AI visibility audit: what to check, and in what order
A step-by-step AI visibility audit. How to check whether ChatGPT, Gemini, Perplexity and Google AI Overviews name your brand, and what to fix first.

What an AI visibility audit is for
Before you change anything, you need to know where you stand. An AI visibility audit answers four questions:
- When buyers ask about your category, do AI engines name you?
- When they do, what do they say about you, and is it right?
- Who do they name instead?
- Which sources are they drawing on?
The answers give you a baseline to measure every later change against. Without one, you cannot tell whether your work moved anything.
Step 1: Pick the questions
Write 10 to 30 questions your buyers actually ask. Mix three kinds:
- Category questions: "What are the best [service] companies for [type of business]?"
- Problem questions: "How do I fix [problem your service solves]?"
- Comparison questions: "[Your brand] vs [competitor]" and "Alternatives to [competitor]."
Use real wording. Sales call notes, support emails, Google's "People also ask" and Search Console queries are good sources. Keep the list fixed so you can repeat the audit later and compare like with like.
Step 2: Run them on each engine
Ask every question on each engine your buyers use. At minimum: ChatGPT, Google (AI Overviews and AI Mode where available), Gemini, Perplexity and Microsoft Copilot. Add Claude and Grok if your buyers use them.
A few rules make the results comparable:
- Use a logged-out or fresh session where you can, so past chats do not colour the answer.
- Record the date, the engine and the full answer text.
- Save the list of sources each engine shows.
Step 3: Score the answers
For each answer, record:
| What to record | Why it matters |
|---|---|
| Named or not | The core visibility measure |
| Position: first, in the list, or missing | Being first in a short list carries more weight |
| Accuracy of the description | A wrong description can cost you the buyer |
| Competitors named | Shows who is winning the answer today |
| Sources cited | Shows where the engine gets its view of your category |
Add up how often you are named across all questions, per engine. That share is your baseline. It is the core of what our AI Visibility Score measures.
Step 4: Find the sources
Look at the sources the engines cite for the questions where you are missing. You will usually see a pattern: a handful of directories, review sites, industry publications, comparison articles or competitor pages come up again and again.
Those are the places that shape the answer. For each one, ask: are we listed there? Is our listing accurate? Could we be?
Step 5: Check your own site and profiles
Now look inward. Check:
- Crawler access. Does your robots.txt allow the AI crawlers you want, such as GPTBot, OAI-SearchBot, PerplexityBot, ClaudeBot and Google-Extended?
- Plain description. Does your home page or About page state, in one sentence, who you are, what you do and who you serve?
- Answer-first pages. Do you have pages that answer the questions from step 1, with the answer in the first sentence?
- Structured data. Do you have Organization markup, and FAQPage markup on real question-and-answer sections?
- Consistent facts. Are your name, address, phone, categories and description the same on your site, Google Business Profile, LinkedIn and key directories?
Step 6: Rank the fixes
You now have a list of problems. Sort them by effort and impact:
- Errors first. If an engine describes you wrongly, correct the source it is drawing on.
- Quick technical fixes. Crawler access, structured data and clear descriptions take days, not months.
- Missing answers. Write the pages that answer questions where you are absent.
- Missing listings. Get onto the sources engines cite most in your category.
- Longer-term authority. Press, expert commentary and partnerships that build independent evidence over time.
How often to repeat it
AI answers change as models and their sources update. Re-run the same question set every quarter, and after any major change to your site. Track the trend per engine rather than reacting to a single result.
Doing it yourself, or with help
You can run a small audit by hand in an afternoon with the steps above. The work grows quickly with more questions, more engines and repeat runs, which is why we built the AI Visibility Score. It runs your buyer questions on each engine and gives you the baseline, and the first stage of every LeadAIO engagement is a full audit that goes deeper. See how it works, or read what generative engine optimization is for the bigger picture.