GameAnalytics’ median retention is a population description

GameAnalytics describes median mobile D7 retention as just under 4% during 2025. Its sample applies a 1,000-MAU threshold, and the report does not include genre cuts. Report methodology and D7 section.

Identify the comparison you actually need

A studio deciding whether to scale a puzzle game needs a different comparison from a portfolio team surveying broad engagement. The report’s median can orient the latter question while remaining too broad to resolve the former. Passing a mixed-population median does not establish that acquisition economics work.

GameAnalytics’ metric documentation describes strict exact-day retention. Match that definition before placing a studio number alongside the report. A rolling-return metric or a return-after-tutorial cohort can produce a different answer without either calculation being wrong.

Replace the automatic gate with a diagnosis

Our recommendation is to compare the game with its own previous build and acquisition cohorts first. Inspect where the curve changes and whether the traffic mix moved. Then use the external distribution as context, with the sample restriction still visible.

The absence of genre detail is a meaningful limit, not a blank to fill with an invented puzzle or strategy estimate. The accompanying card is deliberately unsuitable for claiming that every game below one percentage should be killed. It is suitable for identifying what additional evidence a greenlight discussion needs: product comparability, cohort maturity and the economics of the users actually being acquired.

Inspect the evidence

Download the source-population and comparability card. The extraction separates documented facts from our analysis and unknowns.

Further reading in the existing archive: The retention curve you should be showing your creative team; Retention is a UA metric. These links provide background; this check does not independently verify their full contents.