FMCG · Food & spices / Anonymized case study
Giving gaming inventory its own quality review
Five gaming placements retained 63.45% of gross impressions after filtering, with distinct contributions from GIVT and duplicates.
Observed window: 12 Aug 2026 to 24 Oct 2026 (UTC)
Snapshot: 25 Oct 2026
17.27%
Duplicate impressions, as a share of gross impressions.
The buying question
Gaming delivery should be reviewed in its own context. Combining it with streaming or display inventory can make it harder to see which part of the buy explains a discrepancy. For this food-and-spices advertiser, five gaming placements formed a useful separate scope.
What AdProof measured
The five placements recorded 8.76 million gross impressions. The count after GIVT filtering was 7.07 million, and the final net count was 5.56 million. GIVT exclusions represented 19.28% of gross delivery; duplicate impressions represented 17.27%. Together with the other applicable filtering, the final net count represented 63.45% of gross impressions.
What the result does and does not show
The finding describes the measured placements during their observed window. It does not establish a quality rate for all gaming inventory or for an entire app. It also does not establish whether an ad was seen, remembered, or followed by a purchase. Those questions require their own eligible populations and measurement methods.
A practical review sequence
Keep the five-placement scope intact for the initial reconciliation. Then inspect the placement-level contribution to GIVT and duplicates, check the relevant implementation evidence, and agree how the buyer and partner will report counted impressions. Evaluate attention and business outcomes separately, using the evidence available for those measures.
Take this into your next campaign review
- Review gaming delivery separately from other inventory formats.
- Keep traffic quality, deduplication, and viewability distinct.
- Use placement-level findings to guide the next investigation.
More from this portfolio
A shared measurement basis across four streaming placementsPut the method into practice
Why viewability needs a separate measurement basis