// guide
How do you get cited by ChatGPT?

Written by: Dmitry Filippov, Founder, GET-GEO.AI
Published: 2026-07-31 · Updated: 2026-11-18
// short answer
To be cited by ChatGPT, work on three things: crawler access for OAI-SearchBot, the only OpenAI agent that decides whether a site appears in search answers; server-rendered pages where one question's answer comes first and contains a checkable fact; and corroboration, independent sites describing you the same way. Access takes a day, pages weeks, corroboration quarters. Placement is never guaranteed.
- Only OAI-SearchBot decides whether ChatGPT can cite you; OpenAI documents four user agents, and GPTBot (training) and ChatGPT-User have no bearing on search eligibility.
- OpenAI's search help page says ChatGPT rewrites a question into one or more targeted queries and lists two eligibility conditions; placement is not guaranteed.
- Ahrefs found 28.3% of ChatGPT's 1,000 most-cited pages had no organic keywords, and 76.4% of the dated ones were updated within the previous month.
- The page format we use is one buyer question as heading and a 40–60-word answer with a checkable fact; in our September 14, 2026 run ChatGPT cited provider pages in all nine fintech answers.
- Plan five layers on five horizons: crawler access about a day, rendering about a week, pages weeks to a quarter, entity weeks to a quarter, corroboration quarters.
Which OpenAI crawler decides whether you can be cited?
Only one of OpenAI's crawlers decides whether ChatGPT can cite you: OAI-SearchBot. OpenAI documents four user agents; three of them matter to a publisher (the fourth, OAI-AdsBot, only visits pages submitted as ads), and the crawler documentation is explicit about which one matters for search. OAI-SearchBot builds the index behind ChatGPT search; sites that opt out of it "will not be shown in ChatGPT search answers, though can still appear as navigational links". GPTBot crawls for model training and has no bearing on whether you are cited. ChatGPT-User fetches a page when a user's action calls for it, but OpenAI says it "is not used to determine whether content may appear in Search" and that robots.txt rules may not apply to it at all.
So the earlier advice, ours included, that you must allow both the search agent and the user agent was wrong: the decision rests on OAI-SearchBot alone.
The robots.txt file is not the whole gate. OpenAI also asks that your host or CDN (the service in front of your site that filters traffic) allow requests from the IP ranges it publishes for OAI-SearchBot, so check there as well: an anti-bot rule at the CDN can block the crawler before it ever reads the file, even when robots.txt is clean. After you change robots.txt, OpenAI says its systems take roughly 24 hours to adjust, so checking the same evening proves nothing.
Those two conditions are the whole eligibility list on OpenAI's search help page; there is no submission form (more on that in the FAQ below). If you block pages on purpose, keep the limit quoted above in mind: a site opted out of OAI-SearchBot stays out of ChatGPT search answers but can still appear as a navigational link. Until October 2026 an OpenAI help page for publishers added a remedy: it said a disallowed page's link and title could still surface in ChatGPT Atlas, OpenAI's browser, and advised a noindex meta tag to prevent that, with the crawler allowed to read the page so it can see the tag. OpenAI has since retired Atlas and withdrawn the page (the Internet Archive holds a copy dated 4 October 2026), and neither its crawler documentation nor its search help page mentions noindex, so treat the tag as former guidance. For the complete list of AI crawlers and ready-made configurations, see our guide on robots.txt for AI crawlers.
- OAI-SearchBot — indexes for ChatGPT search. Allow it, and allow OpenAI's published IP ranges at the CDN. This is the one that decides.
- GPTBot — crawls for training. Blocking it keeps your content out of training and does not remove you from search.
- ChatGPT-User — fetches a page on a user's request. Does not determine search eligibility; robots.txt may not apply.
Why does the HTML you serve matter more than your framework?
The HTML your server returns matters because it is the one version of the page every crawler is certain to receive, whatever framework produced it. OpenAI does not publish how much JavaScript its search crawler executes, and we have not measured it, so we work on the assumption that text which appears only after the browser has run your JavaScript may not be read. Treat this as a working assumption, not a documented rule, and test your own pages rather than trust it.
The test for ChatGPT citations costs nothing: open the page with JavaScript disabled, or request it with curl, and check that the sentence you want quoted is in the markup. While you are there, check the rest of what a crawler sees: a 200 status without a redirect chain, no noindex on pages you want cited, and one canonical host, meaning a single hostname (with or without www, https only) that every internal link and every external profile points to. A brand split across two hosts is two half-brands to a retrieval system.
