The headline of every report is a single letter grade. Here is exactly how it is computed.
Mention rate
Every scan asks 29 buyer questions across 3 AI channels — roughly 90 sampled answers. Your mention rate is the share of those answers that mention your brand:
mention rate = answers mentioning you ÷ total answers sampled
A mention is counted when your brand name (or a close variant of the name you scanned with) appears in the answer text. Mentions are attributed at sentence level, so the dashboard can also show how you were mentioned, not just that you were.
The grade bands
- A — mention rate ≥ 60%. AI systems recommend you consistently.
- B — 30–59%. You appear regularly but competitors still take many answers.
- C — 10–29%. You appear occasionally; most buyer questions go to others.
- F — below 10%. AI answers effectively don't know you exist.
Why your grade can move between scans
AI model outputs are not deterministic — the same question can yield different answers on different days. Small movements (a few percentage points) are normal noise. What matters is the trend across weeks, which is exactly what monitoring shows. Treat a single scan as a photograph, not a verdict.
What a realistic target is
Very few businesses score an A across all questions — generic questions ("best X in the world") are dominated by household names. A practical goal:
- Get out of F — ensure the AI can find and cite you at all (start with the website diagnosis).
- Win the specific questions — location- and niche-qualified prompts are winnable; broad ones mostly aren't.
- Track share of voice against your actual competitors rather than chasing an absolute number (SOV, sentiment and trends).