Visibility answers one question: when AI engines are asked the questions that matter in your category, how often are you part of the answer? It is shown as a percentage on your overview and dashboard, and it is the first of the five factors in the Flare score. In the wider industry the same idea goes by AI visibility score or mention rate.
From question to number
Every figure comes out of the same four steps:
- Ask. Your brand's questions are put to five engines: ChatGPT, Gemini, Google Search, Perplexity and Claude.
- Read. Each answer is read for brand mentions: yours, and every competitor you track.
- Judge. A frontier-model judge records where each brand appears in the answer and how it is described, and checks what is said about you against your fact sheet.
- Add up. Results roll up by question, by engine and by day into the numbers on your pages.
The schedule and the detail of each step are in how checks work.
The questions
Measurement starts from a set of questions: what a real buyer in your category asks an AI assistant. "Best X for Y", "X vs Y", "is X worth it". They're written for your brand when you add it, in the market you chose, and they stay largely the same from week to week, because a trend only means something on a stable yardstick. Your plan sets how many you can track (see plans and limits).
Questions that name your brand don't count toward Visibility. Being named in the answer to "is Acme any good?" tells you nothing about whether AI recommends Acme.
Spotting a mention
Engine answers are prose, so counting mentions is itself an AI task. brandflare uses an LLM-as-judge, a frontier model given a narrow job, to work out for each answer:
- whether your brand is named (a brand mention), under its own name or any alias or product name registered for it;
- which competitors are named alongside it, and in what order;
- how each brand is described.
A mention is credited to your brand by matching its name and aliases, so a similarly named organisation isn't counted as you.
The numbers on your overview
- Visibility is the share of answers that named you, across all five engines. A Visibility of 62% means you were part of roughly six answers in ten.
- Share of voice is your share of all mentions among you and the competitors you track, with your place among them. See competitor tracking.
- Average position is where you usually land when you are named: "Usually named 2nd".
- All engines counts the questions where every engine named you.
- By engine shows the same Visibility figure for each engine on its own, including any engine that hasn't been checked yet.
Each one carries a bar coloured by strength, a sentence in plain words, and the change since 14 days ago.
Engines disagree, and that disagreement is useful. The same question can also get a different answer from one run to the next. Measuring a full set of questions across five engines, week after week, is what makes the number steady enough to act on.
Accuracy is measured separately
Whether you're named and whether what's said is true are different measurements. Alongside mention detection, the judge compares every claim an engine makes about your brand with your fact sheet, and raises false claims by severity: wrong prices, discontinued products described as current, invented features. Things an answer left out are never raised; only things AI states that are wrong. Full detail is in Answer Accuracy.
Both feed the Flare score, as separate factors. A brand that is named constantly but described wrongly doesn't look the same as one named constantly and described correctly.
Reading it
There is no universal "good" figure: a household name and a two-year-old challenger occupy different ranges (see what is a good AI visibility score?). Three readings matter:
- Direction. Is the trend up after the work you shipped? Trends on your overview charts visibility, share of voice and position over time.
- Gap. How far are you from each competitor on the same questions (competitor gap)?
- Distribution. Which engines and which questions drive the number? Questions & Prompts shows exactly where you're absent.
Every number has receipts
Every answer is stored. Any figure opens into the questions and engines behind it, and from there into the answer itself, word for word, with your brand highlighted, the brands it named in order, and the websites it used. If a number surprises you, the reason is one click away.