SERVICES · AI VISIBILITY
We measure in ChatGPT, Gemini, Perplexity, Claude and Google AI, with the questions your buyers actually ask. Measure first; improve second.
Short answer
AI visibility measurement establishes how often AI assistants name a company when asked the questions its buyers actually ask. A fixed prompt set is run repeatedly across ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews, every answer is stored verbatim with its date, and you receive three numbers: mention rate, citation rate and share of voice.
RESULTS
A Greek B2B professional services firm with years of serious SEO work. The same set of 172 prompts, before and three months after. The percentage shows how often the company was named in each engine.
ChatGPT
0% 65%
Gemini
0% 34%
Claude
9% 19%
Google AI
85% 87%
PHOENIX VERUS measurement, 2026. Not a forecast for a new client. The prompts and the verbatim answers are available to the client. How the audit works.
Most companies find out whether AI systems mention them by asking ChatGPT once and reading the answer. That is one sample from a system that varies between runs, sessions and accounts. It tells you almost nothing, and nothing you can compare next month.
A measurement is different. Twenty or more questions, phrased the way your buyers phrase them. The same questions on five engines, repeated, because a single run is noise. Every answer stored as written, with the date. From that come three numbers you can put in front of management: how often you are named, how often your site is cited, and what share of all the category’s mentions is yours.
The part that surprises people is the timing. The baseline has to be taken before anything on the site changes, because once it changes the clean before-state is gone for good. We have seen companies invest a year in their presence in AI systems and be unable to prove what it brought, because nobody measured first.
A fixed prompt set, run monthly across five engines, is the only way to see a trend rather than a reading. We built the software that does it.
The check happens once and tells you where you stand. Monitoring runs every month and tells you where you are heading. Both start with the same measurement; the second needs the first.
ONE-OFF
Baseline across five engines, an audit of the site as the AI crawlers read it, an llms.txt file, an Excel report and a one-hour presentation of findings. On Standard, a follow-up audit one month later. Delivered in two to three weeks.
From €550, published prices per tier.
MONTHLY
The same prompt set runs every month in the dashboard: mention rate, citation rate and share of voice per engine, with the verbatim answer behind every number and a history over time. Starts after the check, on its baseline.
Monthly subscription, quoted after the check, depending on prompts and engines.
How often, and in what terms, AI systems name you when someone asks them a question in your category. It is measured with a fixed prompt set run repeatedly across several engines, not with one question typed into ChatGPT.
ChatGPT, Gemini, Perplexity, Claude, and Google’s AI Overviews and AI Mode. Each engine is reported separately, because they do not behave the same way: a company can appear often in Google and not at all in ChatGPT.
Because a single answer is one sample from a system that varies between runs, sessions and accounts. A rate across repeated runs on five engines is a measurement; a screenshot is an anecdote.
No. Google’s AI Overviews are built on Google’s index, so SEO helps there. ChatGPT, Claude and Perplexity crawl independently and answer from what they hold. In a measurement we ran, a company with 85% presence in Google’s AI surfaces had 0% in ChatGPT. The two are measured and worked on separately.
It depends entirely on the category. In a soft market, being named in eight of twenty prompts is strong. What matters more is the direction of travel from your own baseline.
The AEO check happens once: we measure where you stand today, audit the site and deliver the findings and the fixes. Monitoring is monthly: the same prompt set runs every month in the dashboard, so you see a trend rather than a reading. Monitoring starts after the check, because it needs the check’s baseline.
Businesses that get chosen by asking a question: hotels, professional services, medical and dental practices, e-commerce, B2B suppliers. If your customers ask “which one should I pick” before they find you, it applies to you. Prices start at the first tier of the check, so it fits small companies too.
Yes, and it is the one step that cannot be done retrospectively. Once the site changes, the clean before-state is gone permanently.
If you want to understand first what we examine and why, the guide “How to Get Cited by ChatGPT” contains the nine steps, the prompt-set worksheet and the 172-prompt measurement before and after. A 15-page PDF, with a business email address.
Send us the brief or book a short call. You get concrete next steps in reply.