What an AI visibility report shows about your business.
An AI visibility report asks ChatGPT, Claude and Perplexity the questions your customers ask, and records who gets named: you or your competitors. It shows where you stand today and gives you something to measure changes against. It can't give you a ranking, because AI tools don't keep stable ones.
Plenty of tools track where you rank on Google. There's nothing like that for AI answers. No company publishes who ChatGPT recommends, and the answers change every time someone asks.
A visibility report is the practical way round that. You ask the tools what your customers would ask, and you write down what they say.
It starts with the questions your customers ask.
The questions matter more than anything else in the report. They should be the ones real buyers type, in their words, about the work you want more of.
For a commercial roofing contractor in Denver, they might look like this:
| Kind of question | Example | What it tells you |
|---|---|---|
| Who's best | "Who are the best commercial roofing contractors in Denver?" | Whether you're on the shortlist |
| Who can do it | "Who can replace the roof on a 60,000 square foot warehouse near Aurora?" | Whether your services are clear enough to match |
| About you | "What does [your company] do and where do they work?" | Whether the tools know you and describe you correctly |
| Comparison | "Is [your company] or [competitor] better for school roofs?" | How you come across next to a named rival |
We write 10 to 20 of these with you. We lean on the first two kinds, because that's where new work comes from.
Then we ask ChatGPT, Claude and Perplexity.
Each question goes to each tool, and we record who gets named. For every question you can see whether you came up and which competitors did.
Read what the tools say about you as well. A wrong service area, an old address or a service you stopped offering years ago all point to pages that need fixing, on your site or elsewhere. The sources a tool cites are worth a look too. If the same directory or review site keeps coming up, your listing there matters.
One answer on its own tells you little.
AI answers vary from run to run. In a study published in January 2026, SparkToro and Gumshoe had 600 volunteers put 12 prompts through ChatGPT, Claude and Google's AI 2,961 times. The chance of ChatGPT or Google's AI returning the same list of brands twice was below 1 in 100. Getting the same list in the same order was closer to 1 in 1,000.
The same study found something steadier: how often a brand appears across many questions and runs. That's the number to watch. Your position in any one list tells you very little, so be wary of any tool or agency that reports an "AI ranking position" as if it were a Google ranking.
Answers also depend on who's asking. OpenAI says ChatGPT may use a person's approximate location to answer local questions, so a buyer in Aurora may see a different list from one in Boulder. A report is a sample taken on a given day. It's most useful for comparing one month with the next.
For more on why the lists move, read how ChatGPT, Claude and Perplexity decide who to recommend.
Each gap points to a fix.
The report is only useful if it leads somewhere. Most findings point to something specific:
| What the report finds | What usually needs fixing |
|---|---|
| A tool doesn't seem to know your business at all | Its crawler can't get in, or your pages say too little about who you are |
| You're known, but not named for your main service | A clear page for that service that says where you do it |
| You're described with an old address or old services | The pages, listings or profiles that still carry the old details |
| Competitors are named and the same directory is cited each time | Your own listing on that directory |
| You're named in one town and missing in the next | Your service areas, written out on your site and your listings |
The fixes on your website are what the tune-up covers. The ones elsewhere, like directory listings and reviews, are yours to follow up, and the report shows where to start.
We run it before the work and after it.
The report is part of the AI Search Tune-Up, which starts at $1,500 for a site of up to 30 pages. We run it before we change anything, so you have a starting point, and again after the work.
Don't expect the second report to look different straight away. OpenAI says a change to your robots rules takes about 24 hours to reach ChatGPT's search. Changes to your pages take longer to show, because each tool reads them again on its own schedule, and the answer also depends on what else it finds about you.
That's what monthly monitoring is for. For $149 a month we rerun the same questions every month and send you a one-page report of what moved.
What a report can't tell you.
It can't promise future mentions. Nobody controls whether an AI tool names a business, and the tune-up doesn't promise that either.
It can't cover every question a buyer might ask. It's a sample, so the questions need to be the right ones.
And it can't separate the effect of things off your site, like a run of new reviews or a piece in a trade magazine, from the effect of the website work. What it can show is where you stand for the questions that matter to you, who you're up against, what the tools believe about your business and whether that changes over the months.
To see the website side of the picture first, run the free check. In about 30 seconds it shows whether AI tools can get into your site and read it, and what the tune-up would cost for a site your size.