When to turn rewarded ads off for payers (and how to measure the loss)
Operator brief: treat payer rewarded suppression as a measurable IAA opportunity-cost test using AdMob reward callbacks and MMP ad-revenue events — not a vibe check.
Turning rewarded ads off for payers is a UX experiment with a cash cost. Measure the IAA opportunity cost with the same event discipline you use for IAP, or the test will “win” on sentiment while ROAS quietly worsens.
What the docs define
AdMob rewarded ads require an OnUserEarnedRewardListener so reward grant is an explicit callback — use that (or your mediation equivalent) as the exposure flag for “saw rewarded”. AppsFlyer partner ad-revenue sharing documents af_ad_revenue as the event class for mediation revenue shared to UA partners. Neither page publishes a “payers should never see rewarded” rule.
Measurement brief
- Define payer (first purchase timestamp; store-validated).
- Randomise eligible payers into rewarded-on vs rewarded-off.
- Primary readout: IAA revenue / payer / day via ad-revenue events; secondary: IAP revenue / payer / day; guardrail: D7 retention / refund rate.
- Stop if IAA loss exceeds a pre-registered contribution threshold after store fees — do not stop on CPI alone.
Facts versus limits
| Field | Evidence |
|---|---|
| Reward callback | AdMob rewarded docs |
| Ad-revenue event class | AppsFlyer `af_ad_revenue` partner article |
| Recommended payer policy | Not stated on cited pages |
