Why is my brand not showing up in ChatGPT?

Usually one of four things. ChatGPT cannot crawl your pages, your content is locked behind JavaScript it does not run, you rank in Google but not in the sources ChatGPT actually pulls from, or nothing outside your own site confirms you exist. Each cause is measurable, and each has a specific fix.

Being absent from ChatGPT is not a mystery, and it is rarely bad luck. With more than 900 million weekly active users by early 2026 (OpenAI), ChatGPT now sends real referral traffic to websites: AI platforms drove 1.13 billion referral visits to the top 1,000 sites in June 2025, up 357% year over year, and ChatGPT accounted for over 80% of them (Similarweb data via TechCrunch). If your brand never appears, you are missing a channel that is growing while it is still cheap to win.

Here is the diagnostic in one table, then each row in detail.

Is ChatGPT even reading your site?

Two blockers stop it before content ever matters. If your robots.txt disallows OAI-SearchBot, OpenAI says your site will not be shown in ChatGPT search answers. And because AI crawlers do not execute JavaScript, anything rendered client-side is invisible to them, even when a normal browser shows it fine.

OpenAI runs separate crawlers for separate jobs, and the naming trips people up. GPTBot gathers training data. OAI-SearchBot is the one that surfaces sites inside ChatGPT's search answers. The distinction is not cosmetic: OpenAI's own crawler documentation states that "sites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers, though can still appear as navigational links" (OpenAI). Block that one bot, on purpose or by inheriting a template, and you have quietly removed yourself from the answer layer.

This happens more than you would guess. On the live tracker that watches the robots.txt files of major news publishers, 632 of 1,154 sites block at least one of the OpenAI, Google AI, or Common Crawl bots (News Homepages tracker). Many of those blocks were deliberate. Plenty were copy-pasted without anyone deciding to embargo ChatGPT. We cover the full picture of which bots honor these rules in Do AI crawlers respect robots.txt.

The second blocker is quieter still. Vercel analyzed real crawler traffic across its network and found that "none of the major AI crawlers currently render JavaScript." GPTBot alone made roughly 569 million requests in a single month, and it read raw HTML every time (Vercel). If your pricing, your product copy, or your key claims only appear after a client-side framework hydrates, the crawler sees an empty shell. Your browser lies to you here: it runs the JavaScript, so the page looks complete to you and blank to the bot.

Does ranking in Google put you in ChatGPT?

No. ChatGPT's citations and Google's top results overlap far less than most teams assume. Ahrefs found that only 12% of AI-cited URLs rank in Google's top 10 for the same prompt, and 80% do not rank anywhere in Google for it. A strong Google position does not reserve you a ChatGPT seat.

The comfortable belief is that good SEO carries over automatically. The data says it mostly does not. In Ahrefs' study of 15,000 long-tail queries, ChatGPT's in-text citations overlapped with Google's top 10 only 8% of the time (Ahrefs). A separate analysis of the 1,000 URLs ChatGPT cited most in one week found that 25% of them had zero organic visibility in Google at all (seoClarity).

Read those two numbers together and the takeaway is uncomfortable but useful: ChatGPT is running a different selection than Google, so a page that dominates classic search can still be missing from the answer, and a page that ranks nowhere can get cited. That is why the two disciplines need separate scoring, which we lay out in GEO vs SEO. Optimizing only for the ten blue links leaves the answer engine unaddressed.

Why does ChatGPT cite competitors instead of you?

Because the model trusts what other sites say about you more than what you say about yourself. In Muck Rack's analysis of 25 million cited links, earned media accounted for 84% of all AI citations. When review sites, directories, and comparison lists do not mention your brand, ChatGPT has little reason to reach for it.

This is the reason most brands find hardest to accept, because it is not fixable on your own domain. Muck Rack's ongoing study, built on more than 25 million links cited by ChatGPT, Claude, and Gemini, puts earned media at 84% of citations (Muck Rack). Your homepage is not where the model looks first. It looks at who else vouches for you.

