AI visibility

Whether AI names you is measurable. We measure it.

We put your buyers' questions to ChatGPT, Gemini, Perplexity, Claude and Google's AI results, many times over, and record who is named, which sources are cited, and which passages the answers were built from: yours or your competitors'.

Apply for a free 15-minute consultation

The measures

Five numbers, each with a definition.

Calculated per query, per platform, for you and every competitor in the set.

Answer share

Of n answers to the same query, the number that name you. AI answers vary from run to run, so one answer proves nothing.

answer share = k / n

Share of voice

Your mentions as a proportion of all mentions across the competitor set, for the same queries.

SoV = mᵧ / Σ mᵢ

Query coverage

The proportion of the query set, from problem to selection, in which you are named at least once.

coverage = |Q named| / |Q|

Citation coverage

Of the sources the answers cite, the proportion that mention you. AI systems repeat what their sources say.

citations naming you / citations

Retrieval margin

How closely your best passage matches the query, scored with our own embedding model, against the best competing passage. A negative margin usually explains why you were left out.

cos(q, pᵧ) − max cos(q, pᵢ)

How it is measured

From query set to cause.

01
Build the query set

The questions buyers ask at each stage: the problem, who provides it, how providers compare, and which one to choose.
02
Sample every platform

Each query is run repeatedly on each platform. Every answer is stored with the companies it names and the sources it cites.
03
Score the passages

Your pages and your competitors' are split into passages and scored against each query, so we can see which passage won and by how much.
04
Trace each absence

Every query you are missing from is traced to a cause: no matching passage, an entity the system could not resolve, or sources that do not mention you.

What the numbers are

Samples, stated as samples.

Every figure is reported with its n, the platform and the date it was sampled
The same queries and measures are re-run after changes, so results are compared like for like
Answer share and citations are observed directly. Passage scores come from our own models, which approximate retrieval systems rather than reproduce them
No traffic or enquiry forecasts: we report what was retrieved, not what it will earn

Platforms

Measured on every platform your buyers use.

Free 15-minute consultation

Find out where your business stands.

How search and AI systems currently interpret your business
Where competitors are better understood than you
Specific retrieval weaknesses, checked before we speak
Whether there is work worth doing, and what it is worth to you