Can ChatGPT find your website? AI visibility, free on every GhostStat plan
More people ask an AI assistant instead of searching. GhostStat now shows, on the dashboard of every site, who those assistants send you, which AI crawlers read your pages and whether your site lets them in. Free on every plan.
More and more people ask ChatGPT, Perplexity, Copilot, Gemini or Claude instead of typing a search into Google. That raises a new question for anyone with a website: can those assistants find you, and do they send you anyone? GhostStat now answers that on the dashboard of every site you track. We call it AI visibility, and it is free on every plan, the Free plan included.
Why AI visibility matters now
When an assistant answers a question, it often names a few websites and links to them. A visitor who clicks that link arrives with a question already half answered, so these visits tend to be few but focused. If your site is easy for AI to read, you have a chance to be one of those links. If your site blocks AI by accident, you are left out without ever knowing it.
Most analytics tools file these visits under “referrals” next to every other website, or under “direct” when the assistant sends no referrer. So the question “does ChatGPT send me visitors?” is hard to answer. AI visibility puts it in one panel.
What the AI visibility panel shows
Open the dashboard of any site and the panel sits right under Traffic quality. Pick the last 30, 90 or 365 days. It answers three questions.
1. Who do AI assistants send you?
Every visit whose referrer or utm_source is an AI assistant is counted per assistant: ChatGPT, Perplexity, Copilot, Gemini, Claude, DeepSeek, Grok, Meta AI, Le Chat, You.com, Phind and Poe. You see the visits and the people per assistant, the change against the period before, and the pages those visitors landed on. That last list is the useful one: it shows which of your pages AI already recommends.
2. Which AI crawlers read your pages?
AI companies use several crawlers, and they don’t all do the same job. GhostStat groups the ones it recognises by what they do:
- Training, such as GPTBot and ClaudeBot: they collect pages that may be used to train AI models.
- AI search, such as OAI-SearchBot, Claude-SearchBot and PerplexityBot: they read pages so an assistant can show them in its answers.
- On request, such as ChatGPT-User and Claude-User: they fetch one page because someone asked the assistant about it right then.
- Search engines, such as Googlebot and Bingbot, shown alongside for comparison.
For each crawler you see how often it came, how many pages it read and when it was last there, plus the pages AI crawlers read most. One honest note: GhostStat sees a crawler when it loads a page and runs the tracking script, the way a browser does. Many crawlers fetch only the HTML and never run scripts, so treat these numbers as a minimum. Your server’s access log has the full count.
3. Does your site let AI in?
GhostStat reads your site the way an AI crawler would, at most every 12 hours or when you press Check the site again. It looks at:
- whether the home page loads;
- whether your
robots.txtlets the AI search and on-request crawlers in, checked bot by bot with the same rules crawlers use; - whether Googlebot and Bingbot may read the site, because several assistants lean on those indexes;
- whether the home page is indexable (no
noindex) and allows snippets; - whether there is a sitemap, a canonical link, structured data (JSON-LD) and a meta description;
- whether there is an
llms.txt.
The result is one score, the AI-ready percentage, with every check listed so you can see what to fix. Blocking training crawlers is your choice, and it doesn’t lower the score: you can say no to training and still be found in AI answers. Blocking the AI search or on-request crawlers does lower it, because then assistants can’t read your pages when someone asks.
Training versus answers: you can choose
This split matters, and the AI companies themselves draw it. OpenAI says GPTBot collects pages that may be used for training, OAI-SearchBot decides whether a site can appear in ChatGPT search answers, and the two settings are independent. Anthropic describes the same three roles for ClaudeBot, Claude-SearchBot and Claude-User. Google uses a separate Google-Extended token in robots.txt: it controls whether Google may use your pages to train Gemini models and to ground answers in Gemini apps, and it does not affect Google Search.
So a robots.txt like this keeps your pages out of training but lets assistants read them to answer questions:
User-agent: GPTBot
Disallow: /
User-agent: ClaudeBot
Disallow: /
User-agent: *
Allow: /Adding Google-Extended to that list also keeps your pages out of answers in Gemini apps, so weigh that one separately. The panel shows which group each rule hits, so you can check that a robots.txt change does what you meant.
What llms.txt is, and the draft GhostStat writes for you
llms.txt is a short plain-text guide to your site written for AI assistants: a name, one line on what the site offers and a list of the pages that matter most. It was proposed by Jeremy Howard of Answer.AI in September 2024 and sits at the top of your site, next to robots.txt.
It is a proposal, not a rule crawlers must follow, and the large AI companies have not said their crawlers use it. That is why it doesn’t count in the AI-ready score: letting AI search crawlers in and being indexed matter far more. It costs little to add, though, so the panel writes a draft for you. It is built from your home page title and description and your most visited pages in GhostStat. Fill in the line in brackets, then copy it or download it as llms.txt and put it in the main folder of your site.
What to do with the numbers
- Score below 85%? Open the checks. A missing sitemap or meta description is a quick fix. A robots.txt that blocks OAI-SearchBot or PerplexityBot is usually an old “block all bots” rule doing more than you meant.
- Few AI visits, many AI crawler reads? Assistants read you but rarely link to you. Look at the pages they read and make sure each one answers a clear question in its first paragraph.
- A few pages get all the AI visits? Those are the pages AI trusts. Link from them to the pages you want found next.
- Want more detail? The AI Crawlers plugin shows every crawler visit, a score per bot and the crawl patterns over time. The features page has both side by side.
Free on every plan
AI visibility needs nothing switched on and uses no allowance: it is on the dashboard of every site, on every plan, from the first day. The same is true for Traffic quality, the panel right above it that shows how many of your visits were real people. Create a free account, add your site and the panel fills as visits come in. Plans and allowances are on the pricing page.
New to GhostStat? Read how cookieless analytics works without a cookie banner.
Sources
Frequently asked questions
Can I see whether ChatGPT sends me visitors?
Yes. The AI visibility panel on your GhostStat dashboard counts every visit whose referrer or utm_source is ChatGPT, Perplexity, Copilot, Gemini, Claude or another assistant, and shows the pages they landed on.
Is AI visibility a paid extra?
No. It is on the dashboard of every site on every plan, the Free plan included, and uses no allowance. The deeper AI Crawlers plugin follows the normal plan rules.
Does blocking GPTBot hide my site from ChatGPT?
Not from ChatGPT search. OpenAI says GPTBot is for training and OAI-SearchBot decides whether a site appears in ChatGPT search answers, and the two settings are independent.
Why does GhostStat show fewer AI crawler visits than my server log?
GhostStat sees a crawler when it runs the tracking script on a page. Many crawlers fetch only the HTML, so the panel shows a minimum. Your server access log has the full count.
Do I need an llms.txt?
It is optional. It is a proposal, not a rule crawlers must follow, so it does not count in the AI-ready score. GhostStat writes a draft you can copy if you want one.