There is corroborating signal from the ranking side. Ahrefs correlated AI visibility against on-page and off-page factors across 75,000 brands and found that branded web mentions correlated with AI visibility at 0.664, more than three times the 0.218 correlation for backlinks (Ahrefs). That study measured Google's AI Overviews rather than ChatGPT specifically, so treat it as directional for the broader mechanism rather than a ChatGPT number. The mechanism it points to is consistent: being talked about across trusted sites matters more than the links pointing home.

Format of the source matters too. When Ahrefs looked at what ChatGPT cites for "best X" style prompts, 43.8% of the cited pages were listicles and comparison posts (Ahrefs). In a separate Ahrefs study of ChatGPT's 1,000 most-cited pages, 28% of them had no organic search visibility of their own (Ahrefs). If every "best tools for" roundup in your category omits you, the model is being handed a shortlist you are not on. Getting into those lists, and earning the mentions behind them, is the off-page half of how to get cited by ChatGPT.

Is your content written the way answer engines lift it?

Answer engines quote self-contained passages, not whole pages. A Princeton study found that adding citations, quotations, and statistics to a page raised its visibility in generative engines by up to 40%, while keyword stuffing scored worse than doing nothing. On this axis, format is a lever, not decoration.

Even a crawlable, well-corroborated page can go unquoted if the model has to reconstruct your point from scattered sentences. The peer-reviewed GEO study out of Princeton tested content edits against real generative engines and reported gains of up to 40% in visibility from three specific moves: citing sources, adding direct quotations, and adding statistics. Tested live on Perplexity, the quotation edit alone lifted visibility about 22%, and old-school keyword stuffing performed roughly 10% worse than the untouched baseline (Aggarwal et al., arXiv).

The practical shape that follows is the one this article uses on itself:

• Put a question a buyer would actually type in the heading.
• Answer it in the first 40 to 60 words, self-contained, so the passage survives being quoted without the paragraph around it.
• Back the claim with a real number and a live source link.
• Keep one comparison table or definition list per piece so the facts sit in a structure a model can lift cleanly.

This is the same passage format behind answer engine optimization. It reads a little repetitively by design, because the repetition is what makes each block quotable on its own.

Will schema markup fix it?

Probably not on its own, despite how often it is sold as the fix. The evidence that schema drives AI citations is weak, and one controlled Ahrefs test saw Google AI Overview citations fall 4.6% after JSON-LD was added. Use Organization schema for clean entity data, but do not expect markup alone to buy you citations.

Schema is the tempting answer because it feels technical and complete, a box you can tick. The honest reading of the evidence is more skeptical. A review of roughly ten studies on schema and AI citation found no credible proof that markup improves citation rates once you control for where a page already ranks; in one Ahrefs test, pages that added JSON-LD saw their Google AI Overview citations drop 4.6% against a control group, while movement on ChatGPT was statistically zero (evidence review by Daniel K. Cheung). Structured data still earns its place for entity clarity and for shopping results, so add Organization markup. Just do not present it to your team as the reason citations will arrive. Nothing here is guaranteed by a script tag.

How do you know if you are showing up at all?

You test it, on a schedule. ChatGPT answers change from run to run, so a single screenshot proves nothing. The honest measure is a dated matrix of buyer questions, asked repeatedly, recording whether you were cited each time. The delta across a fix cycle is the only proof the work moved anything.

Because the output is generative, one lucky citation and one disappointing one can both be noise. Muck Rack found ChatGPT cites sources in about 96% of responses, averaging five citations per answer (Muck Rack), so there is a seat to win on nearly every query. Whether you hold it is a question you answer with repetition, not a hunch.

That repeatable before-and-after is what BeCited automates: it scores whether AI crawlers can reach your pages, checks your content against the citable-passage rubric, and re-audits on a cadence so you can see the change after a fix instead of guessing. Nothing unmeasured is presented as proof, and exactly what gets measured is public on the methodology page. Start by finding out which of the four causes is yours, then fix that one. The rest of the library, from crawler rules to the passage format, is on the blog.