What is an AI visibility audit?
An AI visibility audit is a systematic analysis of a company's presence in the responses generated by the main language models (LLMs). It measures whether, and how, ChatGPT, Claude, Perplexity, Gemini and Copilot cite, recommend or mention your brand when a user asks a question related to your sector.
Unlike a classic SEO audit, which analyses your ranking in Google's results, an AI visibility audit assesses your citability in conversational AI responses. This distinction between GEO and SEO is key to understanding the issue.
Why this matters to executives now
The figures are clear:
- 73% of senior executives use generative AI weekly for decision-making (McKinsey, 2025)
- 40% of B2B searches now begin on an AI assistant, not on Google
- Companies cited by LLMs benefit from a trust rate 3x higher than those appearing only in traditional results. Our guide on AI visibility for businesses details the 7 levers of citability
The problem: most companies have no idea what AI says about them. Worse still, some discover that LLMs routinely recommend their competitors.
The 5 steps of an AI visibility audit
1. Scope definition
The audit starts with the identification of the strategic queries your potential clients ask AI. For example:
- "What is the best [your sector] consulting firm in Switzerland?"
- "How do I choose a provider for [your service]?"
- "Which companies are leaders in [your field]?"
At MCVA Consulting, we test a minimum of 50 strategic queries per audit.
2. Multi-LLM querying
Each query is submitted to the 5 major LLMs: ChatGPT (GPT-4), Claude (Anthropic), Perplexity, Gemini (Google) and Copilot (Microsoft). Responses are collected and analysed systematically.
Why test 5 LLMs? Because each model has its own data sources, biases and update frequency. A company can be well cited on Perplexity (which accesses the web in real time) but completely absent from Claude (which depends on its training data).
3. Citability analysis
For each response, we assess:
- Direct mention: is your company named?
- Position: where in the list of recommendations?
- Quality of description: does the AI describe you correctly?
- Sentiment: is the tone positive, neutral or negative?
- Competition: which competitors are cited in your place?
4. Competitive benchmark
The audit compares your AI visibility with that of 3 to 5 direct competitors. This benchmark often reveals surprises: smaller companies that are better GEO-optimised are sometimes cited more than the historic market leaders.
5. GEO Score and action plan
The audit concludes with a proprietary GEO Score (0-100) that summarises your AI visibility, and a prioritised action plan to improve it. Typical actions include:
- Optimising structured data (JSON-LD)
- Reworking the semantics of key content
- Strengthening internal linking
- Creating targeted content for missed strategic queries
Detailed methodology: how the GEO Score is calculated
The GEO Score relies on a rigorous protocol. Here is how the methodology we apply at MCVA Consulting works in practice.
Test matrix: 10 queries x 4 LLMs
For each audit, we build a test matrix made up of at least 10 representative queries submitted to 4 LLMs (ChatGPT, Claude, Perplexity, Gemini). This produces a minimum of 40 responses to analyse. On a complete audit, we expand to 50 queries and 5 LLMs, i.e. 250 responses.
Each response is evaluated on 5 independent components, scored from 0 to 20:
- Presence (0-20): is the company mentioned in the response?
- Position (0-20): in what order does it appear relative to competitors?
- Accuracy (0-20): is the information provided correct and up to date?
- Sentiment (0-20): is the tone of the mention positive, neutral or negative?
- Context (0-20): is the company recommended in the right sectoral context?
The weighted average of these 5 components, aggregated across all queries and LLMs, produces the final GEO Score out of 100.
The 4 GEO Score levels
The GEO Score reads against 4 clearly defined thresholds:
| Score | Level | Meaning |
|---|---|---|
| 0 – 25 | Invisible | LLMs do not know you. Your company does not exist in the conversational AI universe. Critical situation: prospects who query AI will never find you. |
| 26 – 50 | Emerging | You are mentioned sporadically, often with incomplete or imprecise information. AI knows you but does not actively recommend you. |
| 51 – 75 | Visible | AI cites you regularly and correctly. You appear in recommendations, but not systematically in first position. Significant room for improvement. |
| 76 – 100 | Leader | AI recommends you as a priority on your strategic queries. Your brand is a reference in generated answers. A position to defend and consolidate. |
The majority of Swiss SMEs we audit score between 5 and 30, in the Invisible to Emerging zone. Companies that invest in a structured GEO strategy reach the Visible level in 3 to 6 months. This dynamic is part of the online search revolution reshaping acquisition channels.
Signals that influence your GEO Score
Some elements of your online presence directly improve your citability by LLMs. Others degrade it. Here is a summary:
| Positive signals (improve GEO Score) | Negative signals (degrade GEO Score) |
|---|---|
| Complete JSON-LD structured data (Organization, LocalBusiness, FAQ) | No structured markup on the site |
| "Definition + expertise" content with citable sentences | Purely commercial content with no informational value |
| Mentions and citations on authoritative third-party sources (press, professional directories, Wikipedia) | No mention on sources external to the site |
| Expert biographies with verifiable credentials (LinkedIn, publications) | Anonymous authors or no public profile |
| Structured FAQ pages answering frequent sector questions | Complex navigation and content buried in non-indexable PDFs |
| Coherent internal linking connecting the site's key entities | Broken links and orphan pages |
| Verified customer reviews on Google Business, Trustpilot or sectoral platforms | No reviews or mostly negative reviews |
| Content updated regularly with visible publication dates | Outdated content with no date or with old dates |
| Active presence on professional networks (LinkedIn, conference talks) | Inactive or non-existent social profiles |
What an audit concretely reveals
Here are the most frequent findings during our audits:
AI ignores you. Your company is not mentioned in any response, even for queries directly related to your activity. This is the case for 60% of the Swiss SMEs we audit.
