Cohort quality: D1 retention vs D7 payer rate divergence lab
High D1 with low payers is a creative-market mismatch signal. Illustrative divergence index with retention definition context.
A creative that lifts D1 retention while D7 payer rate falls is often a promise mismatch, not a UA win. Measure both on the same cohort before scaling.
Definitions
GameAnalytics on retention (updated 23 April 2026) defines retention as the share of players who installed on a day and returned on a later day, and warns that shorthand curves such as 40/20/10 are not universal across genres. Purchase and revenue events remain operator KPIs in the MMP (AppsFlyer in-app events overview).
Synthetic divergence
| Metric | Value | Status |
|---|---|---|
| Installs | 10,000 | illustrative |
| D1 retained | 3,200 (32%) | illustrative |
| D7 unique payers | 180 (1.8%) | illustrative |
| Payers per 100 D1 retained | 5.625 | calculated |
Compare payers_per_100_d1 across creatives on the same geo and bid type. A rising D1 with a falling ratio is the mismatch signal.
Facts vs limits
| Fact | Limit |
|---|---|
| Retention is return-rate by definition in the cited GA piece | No industry “good” payer-per-D1 average is claimed here |
| Same-cohort comparison is the method | SKAN-only iOS cells may lack payer granularity |
CSV.
