Trillboards classifies every creative that flows through its DOOH network across 24 IAB tier-1 content categories — Arts & Entertainment, Automotive, Business, Education, Food & Drink, Health & Fitness, and so on through the full IAB Content Taxonomy. Each classification combines automated vision-and-text analysis (a Gemini-based classifier) with a human-review fallback for low-confidence outputs. The classification metadata then feeds the serving-decision layer, which enforces per-venue and per-buyer brand-safety policies in the openRTB auction.
The pages below break down the network's brand-safety enforcement by IAB category. Each per-category page shows the classification volume, the family-safe pass rate (share of creatives that clear the alcohol / tobacco / gambling / adult / violence / profanity gates), the classifier's average confidence for that category, the venue types where category creatives most commonly serve, and the common pitfalls that the review queue catches. The data is aggregate-only — no per-creative identifiers, no advertiser names, no demand-partner attribution. It is intended to give buyers and venue owners a structural view of how programmatic DOOH brand safety operates in practice.