To get cited by ChatGPT, let OpenAI's search crawler reach your pages, put each answer in plain text on an indexable page, and earn mentions on the third-party lists the engines read. Nothing guarantees it: OpenAI says placement "is not guaranteed". In our own tests on 13 October 2026, ORRJO was named in 6 of 46 category prompt runs.
The same steps cover Perplexity, Microsoft Copilot and Google's AI Overviews and AI Mode, because each one works from a search index. This guide sets out what each engine's own documentation says it reads, the steps in order, what our own test runs showed, the claims that do not hold up, and a checklist.
How to read the tags on this page.
Reported The engine's operator or a published study says it. We have not tested it ourselves.
Verified We saw it in our own dated test runs on 13 October 2026, under the conditions stated below.
- Reported ChatGPT: "Any public website can appear in ChatGPT search." OpenAI says so in its publishers FAQ, which also says not to block its search crawler, OAI-SearchBot (OpenAI Help Center).
- Reported Google AI Overviews and AI Mode: "There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary." (Google Search Central).
- Reported The research: a 2026 review of 45 studies found that "already-retrieved content can causally alter its citation or use, but no reviewed technique shows a stable, longitudinal, cross-platform causal effect on organic discoverability or downstream behavior." (Martinez, arXiv).
- Verified Our tests: 6 of 46 category runs on ChatGPT, Gemini and Claude named ORRJO. Google AI Mode named us in 4 of 9 category prompts. Perplexity, logged out, named us in 0 of 10 runs.
Disclosure: ORRJO publishes this guide and sells an AI visibility audit and SEO and AEO work. Our own results below include the many runs where the engines did not name us. Every third-party figure links to the page it came from.
What does it mean to be cited by ChatGPT?
Being cited means ChatGPT links to your page as a source for its answer. Being mentioned means it names your company in the answer text. The two are separate, and you can have one without the other.
Reported BrightEdge analysed thousands of ChatGPT prompts and found that "ChatGPT mentions brands 3.2x more than it cites them", and that 44% of the prompts it studied produced no brand mention at all. So count both, and expect more mentions than links.
Verified We saw the gap the other way round too. On 13 October 2026, orrjo.com appeared in the sources ChatGPT or Claude retrieved on 5 of 30 category runs where sources were visible, and ORRJO was named as a recommendation on 3 of those 5. Being read is not the same as being recommended.
Where does each AI engine get its answers?
Each engine answers from a search index, either its own or a partner's, so getting your pages into that index comes first. This is what the operators themselves say:
| Engine | Where its answers come from Reported | What lets your pages in Reported |
|---|---|---|
| ChatGPT | Searches the web "when your question would benefit from current information". It "sometimes partners with other search providers", and when it does, it "typically rewrites your query into one or more targeted queries" sent to them. The help page links Microsoft's and Shopify's privacy statements for those providers (OpenAI). | Allow OAI-SearchBot, and let "OpenAI's published searchbot IP addresses" through your host or CDN (OpenAI). A robots.txt change can take "~24 hours" (OpenAI). |
| Google AI Overviews and AI Mode | Google's own index: it relies "on our core Search ranking systems to retrieve relevant, up-to-date web pages from our Search index" (Google). | To be shown as a supporting link, a page must be "indexed and eligible to be shown in Google Search with a snippet" (Google), and the site must be included in Search generative AI features in Search Console, which is the default (Google). |
| Perplexity | "search the internet in real-time" for each question, and "Each answer includes numbered citations" (Perplexity). | Allow PerplexityBot and its published IP ranges. It "is not used to crawl content for AI foundation models" (Perplexity). |
| Microsoft Copilot | "Powered by Bing's search index" (Microsoft Advertising). | Be indexed in Bing. Bing Webmaster Tools reports how often your pages are cited in Copilot and Bing's AI summaries (Bing). |
| Claude | Web search. Anthropic's crawler article does not name the index it searches (Anthropic). | Allow Claude-SearchBot and Claude-User. Blocking either "may reduce your site's visibility" (Anthropic). |
| Gemini app | Google Search grounding: Google-Extended governs whether your content is used to train Gemini models and "for grounding" in Gemini Apps (Google). | Do not block Google-Extended if you want Gemini Apps to ground answers on your pages. It does not affect Google Search (Google). |
What each engine did with our own pages on 13 October 2026 is in the results table further down.
Google adds a point worth knowing before you buy any measurement: "No third-party tool has access to our internal ranking or AI systems." (Google Search Central). Our own audit has no such access either: it asks the engines questions and counts the answers.
