Can AI search tools reach your site at all?
Check access first, because a blocked crawler makes every other GEO effort pointless.
Many sites block AI crawlers by accident, through a CDN bot setting or an old robots.txt rule. Each tool uses its own crawler, and several vendors separate search crawling from training.
| Crawler or token | What it is for (per vendor docs) | Block it? |
|---|---|---|
| OAI-SearchBot | Surfacing sites in ChatGPT search | Allow if you want ChatGPT citations |
| GPTBot | Collecting content that may train OpenAI models | Your choice; independent of search |
| PerplexityBot | Surfacing and linking sites in Perplexity | Allow if you want Perplexity citations |
| Bingbot | Bing index, which Copilot draws on | Allow |
| Googlebot | Google Search, including AI Overviews and AI Mode | Allow |
| Google-Extended | Gemini training and some grounding; not Search | Your choice; no effect on Search |
Run your domain through the GEO readiness checker to see how your robots.txt treats these agents.
What makes a page more likely to be cited?
A page that states a clear, specific, current answer, from a source the system can identify and trust.
We rewrite priority pages so each question heading is followed by a direct answer, supporting facts, a date and a source. Our article on how ChatGPT search picks sources covers what is publicly known about selection. Google's guidance is plain: no special markup or files are required for AI Overviews, only that the page is indexed and eligible for a snippet.
Why do entities and third-party mentions matter?
AI systems describe brands using what many sources say about them, so consistent facts across the web shape the answer.
We align your Organization schema, About page, profiles and directory listings so they tell one story. Then we plan earned mentions in reviews, comparison articles and trade press, since assistants often cite those instead of your own site.
How do you track something that changes every run?
We sample: a fixed prompt set, run on a schedule across several tools, reported as a share of appearances with the variance shown.
One screenshot proves nothing. Fifty prompts run monthly shows a trend. We pair that with AI referral traffic in analytics, as described in tracking AI search traffic in GA4, plus Bing Webmaster Tools and IndexNow so Bing, and Copilot with it, see changes quickly. Where Search Console offers AI-specific views, we treat them as directional; reported AI data is mostly impressions, not clicks or revenue.
Scope depends on site size and how many prompts and tools you want tracked. Request a proposal for a scope and timeline range.
Frequently asked questions
Can you guarantee we will be cited by ChatGPT or Google AI Overviews?
No, and no one honestly can. These systems pick sources per query and the same prompt can return different sources on different runs. We improve the conditions for being cited and measure the share of sampled prompts where you appear.
If we block Google-Extended, do we disappear from Google Search?
No. Google states that Google-Extended does not affect inclusion in Google Search and is not a ranking signal. It controls use of content for some Gemini training and grounding. AI Overviews and AI Mode rely on normal Googlebot crawling and indexing.
Do we need an llms.txt file?
Google's AI features guidance says you do not need machine-readable or AI text files to appear in Google's AI features, so it adds nothing for Google Search. Whether other assistants rely on it is unproven as of November 2026. It is cheap to add, so we treat it as optional rather than a priority.
Can we block AI training but stay in AI search?
Often, yes. OpenAI documents GPTBot (training) and OAI-SearchBot (ChatGPT search) as independent settings, and Perplexity says PerplexityBot surfaces sites in search results and is not used for foundation model training. Each vendor's documentation should be checked before you change robots.txt.
How do you measure AI search visibility?
We run a fixed set of buyer prompts on a schedule and record citations and mentions, add AI referral sessions from analytics, and read the AI-related data Google and Bing expose in their webmaster tools. Google counts AI feature traffic inside overall Search Console web data.