AI-assistant queries are not a census of marketer behaviour

AppsFlyer reports that 46% of AI-assistant queries concerned reporting and performance breakdowns. A query share does not mean that 46% of marketers use AI primarily for reporting. Vendor release.

Identify who can enter the denominator

Product telemetry can reveal how a particular tool is used. It may also reflect the tasks that tool makes easy, the customers who have access and repeated activity by a small set of frequent users. These are reasons to ask about sampling, not evidence that the reported share is wrong.

Our recommendation is to keep query counts, active accounts and unique users in separate columns. A team evaluating its own assistant can then distinguish broad adoption from intensive use by a few specialists. Also separate completed work from prompts submitted: repeated attempts may increase query volume without increasing useful output.

Use the finding as a workflow question

The gaming report page places the finding in a wider marketing analysis, but does not make it an independent survey of every AI workflow. The card records the missing account and user denominators explicitly.

For a studio, the actionable question is whether reporting work is repetitive, well-defined and verifiable enough to improve with assistance. Measure time saved and correction effort on actual tasks before treating usage as value. The external percentage can suggest a place to investigate; it cannot establish a staffing saving or the quality of an automated decision.

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.