This tool samples a small number of live AI answers and reports what share of them mention your brand. It is a spot check, not a measurement: a sample this size tells you whether you appear at all, and it cannot tell you whether your visibility went up or down since last month.
What the number actually is
The score is mentions divided by prompts sampled, expressed as a percentage. Each prompt is sent to a live answer engine and the returned answer is searched for your brand name and domain. Nothing is estimated and nothing is modelled — if the answer did not name you, it counts as zero.
Two things it deliberately does not do. It does not weight prompts by how much commercial intent they carry, so a brand-name lookup counts the same as a buying question. And it does not separate "mentioned in passing" from "cited with a link", which are very different outcomes for traffic.
Why a small sample cannot show a trend
Answer engines are non-deterministic. The same prompt asked twice can return different sources, so a single observation per prompt has no error bar. The scorer behind our paid audits refuses to report a change as real below 40 distinct prompts, and flags anything under that as insufficient coverage.
That threshold is not a sales gate. It is the point below which a difference between two runs is indistinguishable from the engines' own variance. If a vendor shows you a visibility percentage moving month to month off a dozen prompts, ask how many observations per prompt they took.
What to do with a zero
A zero on a brand-name prompt and a zero on a category prompt mean different things. Zero on your own brand name usually means an indexing or entity problem — the engines cannot resolve who you are. Zero on a category question ("best X for Y") is normal for most companies and is a content and third-party-source problem, not a technical one.
Check the prompts list in the result. If the answers cite listicles and forum threads rather than vendor sites, the path to being mentioned runs through those sources, not through your own pages.
Frequently asked questions
- How many prompts do you need to measure AI visibility reliably?
- At least 40 distinct prompts, with more than one observation per prompt. Below that, our scorer reports coverage as insufficient and will not call a run-to-run difference statistically reliable. This free tool samples far fewer than that on purpose — it is a spot check to see whether you appear at all.
- Which AI engines does this check?
- OpenAI web search, Gemini grounding, and Perplexity. Results are live answers, not cached.
- Why did I get a different score running it twice?
- Answer engines are non-deterministic — the same prompt can retrieve different sources on different runs. That variance is exactly why a small sample cannot be read as a trend.
- Is being mentioned the same as being cited?
- No. A mention is your name appearing in the answer text. A citation is your URL appearing in the source list. Citations drive traffic; mentions mostly do not. This tool measures mentions.
Method last reviewed . Thresholds and engine list mirror the scorer used in SurfIO's paid audits.