What is Trust block (AI shopping feed)?
The trust block is the segment of an AI shopping feed that carries the ranking-tiebreaker and answer-eligibility signals the model uses to decide which SKUs surface and which get gated. It contains aggregate rating, review count, a review sample (5–10 recent reviews as text so the model can quote in the answer), return rate, dispute rate, and category safety flags (health claim, financial claim, gambling, minors’ audience). AI shopping surfaces prefer 4.4/5 across 2,000 reviews to 4.9/5 across 12 reviews because distribution matters more than absolute level; return rate is treated as ‘unknown’ (not ‘low’) when missing, which is a ranking penalty; dispute rate is a hard downweight signal. The category safety flags gate answer eligibility — a missing flag on a regulated SKU can cause it to surface in a context where it shouldn’t, which typically triggers a retailer-side clawback and a merchant strike.
Key statistics
- AI shopping surfaces prefer higher review count with slightly lower rating over lower review count with higher rating — distribution beats absolute level.
- Missing return-rate data is treated as ‘unknown,’ not ‘low,’ and functions as a ranking penalty against feeds that expose the field.
- Category safety flags are a hard gate on regulated-category answer eligibility — missing flags on regulated SKUs risk merchant-side clawbacks.