Every guide on this site is rendered on the server and can be read with JavaScript disabled. This gives crawlers access to the content before the work on individual pages, independent coverage and brand information described below.
What does a page ChatGPT actually quotes look like?
The page format we use for ChatGPT is one buyer question as the heading, a 40–60-word answer with a checkable fact directly beneath it, and the detail after. It is a working format, not a threshold OpenAI documents; what we saw on pages that were actually cited is set out further down. To see why we build pages this way, start with how ChatGPT chooses sources: it searches rather than recalls. OpenAI's help documentation says that when ChatGPT hands a question to a search provider it typically rewrites it into one or more targeted queries, may send further, more specific queries after reading the first results, and sometimes partners with other search providers. It also says plainly that ChatGPT ranks results "using multiple factors" and that placement is not guaranteed.
The rewriting step is what your page has to survive. OpenAI's own example is a broad question about a drug programme being turned into a query naming the specific compound and the year: the model is looking for a specific fact, so start with the page that states it plainly rather than with a general article on the topic. Our September 14, 2026 fintech run below points the same way: in all nine answers ChatGPT cited providers' own pages, among them pages built around one checkable fact.
Google ranking is not the filter people assume it is. Ahrefs looked at the 1,000 pages ChatGPT cited most often (October 2025) and found that 28.3% of them had no organic keywords in Ahrefs' index. Of the 564 pages with a determinable date, 76.4% had been updated within the previous month. Two conclusions for your pages: you do not need to rank first in Google to be quoted, and the pages ChatGPT cites tend to carry a recent, visible date next to specific claims.
To get ChatGPT to cite a page of your website, we advise a simple pattern: a heading phrased as the question a buyer actually asks; immediately under it, a compact answer of 40–60 words that names the subject ("a UK business account at Starling", not "it" or "this option") and states the fact with a number, a date or a named source; then the detail. Open each page with the answer to one question; related sub-questions can follow under their own headings. In our runs of 14–15 September 2026, a contract guide built as a series of clause questions, each under its own heading, was cited by ChatGPT, Perplexity and Google AI Mode, in five of the seven answers to a question about what a GEO contract should say. When we opened 11 of the pages cited in those runs on 7 October 2026, the most common trait was a question heading answered in its first sentence, on five pages. The six pages ChatGPT had cited were more varied: one built that way, two service pages that open with a sentence on what the company does and for whom, and three with comparison or verdict tables. Length ran from about 900 to 5,400 words, and authorship and dates varied. Eleven pages from one prompt set show what cited pages looked like, not what caused the citation.
Our own run data shows what this looks like from the model's side. On September 14, 2026, we put three fintech buyer prompts to ChatGPT in signed-out sessions (a UK business account, a neobank for freelancers in Spain, the cheapest UK-to-India transfer), three repeats each, nine answers in total. The providers' own pages were cited in all nine, and the pages the model picked were not "About us": Starling's and Monzo's published service-quality survey results, and N26's account terms for self-employed customers in Spain. Pages with a checkable fact for a specific segment. To see which of your pages ChatGPT quotes today, run the signed-out test protocol from our guide on checking your AI visibility for free.
Which prompts go to other people's pages, and why?
In the data below, two kinds of prompt go to other people's pages: shortlist prompts of the "best X" kind and searches for a local service. The working explanation is corroboration: a recommendation is easier to support with independent sources that describe a vendor the same way the vendor describes itself. Ahrefs quantified the shortlist pattern for ChatGPT in December 2025 across 750 "best X" prompts and 26,283 cited source URLs: "best X" blog lists were 43.8% of all page types cited. For a question like "best agency" or "best tool", treat your own site as the claim and a directory, a review platform or a comparison article as the evidence.
