ChatGPT

ChatGPT SEO and Brand Monitoring

ChatGPT SEO comes down to two things you can check: what ChatGPT already says about you, and whether OpenAI's search crawler can reach your site. 99Visibility asks ChatGPT the prompts your buyers type, reads robots.txt and your firewall for each OpenAI crawler, and ranks what to fix.

Last reviewed:

36%of the coverage weight in the 99Visibility score, the most of any platform. The weights are our model of reach, not a measured market share.

How ChatGPT works for AI visibility

A prompt reaches a ChatGPT answer two ways: search over the index OAI-SearchBot feeds, with links, or training data
ChatGPT searches when the model decides it should; otherwise it answers from training data. Our audit asks it with search on. Source: OpenAI crawler documentation and Help Center; how we ask, from our audit configuration.

Two ways to answer

ChatGPT answers from what its model learned in training, or searches the web first; in the app, the model decides per question whether to search. With search off, nothing you publish now reaches the answer. OpenAI lists May 18, 2026 as the knowledge cutoff for gpt-6-luna, the model we ask.

How search finds pages

When ChatGPT searches, OpenAI says it "typically rewrites your query into one or more targeted queries" and sends them to the search providers it works with. Which provider answers which query is not documented. OAI-SearchBot is the crawler OpenAI runs to surface sites in ChatGPT search.

Two gates, not one

OpenAI's help center says that to be eligible you must "allow OAI-Searchbot to crawl the site and confirm that the website host or content delivery network allows traffic" from its published IP addresses. That is robots.txt and your firewall, and either one can shut you out.

How it cites

With search on, ChatGPT links the sources it used, and OpenAI's API documents that the pages it consulted usually outnumber the ones it cites. With search off there is nothing retrieved to cite, so answers seldom carry a link.

JavaScript

Vercel's crawler study found OpenAI's crawler fetched JavaScript files on 11.50% of requests and executed none of them. A price that appears only after scripts run was invisible to it. The data is a one-month snapshot from December 2024, from a company that sells server rendering.

The crawlers that reach ChatGPT

  • OAI-SearchBot (ChatGPT search)

    Sites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers, though can still appear as navigational links.

    Vendor-documented: developers.openai.com →
  • GPTBot (OpenAI model training)

    Your content is excluded from training OpenAI foundation models. This does not affect ChatGPT search — OpenAI states each setting is independent.

    Vendor-documented: developers.openai.com →
  • ChatGPT-User (ChatGPT on-demand browsing)

    ChatGPT cannot fetch your pages live when a user asks about them. Not used to decide search inclusion, and because a user triggers it, robots.txt rules may not apply.

    Vendor-documented: developers.openai.com →

What is not documented

  • How ChatGPT ranks the pages it finds. OpenAI's whole statement is that it ranks results "using multiple factors intended to help users find relevant, reliable information" and that "placement is not guaranteed."
  • How much Bing contributes. OpenAI says ChatGPT search "sometimes partners with other search providers", and nobody has measured how Bing index presence changes what ChatGPT cites.
  • Whether allowing GPTBot changes what a future model knows about you. Nobody has measured it.
  • How often ChatGPT searches. Published estimates run from about 18% (Profound) to 80.5% (cloro), because each measures a different population.

What 99Visibility does for ChatGPT

How we ask it
OpenAI API, gpt-6-luna, with web search
Web search in our audit
On. Answers come from what it finds on the web when it answers
Weight in the score
36% of platform coverage
Plans
Every plan, Free included
Crawlers checked
OAI-SearchBot, GPTBot, ChatGPT-User: robots.txt, plus a firewall probe for each one that crawls

Asks ChatGPT your prompts, as typed

Each keyword goes to the OpenAI API as gpt-6-luna, exactly as you typed it: no brand name, no instructions, no country. Search is on, so the column reflects what it finds on the web. ChatGPT is on every plan, Free included.

