How to get cited by AI: what makes ChatGPT, Gemini and Perplexity link to your page

, SEO and AI visibility · Updated

To get cited by AI, your page has to be reachable by the engine's search crawler, present in the index that engine searches, and contain a short passage that supports part of the answer better than the alternatives. For buying questions you also need to be named on the third-party pages the engines already cite, because many citations go to roundups, reviews and forums rather than to vendor sites.

This guide covers the work that applies to every engine. For engine-specific steps, see how to rank on ChatGPT, Perplexity SEO and Gemini SEO.

A mention is not a citation

AI answers can reward you in two ways. A mention is your brand named in the text. A citation is a link to your page shown as a source. They often come apart: an engine can recommend you by name while linking to a review site, or link your blog post without naming your product. Track both, and treat the citation as the one that sends visitors.

Citations only exist when the engine looks something up. When an answer comes from the model's training alone, there is no source list to appear in. So the work below is about the moment of retrieval: the engine runs a search, reads a handful of pages, writes an answer and links the pages it used.

How each engine finds the pages it cites

Each company runs its own crawlers and documents them separately. Training crawlers and search crawlers are different bots, and that is the detail most robots.txt files get wrong.

EngineWhat fetches pages for answersWhat the vendor says
ChatGPT searchOAI-SearchBot; ChatGPT-User for fetches a user asks forSites opted out of OAI-SearchBot "will not be shown in ChatGPT search answers" (OpenAI)
Google AI Overviews and AI ModeGooglebot and the Google Search indexA page must be "indexed and eligible to be shown in Google Search with a snippet" (Google)
Gemini appContent from the Google Search index at prompt timeThe Google-Extended token covers Gemini training and grounding (Google)
PerplexityPerplexityBot for its index; Perplexity-User for user requestsAllow PerplexityBot so your site appears in its results (Perplexity)
ClaudeClaude-SearchBot; Claude-User for user requestsBlocking Claude-SearchBot "may reduce your site's visibility and accuracy in user search results" (Anthropic)

Blocking a training crawler does not remove you from answers. You can disallow GPTBot or ClaudeBot and keep OAI-SearchBot and Claude-SearchBot allowed. The opposite mistake is common: a blanket rule meant to keep out "AI bots", often set in a CDN or firewall rather than robots.txt, takes the search crawlers out too. When you fix it, the change is not instant: OpenAI says robots.txt updates take about 24 hours to reach ChatGPT search (OpenAI), and Perplexity says up to 24 hours (Perplexity).

Our free AI crawler checker shows what your robots.txt allows for each of these bots, and the robots.txt generator has a preset that allows AI search crawlers while blocking training crawlers. Neither can see firewall rules, so check those separately.

What makes a passage worth citing

Once the engine can fetch your page, it still has to pick it over the other results. Google's guidance for its AI features says "the best practices for SEO remain relevant" and that there is no special schema or AI text file to add (Google Search Central).

Beyond the basics, these changes make a page easier to cite.

Answer one question per section

Engines quote passages, not whole pages. Give each section a heading phrased the way people ask, and answer it in the first one or two sentences, so the passage still makes sense when it is lifted out with nothing around it. Our answer engine optimization guide has a before and after example.

Put a number, a date or a source in the claim

"XeoRank Starter costs $19 a month for one site" (pricing) can be repeated as an answer. "Affordable plans for every team" cannot. The GEO paper by Aggarwal et al. (2023) found that adding quotations, statistics and cited sources raised visibility in generated answers by up to about 40% on the authors' benchmark, while keyword stuffing did little. Link the source in the same paragraph as the claim, so the reader and the engine can both check it.

Say something other pages cannot

If ten pages say the same thing, the engine can cite any of them. In May 2025 Google's John Mueller advised publishers to focus on "unique, non-commodity content" for Google's AI experiences (Google Search Central blog). For a business site, that material is usually close to hand: your real prices and limits, results from your own tests, setup steps you have timed yourself, the questions customers ask on sales calls, and plain statements of what your product does not do.

Cover the sub-questions, not only the head term

Google says AI Overviews and AI Mode may use "query fan-out", issuing multiple related searches across subtopics and data sources, which lets them show "a wider and more diverse set of helpful links" (Google Search Central). A question like "best payroll software for a small agency" may be answered from pages about pricing, integrations and contractor payments. A focused page that answers one of those sub-questions well gives the engine a reason to cite you even where a larger site holds the head term.