Our own runs show where that pattern stops. In the same September 14, 2026 run, across the 24 signed-out ChatGPT answers on GEO-buyer prompts, the type of question decided the type of source. On prompts about what happens if GEO fails after 90 days, or what guarantees agencies offer (six answers), the sources were agency websites: 18 different domains, and only one of them, an agency site on the guarantee prompt, appeared in all three repeats. On prompts about doing the work in-house versus hiring (12 answers), agencies vanished: developers.google.com was cited in 12 of 12 answers, alongside OpenAI's help pages and arXiv, and agency sites in none. On the fintech buyer prompts described above, providers' own pages were cited in nine of nine answers. So in this run the source followed the question, not a commercial-versus-informational split: questions about what vendors offer were answered from the vendors' own sites, how-to questions from platform documentation. In our wider count of the runs of 14–15 September 2026 (52 ChatGPT answers, signed out), ChatGPT cited a page from a vendor's own site in 20 of its 30 answers that listed sources; the breakdown is in our study of which pages AI assistants cite. Two days of our own prompts are a first look, not a rule. For a how-to prompt, write a page that reads like documentation: sourced, dated, specific and free of sales copy.
Local services add a layer of their own. When a prompt names a service and a city, ChatGPT can put a map of local businesses above the answer and name the firms on its cards. In our study of 72 answers to lawyer searches, run on 22 September 2026, it did so in 11 of 12 first-run answers, while the pages it cited were mostly directories and bar associations: firm websites made up 8 of its 71 links. For a local business, the listing belongs to the third-party layer, and it is the first thing a client sees.
Plan for this as the slowest layer and the one you control least. Getting into the directories and comparison pages that are already being cited for your category, keeping review profiles current, and being described with the same name, the same one-line positioning and the same facts everywhere you appear all take quarters, not weeks. Paying for placements in "best of" lists is a separate decision with its own trade-offs, which we work through in our guide on whether to buy listicle placements. If a competitor is already named for your prompts and you are not, the diagnostic order is in our guide on why ChatGPT doesn't recommend your company.
How does ChatGPT work out who you are, and what can you do this week?
If your goal is to get your business recommended on ChatGPT, distinguish that from getting an article cited. A citation can support an explanation without endorsing the company that published it. Test questions describing a customer need without naming your brand, then record recommendations and source links separately. For a business that is still establishing its offer, our guide to GEO for new businesses explains where to start.
Before working on recommendations, make the company unambiguous as an entity: one organization with one name, one description and a set of independent places where it also appears. We advise Organization markup in JSON-LD (a block of structured data on the page that tells machines who owns the site) on the canonical host, an identical name and one-line description on every profile, and a sameAs list that points at real external profiles (a company registry, LinkedIn, Crunchbase, Clutch, a Wikidata item if you qualify). A sameAs list that names only your own homepage misses the point, which is to link the entity to places you do not control.
What this looks like from the retrieval side: on September 15, 2026, we re-ran the four-question set from our Hindi GEO agency case, signed out, two repeats each. GET-GEO.AI was named in eight of eight answers, and on the one prompt we had lost on September 3 it came back at #4 and #5, with the case page itself cited as the source. Across the eight answers the model drew on the homepage, the case page and two different guides: it was drawing on the organization, not on one lucky URL. One assistant, one day, our own prompts: these are the first reproducible citations, not a stable position, and the next re-measurement is scheduled rather than assumed.
Plan these five types of work on different horizons: about a day for access changes, weeks to a quarter for consistent brand information once profiles exist, and several quarters for independent coverage. Our timeline guide explains the estimates. The table below sets out the actions and checks.
If you want the first layer checked and the first ten prompts run for you, the free AI visibility audit does exactly that: ten prompts in one language on which you are invisible today, the current answers, a crawler-access check, and the specific edits that would change the picture.
| Layer | What to do | How to check | Horizon | Who does it |
|---|---|---|---|---|
| Crawler access | Allow OAI-SearchBot in robots.txt; allow OpenAI's published IP ranges at the CDN or firewall | Fetch robots.txt and a key page with the OAI-SearchBot user agent; look for the bot in server logs after ~24 h | Days | You, or whoever runs the site |
| Rendering and host | Server-render key pages; one canonical host; 200 without redirect chains; no noindex where you want citations | Load the page with JavaScript off and find the sentence you want quoted in the HTML | About a week | You, or your developer |
| Pages with a checkable fact | One lead question per page: question as heading, 40–60-word answer with a number, date or named source, then detail | Run the prompt signed out, three repeats; log whether the page is cited and which passage | Weeks to a quarter | You, or an agency; depends on how many prompts |
| Corroboration | Get into directories and comparison pages already cited for your category; keep review profiles current; same facts everywhere | Ask ChatGPT your category's "best X" prompts; count which third-party pages appear and whether you are on them | Quarters | Partly other people; you can only apply and be consistent |
| Entity | Organization JSON-LD on the canonical host; identical name and description on every profile; sameAs to real external profiles | Ask ChatGPT who you are and what you do, signed out; compare the answer with your one-line positioning | Weeks to a quarter | You; the profiles must already exist |
Related questions
Is there a form to submit my website to ChatGPT?