Checks OpenAI's three crawlers

The access check reads robots.txt for OAI-SearchBot, GPTBot and ChatGPT-User, then requests your page as each one to see what your firewall answers. A robots.txt block on OAI-SearchBot is ranked critical; a block on GPTBot alone is ranked low and says it does not affect ChatGPT search.

Flags wrong facts

Every ChatGPT answer is read against the price, founding year, location and features in your brand profile. On Free you see how many answers contradict them; what each flag says, with a severity of critical, moderate or minor, starts on Starter ($49/mo).

Counts the links it gives

The sources ChatGPT cites are stored with the answer, and the citation rate counts the share that point to your site.

A platform note in the fix list

When the ChatGPT column scores more than 20 points under your overall score, the fix list adds a note for it. It tells you to check that the page is reachable by OAI-SearchBot, readable without JavaScript and free of snippet controls, then to look at what ChatGPT cited instead.

Agent V for ChatGPT

Agent V is the assistant in your dashboard, on every plan. It reads your audits, your accuracy flags and pages on your own site, so you can ask what an engine said for a prompt and which fix applies.

Ask for robots.txt lines and V drafts groups for the AI crawlers, OAI-SearchBot and GPTBot among them, each with a note on what it controls. Ask what to write and it drafts page copy that states the corrected fact in its first sentence, ready to paste.

V can also check a page on your site or run a new audit. Each change it proposes shows up as a "V wants to" card and does nothing until you confirm, and V never edits your site: drafts are text you paste yourself.

How Agent V works

ChatGPT optimization: what to fix first

1

Allow OAI-SearchBot in robots.txt. OpenAI: "Sites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers, though can still appear as navigational links."

2

Make sure your CDN or firewall lets OAI-SearchBot through. OpenAI publishes its search crawler's IP ranges at https://openai.com/searchbot.json and warns that bot-protection systems "can mistakenly block legitimate crawlers, often returning 403 Forbidden errors."

3

Block GPTBot only if you mean to opt out of training. OpenAI states each setting is independent, so a GPTBot block neither removes you from ChatGPT search nor helps you into it.

4

Put prices, product names and your location in the HTML your server sends. Check by loading the page with JavaScript turned off.

5

Make every page that states a fact state the same one. If your pricing page and your FAQ disagree, a search can retrieve either.

ChatGPT optimization FAQ

How does ChatGPT decide which brands to name?
With search off, from training data, and OpenAI does not publish how a brand gets into it. With search on, from pages it retrieves and ranks "using multiple factors", which OpenAI does not list. Anyone selling a list of ChatGPT ranking factors is guessing.
Why does my ChatGPT score not move after I fix my site?
Our audit asks ChatGPT with web search on, so a fixed page reaches the column once OAI-SearchBot fetches it again. Re-run the audit after the crawler has been back.
Do you check the ChatGPT app or the API?
The API, with gpt-6-luna, the model reported as the default for ChatGPT's Free and Go tiers. Paid tiers run other models, and the app can search when it chooses, so read this column as one tier.

Sources

OAI-SearchBot, GPTBot and ChatGPT-User, what each is for, and that "each setting is independent of the others".

Source: OpenAI crawler documentation →

Eligibility for ChatGPT search: OAI-SearchBot allowed, and your host or CDN letting OpenAI's searchbot IP addresses through. Also how search rewrites a question into queries.

Source: OpenAI Help Center, ChatGPT search →

ChatGPT and Claude crawlers fetch JavaScript files (ChatGPT: 11.50%, Claude: 23.84% of requests), but "they don't execute them. They can't read client-side rendered content."

Source: Vercel, The rise of the AI crawler (December 2024) →

The May 18, 2026 knowledge cutoff for gpt-6-luna.

Source: OpenAI model documentation →

Related

The other platforms we audit

See what ChatGPT says about your brand

The Free plan audits ChatGPT and Google AI Overviews and checks AI crawler access on your site. No card needed.

Start for Free