ChatGPT does not use a single public ranking in which websites hold permanent positions. When a question requires current or additional information from the internet, ChatGPT Search may reformulate the question into one or more more specific queries, retrieve results from external search providers and partner sources, and then compose an answer with links to selected pages.
OpenAI does not publish a complete list of signals, their exact weighting, or a formula that determines which URL will be cited. For that reason, no one can reliably promise a specific citation position. It is possible, however, to check whether a website is technically eligible, whether the correct page owns the topic, and whether the content provides a clear, current, and verifiable answer.
The most important point is to distinguish between four different processes:
These processes are not the same. A page may be accessible to a crawler but not selected for a particular query. A brand may be mentioned without a link. A link in the Sources panel also does not always mean that the exact page supports every sentence in the answer.
According to OpenAI’s official information, the process can be described as follows:
This means that ChatGPT does not necessarily search for the exact wording entered by the user. A page optimized only for one exact phrase may not match the more specific query the system uses behind the scenes.
No. Citations are associated with answers that use Search, but not every conversation triggers an internet search.
ChatGPT may automatically search for information when a question would benefit from current web results. The user may also activate Search manually. When search has not been used, the absence of a citation does not prove that the website has a technical problem.
When Search is used, the answer may include:
OpenAI notes that the Sources panel may contain both cited sources and other relevant links. For this reason, an analysis should not check only whether a domain appears in the panel, but whether the specific URL is connected to a claim in the answer.
Discovery means that the page can enter the search process. OpenAI uses OAI-SearchBot to crawl websites automatically for ChatGPT Search.
If a website blocks this crawler, OpenAI states that its pages will not be displayed in ChatGPT Search answers, although the domain may still appear as a navigational link in certain situations.
After Search is triggered, the system works with search queries and retrieved results. The official documentation confirms the use of external search providers and partner content, but it does not publish an exact formula for ranking all potential sources.
It has therefore not been established that one universal factor, such as content length, Schema.org markup, the number of FAQ questions, or a position in Google, independently determines source selection.
Citation is the final visual connection between a source and the generated answer. A URL may be considered during the search process without necessarily being displayed as a citation. Conversely, the Sources panel may show a relevant link that is not attached to a specific sentence.
This distinction is important when measuring visibility. The absence of a citation should not automatically be interpreted as proof that the system failed to discover the page.
| User agent | Primary role | Does it determine participation in ChatGPT Search? |
|---|---|---|
OAI-SearchBot | Automatic crawling for ChatGPT search features | Yes. This is the crawler used to manage opt-out from Search and automatic crawling. |
GPTBot | Crawling content that may be used to train generative AI foundation models | No. Allowing it is not a requirement for participation in ChatGPT Search. |
ChatGPT-User | Visiting a page following a specific action or request by a user | No. OpenAI states that it is not used to determine whether content may appear in Search. |
The settings for OAI-SearchBot and GPTBot are independent. A website may allow the Search crawler while blocking the use of its content for training through GPTBot.
An example robots.txt configuration for such a policy is:
User-agent: OAI-SearchBot
Allow: /
User-agent: GPTBot
Disallow: /
This is only an example. The actual robots.txt file must be checked against the website’s other rules, firewall, CDN configuration, and the IP ranges published by OpenAI.
OpenAI also states that after a robots.txt change, it may take approximately 24 hours for the Search systems to reflect the new settings.
For content to appear in ChatGPT Search answers through automatic crawling, the website must allow OAI-SearchBot and permit requests from the published IP ranges.
The check does not end with robots.txt. The crawler may be permitted there but blocked by:
ChatGPT Search uses information from the internet and links to relevant web sources. If the important answer is available only after login, inside a closed interface, or in an element that cannot be extracted reliably, the page has less practical value as a source.
The official documentation shows that ChatGPT may transform the user’s question into one or more targeted queries. The page must therefore solve the real problem rather than merely repeat one keyword phrase.
When the user asks, “Which SEO agency offers ChatGPT optimization in Bulgaria?”, the system may search for a query related to providers, location, and service. When the user asks, “How do I allow the ChatGPT crawler in robots.txt?”, the search query and the appropriate owner URL will be different.
The following points have not been published by OpenAI as an official list of ranking factors. They are a practical interpretation of how search, retrieval, and the citation of specific claims work.
When several pages on the same website provide similar but different answers, it is unclear which one should be used as the primary source.
A useful model is:
The page should quickly show:
General statements such as “high-quality content is important” do not provide a sufficiently specific fact that can be used and cited.
When official documentation does not disclose a mechanism, this should be stated clearly.
Suitable wording includes:
This allows the reader to distinguish a verifiable claim from professional judgment.
For queries about services, companies, specialists, or products, the main information should not conflict across different pages:
Consistency does not guarantee citation, but it reduces the risk of the system encountering conflicting claims.