No. OpenAI's search help page lists two eligibility conditions and offers no form: OAI-SearchBot may crawl the site, and your host or CDN accepts traffic from OpenAI's published searchbot IP addresses. Anyone selling a "submission to ChatGPT" is selling a robots.txt edit. Discovery happens through crawling and partner search providers.
How do I see ChatGPT referral traffic in my analytics?
ChatGPT has added utm_source=chatgpt.com to links in its answers; OpenAI documented this on a publisher help page withdrawn in October 2026. On our site it was the most frequent tagged source from 10 August to 10 November 2026, about one visitor in ten (Vercel Web Analytics). Filter by that source and, where the tag is missing, by the chatgpt.com referrer.
Does an llms.txt file make ChatGPT cite me?
There is no public confirmation that OpenAI reads llms.txt, and OpenAI's publisher documentation does not mention it. Publishing one is cheap and helps any tool that does consume it, so treat it as a small bet rather than a mechanism. Crawler access and page structure are what OpenAI documents. Our llms.txt guide covers what is and is not known.
Should I block GPTBot to keep my content out of training?
You can, and it does not cost you citations. OpenAI states that the OAI-SearchBot and GPTBot settings are independent: a site can allow OAI-SearchBot to appear in search while disallowing GPTBot to opt out of training. One caveat from OpenAI's crawler documentation: if both are allowed, a single crawl may serve both purposes, so make the opt-out explicit.
Does the same approach work for Perplexity, Claude and Gemini?
The conditions carry over; the mechanics differ. Perplexity retrieves live with its own crawler. Claude uses three Anthropic crawlers; in Otterly's June 2026 count of 379,321 citations on SaaS and tech queries, 64% went to company sites. The Gemini app answers from Google Search results, so Google's index and the Google-Extended token apply. Each has a guide: Perplexity, Claude, Gemini.
Related guides
Sources
- 01GET-GEO.AI research: which lawyers ChatGPT, Perplexity and Google AI Mode recommend, 72 answers to 12 lawyer-search prompts, 22 September 2026
- 02GET-GEO.AI research: which pages AI assistants cite, 569 source links from 120 answers, 14–15 September 2026
- 03GET-GEO.AI — AI search statistics for 2026, with our site's tagged traffic sources, 10 August–10 November 2026
- 04GET-GEO.AI — can ChatGPT recommend a new business?
- 05OpenAI — Overview of OpenAI crawlers (OAI-SearchBot, GPTBot, ChatGPT-User)
- 06OpenAI Help Center — Publishers and Developers FAQ (withdrawn in October 2026; archived copy of 4 October 2026)
- 07OpenAI Help Center — ChatGPT Atlas retirement
- 08OpenAI Help Center — Searching the web with ChatGPT
- 09Ahrefs — ChatGPT's most cited pages, 1,000 URLs (October 2025)
- 10Ahrefs — Best lists research, 750 prompts (December 2025)
- 11Otterly — Claude AI citation study, 379,321 citations (June 2026)
- 12Schema.org — Organization type reference
- 13GET-GEO.AI — Robots.txt for AI crawlers: the full list and ready-made configs
- 14GET-GEO.AI — How do you check what AI assistants say about your company, for free?
- 15GET-GEO.AI — Should you buy listicle placements?
- 16GET-GEO.AI — Why doesn't ChatGPT recommend your company?
- 17GET-GEO.AI — How long does GEO take?
- 18GET-GEO.AI — llms.txt guide
- 19GET-GEO.AI — Case: we asked ChatGPT for the best Hindi GEO agency (baseline Sept 3, re-measured Sept 15, 2026)
- 20GET-GEO.AI — How do you get cited by Perplexity?
- 21GET-GEO.AI — How do you get cited by Claude?
- 22GET-GEO.AI — How do you get cited by Gemini and Google AI Mode?
// share
How to cite this page
You may quote and reuse this content with attribution and a link to this page.
“How do you get cited by ChatGPT?” — GET-GEO.AI, 2026-11-18. https://get-geo.ai/en/guides/get-cited-by-chatgpt