Keep the text readable and the markup honest

Google's best practices for AI features include "making sure that important content is available in textual form" and that structured data "matches the visible text on the page" (Google Search Central). If your copy only appears after JavaScript runs, check what the raw HTML contains, because you cannot assume every AI crawler runs scripts.

Get named on the pages AI already cites

For commercial questions, your own page is one candidate among many. Profound's analysis of 680 million citations, covering August 2024 to June 2025, found Reddit was the most-cited source for both Perplexity (6.6% of citations) and Google AI Overviews (2.2%). Our own check points the same way: when we asked ChatGPT, Gemini and Perplexity for a "cheaper alternative to Profound or AthenaHQ" on 13 October 2026, the answers cited about 30 different third-party alternative, review and pricing pages.

Collect the cited URLs for your buyers' questions, keep the ones that recur across a few runs, and sort them into pages you can update, pages whose authors you can contact with current facts, and communities where you can answer questions under your own name. The ChatGPT guide walks through that list step by step. Do not buy reviews: the US Federal Trade Commission's 2024 final rule bans buying and selling fake reviews, including AI-generated ones.

Check whether it worked

  • Ask the same buyer questions on each engine and record mention and citation separately. Profound found that how long a page stays cited on one engine says almost nothing about how long it stays cited on another, with correlations of 0.03 to 0.09; in its words, "The engines decide independently."
  • Expect churn. The same Profound post puts the half-life of an AI citation at about 11 days, so judge a change over several checks, not one.
  • Use the reports the platforms give you. Bing Webmaster Tools added an AI Performance report in public preview on 10 February 2026; it counts citations of your pages in Microsoft Copilot and Bing's AI summaries and shows a sample of the grounding queries. Google includes AI Overviews and AI Mode traffic in Search Console's Performance report, under the Web search type, as part of your overall search traffic (Google).
  • Watch referral visits from AI assistants in your analytics. A citation that never sends a visit is still worth knowing about, but referrals tell you which answers people act on.

Where XeoRank helps, and where it does not

XeoRank asks ChatGPT, Gemini (grounded in Google Search) and Perplexity your buyer questions and shows the result for each engine, next to an overall citation rate. Starter is $19 a month and Pro is $49 a month; see pricing for what each plan includes. Its GA4 reporting breaks out AI referral traffic from ChatGPT, Perplexity, Claude, Gemini and Copilot and compares it with the previous period. The site crawler flags errors in the structured data a page already has, pages whose HTML carries only the template because their content is fetched client-side, and missing titles, meta descriptions and H1s. You or your AI assistant can then fix pages on WordPress, Framer or GitHub and Vercel code sites through the MCP server; your AI assistant asks before each tool call, and WordPress writes stay off for a site until you turn them on.

It cannot guarantee a citation, and nobody can. It does not do your outreach, it does not estimate prompt volume, and it does not track Copilot, Grok or DeepSeek. To see where you stand without an account, run the free three-question Gemini check at AI visibility.

Questions

How do I get cited by ChatGPT?
Allow OAI-SearchBot in robots.txt and in your firewall, because OpenAI says sites opted out of it are not shown in ChatGPT search answers. Then answer your buyers' questions plainly with sourced facts, and get named on the roundups and review pages ChatGPT already cites. Our guide to ranking on ChatGPT covers each step.
Does blocking AI training bots stop AI citations?
Not if you block the right ones. GPTBot and ClaudeBot collect training data, while OAI-SearchBot and Claude-SearchBot fetch pages for search answers, so you can block the first pair and allow the second. Google-Extended covers Gemini training and grounding but does not affect inclusion in Google Search, which is where AI Overviews appear.
Why does one AI engine cite me and another does not?
Each engine runs its own crawler, searches its own index and picks its own sources, so they cite different pages. Profound has measured very low overlap between engines. Check each engine separately and look at which pages it cites instead of yours.
Do I need schema markup or an llms.txt file to get cited by AI?
Google says there is no special schema and no AI text file needed to appear in AI Overviews or AI Mode. Structured data still helps machines read a page when it matches the visible text, but it does not earn citations on its own.
How long does it take to get cited by AI?
Crawler changes are quick: OpenAI and Perplexity both say robots.txt updates take up to about a day to reach their systems. Earning citations for competitive questions takes longer, because it depends on indexing, on how your page compares with the alternatives, and on other sites mentioning you.

Related

See where you stand in Google and AI answers

Starter is $19 a month for one site, with a 14-day free trial and no card to start.