AI visibility audit: what ChatGPT, Perplexity and Google say about you

Most owners find out about AI search the same way: someone asks ChatGPT who to hire, and it names a competitor. An AI visibility audit replaces that one worrying answer with a clear picture. It shows what the tools say across the questions your buyers actually ask, why, and which fixes will change it.

What do buyers say about the audits they have already bought?

The phrase is new enough that it covers very different products, and people who have bought them are candid about it on Reddit's marketing and SEO forums. The complaints repeat:

• Scores nobody will explain. A buyer at a mid-sized B2B firm described inviting six agencies to pitch AI visibility work. All six led with their own visibility score. Two would only explain how it worked under a non-disclosure agreement, and one could not say how many questions it asked or how often.
• Reports full of zeros. A practitioner exported a tool's ten-page brand report and found pages of zeros, an invented "likelihood to buy" figure, and not one actual ChatGPT answer. A business that is invisible gets a report that says so ten times and explains nothing.
• Free audits as sales bait. Threads offering free "AI audits" draw rival vendors offering theirs, and replies that mock the whole idea. A free audit is usually the first step of a sales call.
• One answer treated as a result. Several practitioners make the same point: ask the same question twice and you can get different businesses. One tracked a local electrician daily for a month and found ChatGPT named it almost every day while Perplexity never did.

An AI visibility audit is only useful if you can see the questions, the answers and the sources behind every finding.

How is this audit done?

The method is built around those complaints. Everything is visible to you.

1. A fixed set of buyer questions. Drawn from your brief and agreed with you before the work starts: category and location questions, problem-first questions, comparisons with named competitors, and your business name on its own. Questions about your brand are reported separately, because they almost always look better than the questions that win new customers.
2. Each question asked more than once, in a clean session, in ChatGPT with search on, Perplexity and Google. A single answer is noise; a pattern across runs is a finding.
3. Every answer recorded with the date, the businesses it named, how it described you and the pages it cited.
4. Your facts checked. Wrong prices, old services and outdated addresses are traced to the listing or page the tool took them from, because that is the thing to correct.
5. Access checked. Whether the AI search crawlers can reach your site, and whether Google and Bing have indexed the pages that should answer these questions.

The guide on what a good AI search visibility audit includes explains each part in more depth, if you want to judge any audit, this one included.

What does it usually find?

The patterns owners and practitioners report most often are the same ones an audit is designed to surface:

• Ranking on Google but missing from AI answers. Strong rankings do not guarantee a mention. The businesses named instead tend to appear on more outside lists, directories and review sites, or describe who they serve in one clear line the tools can repeat.
• Named for your brand, absent for your category. One founder found his product appeared whenever he asked about it by name, and in 2 of 15 questions about his category. Checks that only use your name hide the real gap.
• Wrong facts from stale sources. Old prices and discontinued services usually come from a directory listing nobody has touched in years.
• Blocked by accident. Some sites block AI crawlers with a rule written for something else. The guide on which AI crawlers to allow explains the difference.

What do you get at the end?

Within 7 working days:

• The question set and the recorded answers, with sources, for you and your two competitors.
• Your Google rankings for your 10 most valuable searches.
• A review of your website's conversion and your current ads, because visibility is wasted if the visit does not become an enquiry.
• A ranked list of fixes, most valuable first, with an honest sense of effort for each.
• A 45-minute walkthrough with Muhammad to go through it together.

There is no proprietary score and no dashboard to log in to. The list is yours to act on with anyone, or yourself.

How does $700 compare?

Market prices vary widely. People in the forums report AI visibility tracking tools from roughly $50 a month for a small number of questions to several hundred a month for more, with enterprise platforms and agency programmes quoting thousands a month. Those are ongoing costs for monitoring; few include anyone telling you what to fix.

The Visibility Audit is a single fixed fee of $700 for the diagnosis and the plan. It is the same audit described on the SEO audit service page, which covers the Google side in more detail. For what audits of all kinds tend to cost, see how much an SEO audit costs.

What happens after the audit?

You decide. Many fixes are things you or your web developer can do from the list. If you want the work done for you, it continues in Growth Partner, which includes two new pages a month written to be quoted and a monthly AI visibility check against the same questions, so progress is measured against the audit's baseline. The AI SEO services page explains how that ongoing work runs.

Before you book, the free ten-minute AI test will show you whether there is a gap worth paying to investigate.