AI visibility audit: where your company stands in ChatGPT, Gemini, Google and Claude

What is an AI visibility audit?

An AI visibility audit shows how often ChatGPT, Gemini, Google AI Overviews and Claude name your company when your customers ask their questions, whether they recommend you and which sources they rely on. It is also called a GEO audit, after Generative Engine Optimization.

At anewera the audit is a measurement over at least seven days, not a snapshot. It is the second step of every engagement, and you can also commission it on its own.

Why a free check is not enough

A free check asks each question once. But the same question does not get the same answer every time, and each engine answers differently.

A Swiss study of 1,248 queries to four AI assistants shows how big those differences are: of 3,539 hotels, only 26 were named by all four (Netzwoche, 2 June 2026). Check one engine and you see only part of the picture.

So the audit asks every question several times a day on every engine. Many answers give you a share with a margin of error; a single answer gives you chance. How we calculate it is set out in our method.

What the audit covers

• Your questions: we propose questions the way your customers ask them, in the languages that matter to you. You sign off the list before we measure.
• Four engines: ChatGPT, Gemini, Google AI Overviews and Claude, with Switzerland as the location, and on Claude down to your city.
• Repeated runs: every question several times a day on every engine, for at least seven days. Failed answers are counted and reported.
• Named: the share of answers that name your company, per engine, with the number of answers and the margin of error.
• Recommended or mentioned: whether an answer advises people to choose you, only lists you, uses you as a source or names you with a caveat.
• Competitors: who the engines name instead of you, and how often.
• Sources: which pages the engines read and cite, which sentences they take from your website and which follow-up searches they run on their own.
• Your website: whether AI can read it: access for AI crawlers, llms.txt, structured data, a profile for AI agents and fast pages.
• Action plan: technology, content, sources and outreach, in order of impact.

Steps and timing

Step	What happens	Time
1. First meeting	Your market, your competitors, your goals; free	30 minutes
2. Agree the questions	We propose questions, languages and locations; you sign them off	about 1 week, depending on sign-off
3. Measure	Every question several times a day on four engines, every answer stored	at least 7 days
4. Analyse and discuss	Report, website review, action plan and a meeting to go through it	about 1 week

Allow 2 to 4 weeks from the first meeting to the report. We need nothing from you while the measurement runs.

What you receive

1. A report in plain words: every figure per engine, with its date, the number of answers and the margin of error. Where ranges overlap, we say so rather than claim a ranking.
2. The raw data: every answer, failed ones included, as an export. The data is yours.
3. The website review and the action plan, in order of impact, written so that your team or your web agency can work from it.
4. A meeting in which we explain the results and agree with you which targets make sense.

Measurement data is stored in Zurich, with Supabase in its Zurich region. We handle personal data under the Swiss Federal Act on Data Protection (FADP).

An open example: anewera itself

We show what the figures in the report look like, in public, on our own brand. In the live case study Grow in Public we have been measuring anewera itself since 6 October 2026, starting from zero, with the same measures as the audit: the share per engine with its margin of error, failed answers, competitors as groups and the sources cited.

There are no results there yet: figures appear only after three days of measurement. For a measurement on another of our own brands, see the Expat Savvy case study.

Who the audit is for

• Companies with their own market and competitors that want to know whether AI assistants name and recommend them.
• Marketing teams that need a solid baseline before investing in GEO, so that every later change can be measured against it.
• International companies in Switzerland whose customers ask in more than one language. Today we measure in German and English, with French next.

The audit is less suited to small businesses that would rather do it themselves. For them, agentready.ch has a free check and a do-it-yourself profile.

What does the audit cost?

The price depends on scope: how many questions, engines, languages and locations we measure, and for how long. Prices are in Swiss francs, excluding VAT. You receive a quote after a short first meeting.

If you want to carry on after the audit, anewera takes over delivery and monitoring as your GEO agency.