How do you get cited by ChatGPT, step by step?
Let the search crawlers in, get indexed, put your answers in plain text, publish first-hand facts, earn genuine mentions elsewhere, keep pages current, and measure. In order:
1. Let the AI search crawlers in
Reported Allow the crawlers that power search answers in your robots.txt: OAI-SearchBot for ChatGPT, PerplexityBot for Perplexity, and Claude-SearchBot and Claude-User for Claude. OpenAI is blunt: "Sites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers" (OpenAI). Then check your host, CDN or firewall is not blocking their published IP addresses.
Training crawlers are a separate choice. "Disallowing GPTBot indicates a site's content should not be used in training generative AI foundation models." (OpenAI), and Anthropic says the same of ClaudeBot. Our guide to which AI crawlers to allow in robots.txt covers each one. Our own robots.txt, checked on 14 October 2026, allows all of the crawlers named in this step, plus GPTBot, ClaudeBot and Google-Extended.
2. Get indexed by Google and Bing
Reported To be shown as a supporting link in AI Overviews or AI Mode, a page must be "indexed and eligible to be shown in Google Search with a snippet" (Google). Google also requires the site to be included in Search generative AI features in Search Console. That control reached all websites by 31 August 2026, it is "the default control for all properties", and if you exclude your site, "You won't receive any traffic or impressions from these features." (Google Search Console Help). Copilot runs on Bing's index, so verify your site in Bing Webmaster Tools as well.
Verified In our own tests, Google's AI Overviews went where the rankings were. On 13 October 2026 an AI Overview appeared on 27 of the 34 non-brand queries we tested. All five that used ORRJO sat on queries where ORRJO ranked between 2nd and 6th in the normal results, and where ORRJO was not on page 1, no AI Overview used it.
3. Put every answer in plain text on the page
Reported Microsoft warns that "AI systems may not render hidden content, so key details can be skipped", so keep answers out of tabs and collapsed accordions, and "For critical details, use HTML to ensure clarity and better parsing." rather than a PDF or an image (Microsoft Advertising). Google's advice is the same in fewer words: make sure "important content is available in textual form" (Google).
4. Answer the question first, under a heading that asks it
Reported Bing says "Clear headings, tables, and FAQ sections help surface key information and make content easier for AI systems to reference accurately." (Bing Webmaster Blog). Microsoft adds that with question and answer pairs, "Assistants can often lift these pairs word for word into AI-generated responses." (Microsoft Advertising).
The engines do not fully agree on how far to go. Microsoft describes Copilot breaking content into "smaller, structured pieces", while Google says "There's no requirement to break your content into tiny pieces for AI to better understand it." (Google). Write for the reader: a direct answer under each question heading, with the detail after it.
Reported You do not need special markup. Google: "Structured data isn't required for generative AI search, and there's no special schema.org markup you need to add." (Google). If you do use structured data, Google asks you to make sure it "matches the visible text on the page" (Google).
5. Publish first-hand facts, with their sources
Reported Google calls original content the biggest lever: "Creating content that people find unique, compelling, and useful will likely influence your website's presence in generative AI search in the long run more than any of the other suggestions in this guide." (Google).
Reported In the peer-reviewed GEO study (KDD 2024), the top-performing methods were adding citations from reliable sources, quotations and statistics to a page. They "achieved a relative improvement of 30-40% on the Position-Adjusted Word Count metric" (Aggarwal et al.). Read that with its limits. A 2026 review says the paper's gains are "valid within its experimental setting but conditional on a source already being present in a fixed context; they establish neither organic discoverability nor durable traffic effects." (Martinez). Facts and sources can change how a page that is already being read gets used. The review found no evidence that they get a page read in the first place.
6. Earn genuine mentions on the pages the engines already read
Reported Ahrefs studied 75,000 brands and found that for Google AI Overviews, "Web mentions (0.664) correlate much more strongly than backlinks (0.218)." It stresses that this is correlation, not proof of cause (Ahrefs).
Verified In the ChatGPT, Gemini and Claude runs where we could see sources on 13 October 2026, the engines retrieved 599 unique URLs from 362 domains. By type, 321 of those URLs were on vendor sites, 144 were "best X" lists and roundups, 50 were directory pages, 30 were review-platform pages, 19 were news and trade press, 17 were ORRJO pages, 13 were social posts and 5 were Reddit threads. None was a YouTube video. The counts lean towards ChatGPT, which shows its full source set.
