AI brand visibility checker
An AI brand visibility checker asks whether models name your company, not whether they can read your pages. It runs twelve fixed buyer prompts through a model, none of which contain your brand name, then reports your mention rate, your average position when named, and the competitors that get named instead.
By The Promptscope team (Product and testing). Last updated .
How to check whether models name your brand, step by step
- Type your domain into the box above, for example basecamp.com, and start the readiness audit.
- Wait for the twelve readiness checks, because a blocked crawler caps your score and skews everything after it.
- Then open the mention check, confirm the category your buyers would actually name, and run the twelve prompts.
- Finally, save the mention rate and the competitor list, since the numbers only mean something next to an earlier run.
The first 3 runs each day are free. After that, see the pricing page for packs and plans, or read how we test the tool.
AI brand visibility checker run against a real site
We pointed this checker at basecamp.com and kept what came back. So the card below is an actual report rather than a mockup, and you can run the same check yourself and compare.

- Site we checked
- basecamp.com
- Readiness score
- 95 / 100
- Checks run
- 12
- Needs a fix
- 0
- Warnings
- 2
- AI crawlers allowed
- 8 of 8
- Checked on
- 2026-10-16
How we test: every page runs the same checks against a different real site, and we read each report by hand before publishing it. More on the about page.
Readable and named are two different outcomes
Your site can pass every technical check and still never come up in an answer. Readability is only the floor, because it decides whether an engine can use your pages at all. Getting named is a separate outcome, since it also depends on what the rest of the web says about you. Most tools blur those two things together, so a green dashboard ends up promising something it never measured.
Therefore the two halves stay apart here. The free readiness audit scores your markup, your crawler rules and your opening sentence, and it calls no model at all. The mention check then asks a model real buyer questions and writes down which brands come back. Run the audit first, since one stray robots.txt line quietly caps everything the second half could ever show you.
Inside the AI brand visibility checker: twelve buyer prompts
The AI brand visibility checker sends twelve fixed prompts to a model. They read the way a buyer actually types, for example what is the best option in a category, or which one would you recommend for a small team. Some ask for a shortlist, while others ask for a single pick or a comparison. The wording never changes between runs, so a check in March and a check in June stay comparable.
None of the twelve prompts contain your brand name, and that rule carries the whole method. An answer that names you only because you were named in the question measures nothing at all. Instead, the model has to reach for you on its own. The report then records your mention rate across the twelve answers, your average position when you do appear, the competitors named in your place, and the domains the model cited while answering.
What an AI brand visibility checker needs from your markup
A model does not remember a URL. It remembers an entity, which means a name attached to a category and a description. So the first fix is boring and unglamorous: pick one brand name and use it everywhere. A company that writes itself three ways across its own footer, its pricing page and its support docs has split its own signal into three weaker ones.
Next, make the machine readable version match. Organization markup in JSON-LD states your name, your URL and your logo in a format an engine does not have to guess at. In addition, a plain sentence near the top of your home page should say what you sell and who buys it. Vague positioning hurts here, because a model that cannot place you in a category will not offer you when a buyer asks about that category.
What this AI brand visibility checker cannot tell you
A mention check is a sample of model answers at one moment, and it is not a ranking. Models update, temperature varies, and the same prompt can return a slightly different shortlist an hour later. Therefore one run is a reading, not a verdict. Treat a single result the way you would treat one poll, and watch the direction across several runs instead of celebrating or mourning any single number.
The scope is narrow in other ways too. It checks one URL at a time, so a large site with many templates needs a crawler rather than a home page snapshot. It also reads the HTML your server returns, which means content that appears only after JavaScript runs will look missing. Finally, a strong readiness score never promises a recommendation, although it does remove the reasons a model might skip you.
What the mention check reports after twelve prompts
| Result | How to read it |
|---|---|
| Mention rate | how many of the twelve answers named your brand at all |
| Average position | where you land in a list when you do get named |
| Competitors named | the brands a model reaches for instead of yours |
| Cited domains | the pages a model leaned on while it answered |
| Prompt plan | the exact twelve questions, shown before any run costs credits |
| Brand name absent from prompts | keeps the result honest, since you cannot prompt yourself into it |
| Category wording | the label you compete under, which changes who shows up |
| Run date | the anchor for comparison, because a single reading proves little |
AI brand visibility checker FAQ
How is this different from the free readiness audit?
The readiness audit asks whether an engine can fetch and understand your pages. This asks whether a model actually says your name when a buyer asks a question. One is about your site, and the other is about your reputation across the web. Both matter, but fixing the first is far more within your control.
Why do the prompts leave out my brand name?
Because a prompt that mentions you guarantees an answer that mentions you. That result would look great and mean nothing. The twelve questions describe a category and a buying situation instead, so any mention you earn is a real one.
What counts as a good mention rate?
There is no universal number, since a crowded category behaves nothing like a narrow one. Use your own first run as the baseline and watch the trend. Also read the competitor list carefully, because seeing which brands get named in your place usually tells you more than the percentage does.
Will the same run give the same answer tomorrow?
Not exactly. Model versions change, and answers vary slightly even with identical prompts. So we keep the twelve questions fixed and date every run, which makes the comparison between runs meaningful even though no single run is definitive.
Why does the mention check cost credits?
Each run sends twelve prompts to a model, and that has a real cost on our side. The readiness audit calls no model, therefore it stays free with no account. When you are out of free runs we show you the prompt plan and call nothing, so you never see a blurred or invented result.