# Fictional example output

Prioritize k1's ad relevance investigation: it has $900 of the $1,100 observed spend (81.82%) and 30 demos at $30 CPA. k2 has a lower score but only $50 exposure; its free budget tool wording warrants checking the actual product/offer fit before score work. No supplied offer definition establishes that it is off-offer.

k3's -- diagnostic is unavailable, not zero. Its observed demo CPA is $15, but the missing diagnostic does not establish excellent or poor quality.

Do not average QS into an account KPI or treat it as the auction input. k1's below-average ad relevance suggests checking query-to-ad alignment and drafting a bounded copy test. Snapshot components and performance dates have different windows, so a causal performance effect is unproven. No promised CPC reduction or score target was invented. No changes applied.