rise but later renewals do not, inspect paid-feature use, cancellation and renewal failures. These are investigation paths, not diagnoses from a chart. Compare cohorts after their relevant windows close and keep plan, market and measurement rules visible.