Verified We then read the raw HTML of 29 third-party lists that Perplexity and Google cited most often that day, including some cited only in logged-in Perplexity runs or in Google's normal results. 3 of the 29 mentioned ORRJO.
Reported Earn those mentions honestly. Google says "seeking inauthentic 'mentions' across the web isn't as helpful as it might seem" (Google), and Wikipedia's rules say "Do not edit Wikipedia in your own interests" and "Paid editing must be disclosed." (Wikipedia).
7. Keep the facts current, on your site and off it
Reported SE Ranking's study of 129,000 domains found that "Content updated in the past three months averages 6 citations versus 3.6 for outdated pages." That is a correlation in ChatGPT citations, not a test (SE Ranking). Bing also points to IndexNow, which notifies participating search engines "whenever content is added, updated, or removed" (Bing).
Verified Old facts outlive the page that published them. On 13 October 2026 two directory profiles supplied Claude with headline figures we had already retired, Google's AI answers quoted some of our prices with dollar signs although the live pages showed sterling (our log names older Google snippets as the likely source), and ChatGPT quoted a lead generation price we had changed that day. Check what third-party profiles say about you, not only your own pages.
8. Measure mentions and citations, run by run
Reported Ask the questions your buyers ask, without your name in them, and count. Then use what the engines report: ChatGPT "automatically includes the UTM parameter utm_source=chatgpt.com in referral URLs" (OpenAI); Search Console's generative AI performance report shows impressions in AI Overviews and AI Mode (Google); and Bing's Citation Share is "the percentage of citations attributed to your site out of all citations shown across all sites for that same grounding query" (Bing).
Verified Run each question more than once. On ChatGPT, two runs of the same "best B2B lead generation agency UK" prompt shared only 3 of 6 vendors, and on another prompt one run named ORRJO and the next did not, with the same ORRJO page in both source sets. A single answer is a sample. Our AI share of voice definition and our comparison of AI visibility tools cover the measuring in more depth.
What did ORRJO's own tests show?
ORRJO was named in a minority of runs. On Google AI Mode every naming cited one of our own comparison pages or guides, and every AI Overview that used ORRJO came on a query where we already ranked on page 1. These are single-day readings from 13 October 2026, not a benchmark.
How the runs were made. Every chat engine was used from a logged-in account. ChatGPT ran on the Free tier in a Temporary chat with the Unpersonalized option and web search on. Gemini ran in Temporary chat on ORRJO's own Google Workspace account, on the Pro model for the first two prompts and on Flash after a usage limit. Claude ran in Incognito, which did not fully remove account context, so treat its ORRJO mentions with care. The browser's IP address was in Spain. Google's normal results and AI Mode both ran from a logged-in Google account on google.co.uk; personalised results were switched off for the normal results only, so AI Mode answers may be personalised. For Perplexity we count only the 10 logged-out runs; the later runs were logged in and personalised, so we left them out. Copilot could not be tested.
| Surface Verified | ORRJO named | What stood out |
|---|---|---|
| ChatGPT | 3 of 23 category runs | An ORRJO "best X" list put ORRJO into one answer. |
| Gemini | 2 of 16 | Showed sources on 1 of 17 runs. |
| Claude | 1 of 7 | Cited an ORRJO list as the source for other agencies. |
| Google AI Mode | 4 of 9 prompts, cited in 6 | Every naming cited an ORRJO "best X" page or guide. |
| Google AI Overviews | Used ORRJO on 5 of 34 non-brand queries | All five were queries where ORRJO ranked 2nd to 6th. |
| Perplexity, logged out | 0 of 10 | Retrieved an ORRJO page on both runs of one prompt. |
| Microsoft Copilot | Not tested | It required a Microsoft sign-in. |
On the head-term prompt "best B2B lead generation agency UK", none of the 4 chat runs across ChatGPT, Gemini and Claude named ORRJO.
Three lessons from those runs:
- Self-published lists get read, and so do everyone else's. Claude warned in one answer that "most 'top 10' lists are written by agencies that rank themselves". Our own list got ORRJO into one ChatGPT answer, while 3 of the 29 third-party lists we checked mentioned ORRJO.
- Retrieved is not recommended. An engine can read your page and still name someone else. Count names and citations separately.
- Brand answers draw on third-party records. When we asked ChatGPT about ORRJO by name, its sources included a review platform profile, LinkedIn and Companies House as well as orrjo.com. What those records say about you becomes part of the answer.
What does not work, or is not proven?
