How to Get Cited by ChatGPT Search

To get cited, or mentioned, by ChatGPT, your pages first have to be reachable by OAI-SearchBot, the crawler OpenAI uses for ChatGPT search. Then they need to answer specific questions clearly enough to be chosen as a source. OpenAI says placement is never guaranteed, so this guide covers what you can control: access, indexing, page content and measurement. You'll need access to your robots.txt, your CDN or firewall settings and your analytics. This guide is part of our AI search optimization guide.

How ChatGPT search finds pages

ChatGPT decides whether to search the web based on the question, or the user can turn search on. When it searches, OpenAI says it often rewrites the question into one or more targeted queries. It may send those to third-party search providers, review the results, then run more specific follow-up searches.

Separately, OpenAI runs its own search crawler, OAI-SearchBot. OpenAI says sites that opt out of OAI-SearchBot will not be shown in ChatGPT search answers, though they can still appear as plain navigational links.

The answer then includes citations. Users can open a Sources panel to see cited pages and other relevant links.

Two things follow. Your pages need to be reachable by OpenAI's crawler and findable through search providers. And because ChatGPT searches narrow, rewritten queries, each page should answer a specific question well.

Steps 1-3: Let the OpenAI search crawler in

1. 01
   Check robots.txt for OAI-SearchBot.

   Open your robots.txt and look for any group that names OAI-SearchBot, or a catch-all `User-agent: *` group that disallows paths you want cited. If there's no rule naming the bot, the catch-all group applies. Fetch the live file to see what crawlers actually get:

   bash

   curl -s https://example.com/robots.txt

   If you want to be explicit, add a group like this:

   text

   User-agent: OAI-SearchBot
   Allow: /

2. 02
   Allow OpenAI's published IP ranges at your CDN or firewall.

   robots.txt is only half the job. Many CDNs and firewalls block unfamiliar bots by default. OpenAI asks site owners to allow requests from its published search crawler IP ranges, listed at openai.com/searchbot.json. In your CDN or firewall dashboard, find the bot or WAF rules and confirm OAI-SearchBot is not challenged or blocked.

3. 03
   Decide on GPTBot separately.

   GPTBot is a different crawler. OpenAI uses it to collect content that may be used to train its models. Its setting is independent of OAI-SearchBot. You can block training and still allow search:

   text

   User-agent: GPTBot
   Disallow: /
   
   User-agent: OAI-SearchBot
   Allow: /

   Make this a deliberate business choice, not an accident of a default rule.

Steps 4-6: Make your pages worth citing

4. 04
   Get indexed by the search providers ChatGPT uses.

   OpenAI doesn't publish a full list of its search partners. Its help page does point to Microsoft's privacy statement for search providers, so being indexed in Bing is a sensible baseline. Verify your site in Bing Webmaster Tools and submit your XML sitemap. If your platform supports IndexNow, turn it on so new and updated URLs get reported as they change. Neither step guarantees indexing.

5. 05
   Answer one buyer question per page, answer first.

   Because ChatGPT rewrites questions into narrower searches, a page that clearly answers one specific question has a better shot than a page that covers everything vaguely. Pick the questions buyers actually ask. Put the direct answer in the first two or three sentences. Add the detail, examples and specifics below it. Use the question, or close to it, as a heading.

6. 06
   Keep key facts in the HTML.

   OpenAI doesn't document whether its crawler runs JavaScript. Make sure your main answer, prices if you publish them, service details and contact facts are in the HTML the server sends. Check by fetching the page and searching for a sentence you expect:

   bash

   curl -s https://example.com/services/ | grep -c "a sentence from the page"

   A result of 0 means that text isn't in the initial HTML. The same habit serves accessibility: text in the HTML, alt text on informative images and transcripts for video all help people using assistive technology, in line with WCAG's text-alternative criterion.

Steps 7-8: Check and measure

7. 07
   Ask ChatGPT your buyers' questions and log the results.

   Write down 10 to 20 questions buyers ask before they hire you. Ask each one in ChatGPT with search on. For each answer, record the date, whether your site is cited, which page, and which other sources appear. Use a fresh chat each time so earlier messages don't steer the result. Repeat monthly. Answers vary between runs, so look for patterns over several checks, not single results.

8. 08
   Track ChatGPT referrals in analytics.

   In Google Analytics or a similar tool, filter sessions where the source or referrer is chatgpt.com. Treat the count as a floor: some visits arrive without a referrer and show up as Direct. Compare landing pages against your question log from step 7.

If you'd rather have a team run these checks, our in-house SEO team does this work end to end, including reporting on where clients are mentioned in Google AI Overviews and ChatGPT.

Troubleshooting

Your site never appears, even for questions about your brand

Cause: usually access. Either robots.txt blocks OAI-SearchBot, a catch-all rule blocks it by accident, or a CDN or firewall rejects its requests before robots.txt is even read.

Fix: fetch your live robots.txt and read every group that could apply. Then check your firewall or bot-management logs for requests from OAI-SearchBot and note the response codes. If you see 403 or challenge responses, add an allow rule for OpenAI's published IP ranges. Wait about a day after any change, then rerun your test questions. If access is fine, check that the page is indexed in Bing and that your brand name and business details appear in plain text on the page, not only in a logo or image.

ChatGPT shows your link but no summary of your page

Cause: OpenAI says sites opted out of OAI-SearchBot will not be shown in ChatGPT search answers, though they can still appear as navigational links. A bare link with no summary often means the crawler is blocked for that page, by robots.txt or by a firewall rule.

Fix: if you want summaries and citations, allow OAI-SearchBot for those pages in robots.txt, and check that your CDN or firewall allows OpenAI's published IP ranges. Wait about a day, then ask the same question again. If you want a page kept out of ChatGPT entirely, check OpenAI's current crawler documentation before you rely on any one method, because a robots.txt block alone can still leave a navigational link.

robots.txt allows the bot, but your logs show blocked requests

Cause: a security layer sits in front of your site. Bot-fight modes, rate limits and geo rules often block crawlers that robots.txt allows. Some rules also block based on the user agent alone.

Fix: in your CDN or firewall dashboard, create an allow rule that matches the OAI-SearchBot user agent and OpenAI's published IP ranges together. Matching both is safer than user agent alone, because user agents are easy to fake. Then check the logs again after a day. For the same checks on another engine, see how to get cited by Perplexity. For Google's side, see how Google AI Overviews choose sources.