Ask AI who to hire. See who it names.
An AI visibility audit measures whether AI assistants name your business when a buyer asks for a recommendation, which of your pages they cite, and which competitors they name instead. Ours reads six engines and runs every question three times, because the answers change between runs.
Free for qualified businesses. No card. Delivered within 5 business days.
What an AI visibility auditactually measures.
It measures the answer, not your website. A buyer opens ChatGPT and types "who is the best commercial roofer in Tampa." The assistant writes a paragraph, names two or three companies, and links a handful of sources. The audit records that moment: whether you were named, whether you were linked, what the answer said about you, and which competitors took the slots you did not.
That is a different question from the one an SEO audit answers. An SEO audit examines your site. This examines what the assistants say when your name comes up, which is what a lead reads before they call you. The two are connected, because assistants search the web to build their answers, so the pages that rank are the pages that get read.
Background reading: what AI engines actually do when someone asks about your business, and how an AI visibility audit differs from an SEO audit.
Why one check of one enginetells you almost nothing.
Most free AI visibility checkers ask one assistant one time and print the result. We kept every answer we have ever recorded, so we can show what that approach misses. These two figures come from our own stored runs, not from a vendor study:
of the sources cited for a question were cited by all three of ChatGPT, Perplexity and Google's AI Overviews. Another 84% were cited by exactly one of them.
2,376 cited domain and question pairs · 194 questions · 24 auditsof the hidden searches an engine ran for a question showed up in all three runs of that same question. 78% appeared in only one run.
164 question and engine pairs run 3 times each · 5 auditsRead together they say something plain. A tool that watches one engine has measured one engine, and tells you close to nothing about the other five. A tool that asks once has caught one shuffle of a deck that gets reshuffled on every question.
This is also why a competitor who looks dominant in a single screenshot often is not, and why a bad result in one check is not yet a problem. Nothing in our report counts as a finding until it repeats. The reasoning is written up in why we run every prompt three times.
Scope, so the numbers travel honestly: both are computed from our own client mix, which skews to legal, finance, home services, SaaS and ecommerce at US locale. The first counts registrable domains rather than individual URLs. Neither is a controlled experiment, and we would not present them as one.
Want this run on your own questions?
Get my free auditWhat the free auditcovers.
The instrument behind it is Citrion, which we built and run in-house. The full protocol is public: how we measure AI visibility, including the evidence tier attached to every number in the report.
What this auditwill not tell you.
Worth knowing before you ask for one, because the gap between what this measures and what people assume it measures is where most disappointment lives.
- How many customers came from an AI answer. Google reports AI Overview and AI Mode clicks inside ordinary organic totals, so nobody can separate them. We estimate with a stated method and label it an estimate, never a count.
- What every user sees. Assistants personalise, and ours run with memory off on our own accounts. The audit measures a clean baseline, not your specific buyer's screen.
- A guaranteed outcome. We cannot promise a ranking position, a citation, a traffic number or a lead volume, and anyone in this field who does is guessing.
- That schema markup is the fix. It is usually not. We add it because it is cheap and helps Google's classic results, not because it moves citations. The evidence is in does schema markup get you cited.
What it does give you is a measured starting position and a list of the specific questions where a competitor is being named and you are not. That is the thing you cannot buy back later, because once you start making changes the clean before shot is gone.
Questions people askbefore they book one.
How much does an AI visibility audit cost?
The first one is free for qualified businesses. No card, no obligation. Ongoing tracking of you and your competitors sits inside the paid plans, and you can see what those cost on the pricing page.
How long does it take?
Five business days from the point where the question list is agreed. The list comes first on purpose. An audit built on the wrong questions measures the wrong market, and that is the most common way these go wrong.
Do you need access to my website or analytics?
Not for the free audit. It reads public answers from public assistants, so all we need is your business name and your domain. Paid work later may ask for Search Console and analytics access, and that is a separate conversation.
Is this the same as an SEO audit?
No. An SEO audit examines your site: crawling, indexing, speed, titles, links. This examines the answers, meaning what assistants say about you and who they cite instead. They overlap because assistants search the web to answer, so ranking still decides much of what they can find. The longer comparison is here.
Will schema markup or an llms.txt file fix this?
On the evidence, no. Ahrefs tested 1,885 pages that added JSON-LD against roughly 4,000 matched controls and found AI Overview citations moved by -4.6 percent. A separate study of 137,210 domains found 97 percent of llms.txt files received zero requests across a month. Both cost almost nothing and we add them anyway. Neither one is a strategy, and we will not bill you for either as though it were.
What if the audit shows I am already doing well?
Then we tell you that and you keep your money. It happens, usually with established brands that have been publishing for years. We would rather say so than sell you a fix for a problem you do not have.
Get the measurementbefore you change anything.
Ten buyer questions, five assistants plus Google's AI answers, every question run three times. You get the score, the competitors who beat you, and the exact sentences the assistants use to describe you.