Special files, special schema and engineered mentions have no support in the engines' own documentation, and some widely repeated figures are misread. The common claims, checked against their sources:
| The claim | What the source says Reported |
|---|---|
| An llms.txt file gets you cited | SE Ranking: "Including LLMs.txt has shown negligible impact on ChatGPT citation likelihood." (SE Ranking). Google says Google Search ignores the file (Google). Our llms.txt guide covers what the file is for. |
| Buying ChatGPT ads gets you cited | OpenAI: "Answer independence: Ads do not influence the answers ChatGPT gives you." (OpenAI). Its help centre adds that "advertisers have no ability to shape, rank, or alter ChatGPT's responses." (OpenAI Help Center). |
| You need special schema for AI answers | "there's no special schema.org markup you need to add" (Google). |
| Google-Extended controls AI Overviews | "Google-Extended does not impact a site's inclusion in Google Search nor is it used as a ranking signal in Google Search." (Google). The Search Console generative AI control does that job (Google). |
| Wikipedia is cited in 47.9% of ChatGPT answers | In Profound's study, 47.9% is Wikipedia's share "Within ChatGPT's top 10 most-cited sources". Across all ChatGPT citations from August 2024 to June 2025, Wikipedia's share was 7.8% (Profound). |
| GEO rewrites add up to 40% visibility | The GEO paper reports up to 40% in its own benchmark (Aggarwal et al.). A 2026 review of 45 studies says those gains are "conditional on a source already being present in a fixed context" and that "citation-oriented rewrites can impair retrieval" (Martinez). |
| You cannot see AI answers in Search Console | Search Console's generative AI performance report covers AI Overviews and AI Mode. Google says it reached all websites as of 31 August 2026 (Google). |
| A separate page for every variant of a question | Doing that "primarily to manipulate rankings or generative AI responses in Google Search violates Google's scaled content abuse spam policy" (Google). |
Google's summary of the whole field: "From Google Search's perspective, optimizing for generative AI search is optimizing for the search experience, and thus still SEO." For how the three labels differ, see SEO vs AEO vs GEO and our definition of GEO.
Is your site ready to be cited? A checklist
It is ready when every line below is true. Work down the list once, then repeat the last two items every month:
- robots.txt allows OAI-SearchBot, PerplexityBot, Claude-SearchBot and Claude-User
- Your host, CDN and firewall let the engines' published IP addresses through
- Training crawlers (GPTBot, ClaudeBot, Google-Extended) are a decision you made on purpose
- Key pages are indexed in Google and eligible for a snippet
- Search Console's generative AI control is left on include
- The site is verified in Bing Webmaster Tools
- Every answer is in the HTML text, not in a tab, accordion, PDF or image
- Each page answers its question under a heading that asks it, with tables for prices and comparisons
- Any structured data matches the visible text word for word
- Your facts are first-hand, dated and linked to their sources
- Mentions on third-party lists and profiles are earned, accurate and current
- The same blind buyer questions run several times a month, with mentions and citations counted separately
If you want someone else to run the questions, our AI visibility audit does it blind: "We never put your name inside the questions." It costs £750 plus VAT, one off, and we do not guarantee that AI engines will mention you, or where you will rank. Ongoing work sits in our AEO and SEO and AEO service.
Questions buyers ask
How long does it take to get cited by ChatGPT?
Nobody can give you a reliable date. OpenAI says a robots.txt change can take about 24 hours to reach ChatGPT search, and that placement is not guaranteed. Beyond that, we found no engine documentation or study that sets a timeline, so we do not give one.
Does ChatGPT use Bing to find sources?
The OpenAI help pages we cite do not say Bing by name. They say ChatGPT search sometimes partners with other search providers, sends them rewritten queries, and link the Microsoft and Shopify privacy statements for those providers. Microsoft says its own Copilot is powered by Bing's search index.
Can you pay OpenAI to get cited in ChatGPT?
No, according to OpenAI. Its help centre answers the question directly: "No. Ads do not influence ChatGPT's answers. Ads run on separate systems from our chat model, and advertisers have no ability to shape, rank, or alter ChatGPT's responses." Ads are separate, labelled placements. For the answers themselves, OpenAI says placement is not guaranteed.
What is the difference between being mentioned and being cited?
A mention is your name in the answer text. A citation is a link to your page as a source. They are counted separately: BrightEdge found ChatGPT mentions brands 3.2 times more often than it cites them, and in our own 13 October 2026 tests ORRJO pages were retrieved as sources more often than ORRJO was named.
Should I block GPTBot if I want to be cited?