AI describes you poorly. AI mentions your company but with outdated, incomplete or incorrect information. This happens when your web content is not optimised for extraction by LLMs.
Your competitors dominate. AI systematically recommends your competitors for strategic queries. An audit identifies why and how to reverse the trend.
10-minute self-diagnosis checklist
Before commissioning a professional audit, you can run a quick first diagnosis. Open ChatGPT (or Claude, Perplexity, Gemini) and ask the following questions, replacing the bracketed elements:
- "What are the best [your profession] in [your city/region]?" Does your company appear in the response?
- "What do you know about [your company name]?" Does the AI give correct, up-to-date information?
- "I'm looking for a provider for [your main service]. Who do you recommend in Switzerland?" Are you on the list?
- "Compare [your company] and [your main competitor]." Is the comparison favourable? Is the information accurate?
- "What are the advantages of working with [your company]?" Does the AI know your differentiators?
Quick interpretation:
- 5 positive responses: you are probably in the Visible or Leader zone. A complete audit will confirm and identify optimisation areas.
- 2 to 4 positive responses: Emerging zone. Targeted actions can quickly improve your score.
- 0 to 1 positive response: Invisible zone. A professional audit is strongly recommended to establish a roadmap.
Repeat the exercise on at least 2 different LLMs. Results vary significantly from one model to another: being visible on ChatGPT does not guarantee visibility on Claude or Gemini.
What does a professional GEO audit report contain?
An express or complete audit by MCVA Consulting delivers a structured report comprising the following:
- Overall and per-LLM GEO Score: your aggregate score and the breakdown by model (ChatGPT, Claude, Perplexity, Gemini), as gaps between LLMs reveal specific opportunities.
- Component breakdown: detailed scores on the 5 axes (Presence, Position, Accuracy, Sentiment, Context) with identification of priority weak points.
- Competitive benchmark: direct comparison with 3 to 5 competitors on the same queries and LLMs, in radar-chart form.
- Query mapping: complete table of tested queries with citation status per LLM (cited / not cited / poorly cited).
- AI verbatim analysis: extracts of LLM responses about your company, with annotation of factual errors to correct.
- Prioritised action plan: list of concrete actions ranked by impact and ease of implementation (quick wins, medium-term actions, structural projects).
- Technical recommendations: audit of existing structured data, JSON-LD markup suggestions, content optimisations for citability.
The report is delivered as a PDF with a 45-minute video debrief. The aim: leave you with a clear view of your situation and an action plan that can be activated immediately.
Who should run an AI visibility audit?
An audit is particularly relevant for:
- B2B companies whose clients use AI to research providers
- Consulting firms and professional services companies
- Brands observing erosion of their organic traffic
- Executives who want to understand their exposure in this new ecosystem
How MCVA Consulting runs its audits
Our AI visibility audit offer is built on a proprietary methodology developed exclusively for the Swiss and European market. In 2 weeks, we deliver a complete report including the GEO Score, the competitive benchmark and the prioritised action plan.
Request a free diagnosis to discuss your specific situation.
FAQ
Is a GEO audit different from an SEO audit?
Yes. An SEO audit analyses your position in search engine results pages (SERPs): keywords, positions, organic traffic, technical aspects (speed, indexing, markup). A GEO audit analyses your presence in AI-generated responses. The two disciplines are complementary but distinct. Good SEO performance does not guarantee good GEO visibility, and vice versa. In practice, we recommend running both audits in parallel for a complete view of your digital visibility.
How often should you audit AI visibility?
Language models evolve rapidly. Their training data is updated, new models appear, and citation algorithms change. We recommend a complete audit every 6 months to track the evolution of your GEO Score. Between two complete audits, a light quarterly check (10 key queries on 4 LLMs) helps detect drops in visibility and react quickly. Companies in active optimisation phase can opt for monthly monitoring.
What budget should I plan for a GEO audit?
At MCVA Consulting, we offer two packages tailored to Swiss companies. The express audit (10 queries, 4 LLMs, delivery in 5 days) is accessible to SMEs wanting a first diagnosis. The complete audit (50+ queries, 5 LLMs, detailed competitive benchmark, action plan) is intended for companies ready to invest in a structured GEO strategy. Pricing is available on request and depends on sector complexity. Contact us for a tailored quote.
Summary
- An AI visibility audit measures whether your company is cited by ChatGPT, Claude, Perplexity, Gemini and Copilot.
- 73% of executives use generative AI to inform themselves: not being there means being invisible.
- The audit tests 50+ strategic queries on 5 LLMs and produces a GEO Score (0-100).
- The most frequent findings: AI ignores you, describes you poorly, or recommends your competitors.
- MCVA Consulting is the first Swiss firm specialising in this type of audit.
Is your company visible to AI? Don't let your competitors capture the recommendations that should be yours. Request your GEO audit or contact us directly to discuss.
Related articles
GEO vs SEO: why classic search marketing is no longer enough
SEO optimises for Google. GEO optimises for ChatGPT, Claude and Perplexity. Discover why this distinction changes everything for your visibility.
9 min
AI visibility for businesses: getting recommended by ChatGPT, Claude and Perplexity
The reference article on AI visibility. How to be recommended by LLMs, the GEO Score™ explained, and the 7 levers of citability.
10 min
From SEO to GEO: the online search revolution
From directories to AI engines: a look back at the evolution of online search and why GEO is the next SEO.
8 min