Glossary
Perception Score
A perception score is a composite metric summarizing how favorably and accurately AI engines represent a brand, typically combining mention rate, recommendation position, claim accuracy, sentiment, and citation ownership across a fixed prompt set. Its value lies in trend and comparison — tracking movement scan over scan — rather than in the absolute number.
Composite scores exist because 'how are we doing in AI?' needs a one-line answer for executives, but the components matter more than the composite. A score of 72 is meaningless in isolation; a score that dropped eight points because recommendation rate fell on pricing prompts is a diagnosis.
Beware of score-only tools and score-only dashboards. A number without the underlying answers invites two failure modes: chasing the score with actions that don't change what buyers are told, and dismissing real problems because the aggregate looks stable while one high-intent prompt flipped to a competitor.
Use a perception score the way you'd use a health metric: as a trigger for drill-down. The operational questions are always concrete — which questions, which engine, which claim, which source — and the score is just the doorbell.
See what AI currently says about your brand
Perciva runs buyer-intent prompts across ChatGPT, Perplexity, Gemini and Claude and shows you the answers verbatim — including which competitor gets recommended instead of you.
Get a free snapshot