D1 Retention Is a Promise-Keeping Metric
When one scaling creative made the wrong promise, blended D1 drops — not because the game broke, but because one cohort swamped the average.
Ad networks are ruthlessly concentrated. When an algorithm finds an ad creative that delivers cheap installs, it doesn’t distribute budget evenly — it pours a disproportionate share of your spend into that single creative.
If that scaling creative made an exaggerated promise or highlighted a late-game mechanic, that single cohort floods your early funnel with the wrong expectation. Blended D1 retention drops not because the game broke, but because one creative swamped the average.
The diagnostic move is to step into your MMP (AppsFlyer, Singular, Adjust) and join attribution with downstream event telemetry — not to open the funnel dashboard and start rebalancing the early game.
The catch: attribution granularity
- SDK networks (AppLovin, Unity, Mintegral) and Meta often pass creative-level macros or ad IDs into the MMP.
- Google App Campaigns largely black-box creative reporting, leaving you at the ad group or asset-cluster level.
Where creative or ad-group attribution exists, pull retention by creative ID. Compare the baseline cohort against the asset that just scaled up.
If organic and existing campaigns are holding steady while the newly scaled creative sits at half the normal D1, you don’t have an onboarding problem. You have a creative concentration problem.