AppsFlyer’s creative-volume figure is a capacity reference, not a quota

AppsFlyer reports 2,400–2,600 quarterly creative variations among its top gaming advertisers. That is a description of a selected production tier, not a required output for a studio. Source release.

Count the unit before planning the team

A variation can differ from a concept, a localisation or a test hypothesis. Before comparing your output with the reported range, define which changes receive a new asset identifier and which represent a materially different idea. A team that counts every language export separately cannot assume it is measuring the same activity as a team that counts only new concepts.

Our editorial recommendation is to build two internal counts: distinct hypotheses and deliverable versions. The first helps assess learning coverage; the second helps plan production and trafficking work. Neither count alone establishes that an asset reached enough players to support a decision.

Use the report without inheriting its ambition

The report introduction describes the larger gaming dataset, but that population is not the denominator of every creative-volume claim. The public release does not disclose enough detail to reproduce the spending-tier range independently.

Use the figure to ask whether production infrastructure can support the studio’s chosen test plan. Do not multiply headcount until an arbitrary industry number is reached. A useful budget proposal connects each proposed batch to a question, an audience, a delivery plan and an expected decision; the report can supply context around that proposal without dictating it.

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: AI Creative Testing Volume Benchmarks for Mid-2025; The benchmark report is selling you something. These links provide background; this check does not independently verify their full contents.