How to Get Your Brand Cited by ChatGPT

There is no guaranteed way to get ChatGPT to cite your brand in an answer. But the conditions are known, and most of them can be checked in a working day.

Based on OpenAI's own documentation, two academic studies and our scan of the robots.txt files of 175 popular Turkish websites, four conditions stand out:

1. Access: OpenAI's search crawler, OAI-SearchBot, has to be able to reach your site.
2. Index: Your site has to be in the search provider ChatGPT uses.
3. Third-party mentions: Your name has to appear outside your own site.
4. Answerability: Your pages need text that answers the question directly.

This post separates each one into what OpenAI says and what it doesn't, then covers how to measure the result.

Where does ChatGPT find the sources for an answer?

As OpenAI describes it, ChatGPT search draws on two inputs. Announcing search in October 2024, OpenAI wrote that it "leverages third-party search providers, as well as content provided directly by our partners" (OpenAI, Introducing ChatGPT search). It does not say which search provider.

The crawler that decides whether your site can be shown is OAI-SearchBot. According to OpenAI's crawler documentation, it is used to surface websites in ChatGPT's search features. Sites that block it are not shown in ChatGPT search answers, though they can still appear as navigational links (OpenAI, Overview of OpenAI Crawlers).

The closest data on the provider question comes from an academic study. Zhang and colleagues compared six AI search engines with Google and Bing across 55,936 queries. ChatGPT's source selection diverged significantly only from Google's, which the authors read as "likely reflecting its reliance on Bing" (Zhang et al., arXiv 2512.09483).

That is an inference, not proof. The practical conclusion is still clear: a site that wants to appear in ChatGPT has no reason to neglect Bing.

Condition one: OAI-SearchBot must be able to reach your site

Access is set in robots.txt, and OpenAI keeps three crawlers apart:

• OAI-SearchBot surfaces your site in ChatGPT search. If you want to appear, don't block it.
• GPTBot collects content to train OpenAI's models. Blocking it signals that your content should not be used for training; it does not affect search.
• ChatGPT-User acts when a user asks ChatGPT to open a page; it does not crawl automatically. OpenAI notes that because these actions are user-initiated, robots.txt rules may not apply.

OpenAI also says it can take about 24 hours for a robots.txt change to reach its search systems (OpenAI, Overview of OpenAI Crawlers). A site that wants to opt out of training but stay in search would use:

User-agent: OAI-SearchBot
Allow: /

User-agent: GPTBot
Disallow: /

This is an option, not a recommendation. Whether to allow training crawlers is a business decision; our post on AI crawlers covers what each one does.

The Turkish picture: robots.txt is rarely the problem

How often do Turkish sites block ChatGPT? To find out, on 13 October 2026 we scanned the robots.txt files of the 200 most popular .tr domains in the Tranco top 100,000. Tranco is a ranking widely used in academic research that combines five separate popularity lists over a 30-day window (Tranco, list 64VKX). Each file was evaluated against the RFC 9309 rules for the home page.

175 sites could be analysed. Seven block OAI-SearchBot from the home page, or 4.0 percent. Five of those seven are major news sites.

The real split in the chart is between training and search: training crawlers are blocked more often than search crawlers. GPTBot is blocked on 6.3 percent of sites, Anthropic's training crawler ClaudeBot on 8.0 percent. Four sites block GPTBot while allowing OAI-SearchBot, a deliberate choice to opt out of training but stay in search. Yet only 27 of the 142 robots.txt files name any AI crawler at all. Most sites have made no decision; the crawlers are allowed by the general rule.

So in Turkey, robots.txt is usually not why a site is missing from ChatGPT. There is a second gate after it.

The invisible block: firewalls

robots.txt is a permission list, not a lock. A CDN or firewall can still refuse a crawler that robots.txt allows. In our scan, 25 sites could not be reached; 11 returned 403 to our research crawler, and 8 of those did the same to an ordinary browser user agent. Whether those rules also apply to OAI-SearchBot is something we cannot measure from outside.

That is why OpenAI's recommendation has two parts: allow OAI-SearchBot in robots.txt, and allow requests from OpenAI's published IP ranges (OpenAI, Overview of OpenAI Crawlers). The most reliable way to know whether your site is actually open to ChatGPT is your server log. Look for OAI-SearchBot requests and the status codes they received.

Condition two: be in the search provider's index

Even without OpenAI naming its provider, the work here is cheap and carries no risk. Add the site to Bing Webmaster Tools, submit your sitemap and notify Bing of new pages with IndexNow. These steps help Bing crawl your site faster and more completely. Even if Bing is not ChatGPT's provider, being in its index is what it takes to appear in Bing's own search and in Copilot.

The AI Performance report that Bing Webmaster Tools opened in preview in February 2026 is the measurable side of this. It shows how often your site is cited in Microsoft Copilot, in Bing's AI-generated summaries and in "select partner integrations", which pages are cited and which grounding phrases they matched (Bing Webmaster Blog, 10 February 2026).

Two limits matter. The report does not name ChatGPT, and it does not say which partner integrations it includes. Bing's help page also states that the report does not measure rankings, authority or importance; it only shows which content was cited (Bing Webmaster Tools, AI Performance help).

Condition three: be mentioned outside your own site

AI search trusts what others write about a brand more than the brand's own site. Chen and colleagues compared ChatGPT, Perplexity and Gemini with Google across verticals, languages and query phrasings. They found a "systematic and overwhelming bias towards Earned media", meaning third-party, independent sources, in AI search; Google's source mix was more balanced (Chen et al., arXiv 2509.08919).