Blocking GPTBot is a training opt-out. OpenAI says disallowing GPTBot means content should not be used to train its foundation models, while OAI-SearchBot is the crawler that surfaces sites in ChatGPT search. Our own robots.txt allows both.
Does an llms.txt file help you get cited by ChatGPT?
Not on the evidence we found. SE Ranking's ChatGPT study found that including an llms.txt file has shown negligible impact on citation likelihood, and Google says Google Search ignores the file. Our llms.txt guide covers what the file is for.
How do you measure whether AI engines cite you?
Ask the same buyer questions without your name in them, run each one several times, and count mentions and citations separately. Then add the engine reports: utm_source=chatgpt.com referrals in your analytics, the generative AI performance report in Google Search Console, and AI Performance in Bing Webmaster Tools.
Sources
- OpenAI, Overview of OpenAI Crawlers. OAI-SearchBot, GPTBot, ChatGPT-User, OAI-AdsBot and the 24-hour robots.txt note, checked 13 October 2026.
- OpenAI Help Center, Searching the web with ChatGPT. Placement, eligibility, search providers. The live page blocks automated fetches, so we read the Internet Archive copy of 23 September 2026, checked 13 October 2026.
- OpenAI Help Center, Publishers and Developers FAQ, in the ChatGPT Atlas collection. Any public website can appear; utm_source=chatgpt.com. Read from the Internet Archive copy of 4 October 2026 for the same reason, checked 13 October 2026.
- OpenAI Help Center, Ads in ChatGPT. Whether ads influence answers. Read from the Internet Archive copy of 4 October 2026 for the same reason, checked 13 October 2026.
- OpenAI, Our approach to advertising and expanding access to ChatGPT, 16 January 2026. Read from the Internet Archive copy of 7 October 2026, checked 13 October 2026.
- Google Search Central, AI features and your website, last updated 10 December 2025, checked 13 October 2026.
- Google Search Central, Optimizing your website for generative AI features on Google Search, last updated 10 July 2026, checked 13 October 2026.
- Google Search Console Help, Search generative AI control. Rolled out to all websites as of 31 August 2026, checked 13 October 2026.
- Google Search Console Help, Generative AI performance report (Search), checked 13 October 2026.
- Google, List of Google's common crawlers. Google-Extended, checked 13 October 2026.
- Microsoft Advertising, Optimizing Your Content for Inclusion in AI Search Answers, Krishna Madhavan, 8 October 2025, checked 13 October 2026.
- Bing Webmaster Blog, Introducing AI Performance in Bing Webmaster Tools Public Preview, 10 February 2026, checked 13 October 2026.
- Bing Search Blog, New AI Visibility Insights in Bing Webmaster Tools: Intents, Topics, Citation Share, Compare, 16 June 2026. Citation Share, checked 13 October 2026.
- Perplexity, Perplexity Crawlers and How does Perplexity work?, checked 13 October 2026.
- Anthropic, Does Anthropic crawl data from the web, and how can site owners block the crawler?, updated 7 April 2026, checked 13 October 2026.
- Aggarwal, Murahari, Rajpurohit, Kalyan, Narasimhan and Deshpande, GEO: Generative Engine Optimization, arXiv, accepted to KDD 2024 (abstract and v3 full text), checked 13 October 2026.
- Olivier Martinez, Optimizing Visibility in Generative Engines: A Critical Survey of Generative Engine Optimization (2023-2026), arXiv preprint, submitted 15 July 2026 (abstract), checked 13 October 2026.
- Ahrefs, An Analysis of AI Overview Brand Visibility Factors (75K Brands Studied), Louise Linehan, published 26 May 2025, updated 27 April 2026. 75,000 brands, Spearman correlations, checked 13 October 2026.
- Profound, AI Platform Citation Patterns: How ChatGPT, Google AI Overviews, and Perplexity Source Information, updated August 2025. 680 million citations, August 2024 to June 2025, checked 13 October 2026.
- SE Ranking, How to Optimize for ChatGPT: Skip LLMs.txt, Earn Trust on Quora & Reddit, published 24 November 2025, updated 27 March 2026. 129,000 domains and 216,524 pages, checked 13 October 2026.
- BrightEdge, ChatGPT Brand Mentions vs. Citations: What Triggers Visibility, checked 13 October 2026.
- Wikipedia, Conflict of interest guideline, checked 13 October 2026.
- ORRJO, test logs for ChatGPT, Gemini, Claude and Copilot, and for Google, Google AI Mode and Perplexity, both run on 13 October 2026 with verbatim transcripts on file. Not published.