How AI Brand Visibility Is Actually Measured (And Why Most Tools Don't Bother)

Ask ten "AI visibility" or "GEO" tools how confident they are in the score they just showed you, and most will just repeat the number louder. That's the wrong answer — a score with no error bar is a guess wearing a suit. Here's what actually goes into a number we're willing to put our name on.

The question isn't "what's the score" — it's "how sure are we"

Every scan runs real buyer-intent questions against ChatGPT, Claude, Gemini and Perplexity, and samples in rounds — stopping as soon as your brand's mention-rate estimate reaches a target precision (a 95% Wilson confidence interval of a set half-width), not after a fixed number of prompts. A brand that's clearly dominant or clearly absent converges fast; a contested one keeps sampling. The precision behind the headline number is a design goal, not a side effect of whatever ran.

Three real sources of uncertainty, not one invented margin

The confidence interval on every score combines three independently measured things, in quadrature:

• Rerun variance — how much the exact same question, asked twice, moves just from model stochasticity.
• Per-engine extraction reliability — measured against a gold-labeled dataset of real AI answers, so we know how often the parser correctly reads a mention, a rank, or a sentiment, instead of assuming it's always right.
• Category-phrasing sensitivity — the sharpest one. We measured a real brand scoring meaningfully differently under two near-synonym category phrasings, with non-overlapping confidence intervals. That's genuine AI behavior — the model treats close synonyms as different questions — not measurement noise, and pretending otherwise would just be lying about the margin.

Deterministic scoring, honest limits

Given the same collected answers, the scoring engine always produces the same score — no hidden randomness in the grading step. But a score is still a point-in-time snapshot: AI answers shift as models and their sources update, so a re-scan next month can move for reasons that have nothing to do with anything you did. That's exactly why the interval — and the trend across scans — matters more than any single number.

Full technical detail lives on the methodology page. If you want to see it applied to your own brand, run a free scan.