When a claim exists only on a company’s own website, it is a self-published claim. Editorial mentions, official registers, professional profiles, primary data, and independent sources may help verify it.
This does not mean mass-creating profiles or purchasing random links. The external source should genuinely confirm the information and provide value to the user.
GPTBot is associated with the potential use of content for training, not eligibility for ChatGPT Search. OAI-SearchBot is used for Search.
In its current crawler documentation, OpenAI describes OAI-SearchBot, GPTBot, ChatGPT-User, robots.txt, and the published IP ranges. It does not state that a website must have an llms.txt file to participate in ChatGPT Search.
Such a file cannot replace accessible content, the correct owner URL, and an allowed Search crawler.
OpenAI has not published an official rule stating that a particular schema type guarantees a citation. Structured data should describe the real visible content, but it is not a pass for automatic inclusion.
ChatGPT may reformulate the user’s question into different and more specific search queries. Ranking for one monitored phrase therefore does not prove that the same URL will be selected for every conversational variation.
Artificial length does not solve the absence of an accurate answer. An FAQ section is useful only when the questions are genuine, are not covered more effectively in the main content, and do not duplicate other owner pages.
A single appearance shows that the page was used at a particular moment and in a particular context. It does not prove stable visibility across all related queries.
Before analyzing the website, check whether the answer contains inline citations or a Sources button. If ChatGPT did not use web search, the absence of your domain is not a technical diagnosis.
Determine:
OAI-SearchBot;Disallow rule;For every test query, determine which page should be cited.
If you cannot identify one URL unambiguously, the problem is probably architectural rather than limited to ChatGPT. Do not create a new article for every wording variation. First expand or clarify the existing owner page.
The reader should understand the main point without assembling information from five different pages.
Check whether the URL contains:
For every important claim, ask:
A page that mixes facts and assumptions is harder to verify and less useful to the user, regardless of whether it is cited.
The owner URL should receive links from relevant parent and sibling pages. The anchor text should describe the topic clearly, rather than several different URLs receiving competing anchors for the same problem.
An informational article, service page, product category, and company profile solve different tasks.
For a query such as “How does OAI-SearchBot work?”, the appropriate source is likely to be a technical guide. For a query such as “agency for ChatGPT optimization”, the system may look for a commercial page or a provider-selection page. One URL should not attempt to own both intents completely.
There is no official OpenAI report showing every question for which a domain was considered as a potential source. Practical monitoring must therefore be structured.
Create a limited set of queries based on real business tasks:
For every test, record:
| Field | What to Record |
|---|---|
| Prompt | The exact wording without later editing |
| Date and context | Date, country, language, and whether the conversation contains previous context |
| Search | Whether web search was triggered |
| Cited domain | Which website was shown |
| Cited URL | The exact page, not only the domain |
| Supported claim | Which part of the answer is connected to the citation |
| Accuracy | Whether the source genuinely supports the statement |
| Repeatability | Whether the result appears in subsequent tests |
Do not change your strategy after one test. Compare groups of closely related queries and look for recurring problems:
OAI-SearchBot is blocked. The cause may be robots.txt, a firewall, CDN, or anti-bot system.GEO and AEO are industry concepts used for visibility in generated and direct answers. Their definitions, similarities, and differences from SEO are discussed in the separate guide What Are GEO and AEO?.
This page has a narrower purpose: to explain ChatGPT Search, the crawlers, and visible citations specifically. It does not define the entire GEO/AEO discipline and does not present the service.
No. The official materials explain the triggering of Search, query reformulation, external search providers, partner content, crawlers, and how citations are displayed. They do not publish a complete ranking model, a list of weights, or a guaranteed URL-selection formula.
No. ChatGPT Search uses OAI-SearchBot. GPTBot is associated with crawling content that may be used to train models. The settings are independent.
OpenAI states that websites that deny access to OAI-SearchBot will not be displayed in ChatGPT Search answers, but they may still appear as navigational links. This should not be treated as a normal citation-visibility strategy.
Not necessarily. The Help Center states that the panel may contain cited sources and other relevant links. Check whether the link is attached to a specific claim.
OpenAI states that it may take approximately 24 hours for the systems to adapt to the change. After that, the server logs, CDN, and firewall rules should also be checked.
No. Technical eligibility, structure, the evidence base, and alignment with intent can be improved. The final selection of a source for a particular answer remains outside the control of the website or agency.
When it is unclear why your website is not appearing as a source, the problem may be technical, content-related, or architectural.
As part of AI search engine optimization, we review:
OAI-SearchBot and the applicable IP ranges;We do not promise guaranteed citations. The objective is to remove verifiable obstacles and build a website that provides clear, accessible, and reliable sources for real user questions.
Sources checked on July 13, 2026.
Image source: ChatGPT