There is a numerical side too: an AI answer has fewer places to fill.

In Zhang and colleagues' measurement, AI search engines showed 4.3 URLs per query on average, against 10.3 for traditional search engines. Only 38 percent of the domains shown appeared on both sides, and 37 percent appeared only in AI search (Zhang et al., arXiv 2512.09483). Ranking on Google's first page does not guarantee a ChatGPT citation, and a source that never appears on Google can still be cited in ChatGPT.

In practice, this turns into the following work:

• Trade publications: Interviews, expert commentary or data shared with publications that cover your field.
• Comparison and directory sites: A consistent profile on the lists your customers check when they look for a provider.
• A consistent identity: The same company name and description everywhere, linked from the Organization schema on your site. Our post on entity SEO covers this.
• Data worth citing: An original measurement others can quote. The robots.txt scan in this post is that kind of data.

None of this is a "ChatGPT placement" you can buy. An independent publication mentions you when you have done something worth mentioning.

Condition four: pages that can be quoted as an answer

ChatGPT lifts a sentence out of a page and quotes it out of context. A quotable page answers the question in its first paragraph, backs its claims with sources and renders its text on the server. A page whose content only appears after JavaScript runs in the browser may not be readable by every crawler.

Our post on writing content AI can quote covers how. One misconception is worth correcting here: OpenAI's crawler documentation talks about robots.txt and IP ranges for access; it does not list llms.txt, sitemaps or structured data as conditions. Our post on llms.txt explains why it is not a priority for most sites.

How do you measure visibility in ChatGPT?

The honest answer is that you can't, directly. At the time of writing, we could not find an OpenAI report that, like Google's Search Console, tells you how often your site is shown in ChatGPT. There are four indirect measures:

1. Referral traffic: Sessions from chatgpt.com in GA4. Our post on measuring AI search traffic in GA4 covers the setup.
2. Server logs: Which pages OAI-SearchBot visits and how often. No visits means no chance of appearing.
3. Bing AI Performance: Citations on the Copilot and Bing side. Not a direct ChatGPT measure, but an indicator that likely draws on the same index.
4. A fixed question set: Ask ChatGPT the 10–20 questions your customers would ask, every month, and record which sources it shows. Answers vary even for the same question, so look at the trend over months, not at a single run.

Together, these four give a reasonable answer to "are we showing up?". None of them does on its own.

What is being sold without evidence

ChatGPT optimisation is a new market, and offers that guarantee results already exist. OpenAI does not publish how sources are selected or any ranking criteria. A promise of "number one in ChatGPT" is a promise about something the seller cannot control.

Similarly, it is a common mistake to think that blocking GPTBot removes you from ChatGPT search. According to OpenAI's documentation, training and search are handled by separate crawlers; turning off training does not turn off search visibility. Our post on GEO vs SEO covers the same split on the Google side.

Frequently asked questions

What do I need to do to appear in ChatGPT?

There are four conditions. OpenAI's search crawler, OAI-SearchBot, must be able to reach your site; OpenAI says sites that block it are not shown in ChatGPT search answers. Your site must be in the index of the search provider ChatGPT uses, which in practice means Bing Webmaster Tools and a current sitemap. Your brand has to be mentioned outside your own site, in trade publications and independent sources. And your pages need server-rendered text that answers the question directly. None of these guarantees a citation; they remove what stands in the way of one.

If I block GPTBot, will I disappear from ChatGPT search?

Not according to OpenAI's own documentation. GPTBot collects content for training OpenAI's models; the crawler that surfaces sites in ChatGPT search is OAI-SearchBot. They are managed separately in robots.txt, so a site can block GPTBot and allow OAI-SearchBot, opting out of training while staying eligible for ChatGPT search.

Which search engine does ChatGPT use?

OpenAI does not name it. Its October 2024 announcement says ChatGPT search draws on third-party search providers and on content provided directly by its partners. An academic study of 55,936 queries found that ChatGPT's source selection diverged significantly only from Google's, which the authors attribute to ChatGPT's likely reliance on Bing. That is the researchers' interpretation, not a statement from OpenAI.

How can I measure how often ChatGPT cites my site?

There is no direct way; at the time of writing we could not find an OpenAI publisher report that tells you how often your site appears in ChatGPT. There are three indirect signals: referral traffic from chatgpt.com in GA4, OAI-SearchBot visits in your server logs, and asking ChatGPT the same set of questions every month and recording the sources. Bing Webmaster Tools' AI Performance report also shows citations in Microsoft Copilot and Bing's AI summaries; it does not say it covers ChatGPT.

Is there a service that guarantees a ChatGPT citation?

No. OpenAI does not publish how sources are selected or any ranking criteria, and the same question can be answered with different sources at different times. An offer that guarantees placement is selling something it cannot control. What can be done is meeting the access, index, third-party and content conditions and measuring the result regularly.

In short

Four things stand between your brand and a ChatGPT citation: the crawler can't reach your site, the site is not in the search provider's index, your name is not mentioned elsewhere, and your pages don't answer the question. The first two can be checked in a working day. The last two are months of content and visibility work.

Among popular Turkish sites, robots.txt is rarely the problem: only 4 percent block OAI-SearchBot. The real difference lies in whether others mention you and whether your pages are quotable.

Our free SEO checker reads your robots.txt separately for OAI-SearchBot, GPTBot and the other AI crawlers and shows which ones are blocked, in a few seconds. If nothing is blocked and you still don't appear, the work is on the content and visibility side; our GEO service page explains how that works and what it covers. If you'd like us to look at your site with you, get in touch.