Scores & metrics

Every run produces a deterministic score using the same public formula. There is no learned model and no hidden weighting — the same inputs always produce the same output, so your scores can be audited and compared across time.

The top-level AI Visibility Score is a weighted average of four subscores. Each subscore is normalized to 0–100 before weighting.

Visibility = (0.35 × Presence)
           + (0.25 × Position)
           + (0.20 × Coverage)
           + (0.20 × Consistency)

The four subscores isolate different dimensions of AI visibility. A brand can rank well on presence but poorly on position, and the subscore breakdown exposes that asymmetry.

Presence

Share of prompts where your brand is mentioned at all.

Position

Average rank of your first mention within each response.

Coverage

Breadth of prompt categories where you appear.

Consistency

Stability of your mentions across ChatGPT and Gemini.

Every run also carries a confidence band based on the number of prompts that actually returned parseable responses. Runs with fewer than twelve successful prompts render the score in a muted state and warn that the sample is small.