Why Can't ChatGPT Find Your Brand? 5 Technical Causes Almost Nobody Checks

When a brand doesn't show up in ChatGPT's answers, most people think about content or reputation. But there is an earlier step almost nobody checks: whether AI can even read your site. A misplaced rule in a text file, or content loaded with JavaScript, can keep you out of ChatGPT, Claude, or Perplexity answers with no error and no alert.

Why can ChatGPT miss your brand even when your product is good?

A brand becomes invisible in AI for two kinds of reasons. The first are about authority: few sources talk about you, or the model doesn't recognize you as an entity. We cover those in another blog article. The second are technical and are the subject of this guide: AI can't access your content, or accesses it and sees it incomplete. They are the cheapest to fix and the least checked, because they make no noise: your site works perfectly for any person who opens it.

Cause 1: your robots.txt shuts the door on the wrong bot

OpenAI doesn't use one crawler; it uses three, and each does something different. What decides whether your site can appear in ChatGPT search is OAI-SearchBot. Blocking GPTBot, by contrast, only affects model training and, per OpenAI's documentation, doesn't prevent you from appearing in ChatGPT search. If someone on your team blocked “AI bots” by copying a generic list, that list may well include OAI-SearchBot and remove you from answers without anyone noticing.

• GPTBot: collects content to train models. Blocking it doesn't remove you from ChatGPT search.
• OAI-SearchBot: surfaces sites in ChatGPT's search results. If you block it, your site doesn't show up in those answers.
• ChatGPT-User: triggered when a person asks for something inside ChatGPT. Because the action is user-initiated, robots.txt rules may not apply.

The test takes 30 seconds: open yourdomain.com/robots.txt and look for “Disallow: /” rules under OAI-SearchBot, GPTBot, or “User-agent: *”. The last one shuts the door on every bot that respects robots.txt, Googlebot included, for example if it was left over from a staging version of the site. Once you fix it, OpenAI takes about 24 hours to update its systems.

Cause 2: your content loads with JavaScript and ChatGPT doesn't run it

A study by Vercel and MERJ analyzed how AI crawlers read the web and concluded that none of the major ones (OpenAI's, Anthropic's, and Perplexity's) execute JavaScript. They read only the HTML the server delivers. If your text, prices, or reviews only appear after a script loads, that part of your site doesn't exist for those bots. This is most common on sites built as single-page applications. The data is from late 2024, and we found no evidence that it has changed.

The exception is Google: Gemini uses Googlebot's infrastructure and does render JavaScript. That's why your site can look fine in Gemini and be empty in ChatGPT. The test: open your site, right-click, choose “View page source,” and search with Ctrl+F for a sentence from your main copy. If it doesn't appear, ChatGPT can't see it either. The fix is to deliver important content already written into the HTML from the server, using server-side rendering (SSR) or static generation. It's a request for your development team, and now you know which one.

Cause 3: Google can't show you with a snippet (AI Overviews and AI Mode)

According to Google's documentation, to appear as a supporting link in AI Overviews or AI Mode, a page must be indexed and eligible to be shown in Search with a snippet. There are no additional technical requirements. That means a noindex tag, a nosnippet tag, or a very restrictive max-snippet takes you out of those answers. And blocking Google-Extended is not the control for this: that token limits Gemini training and other uses, not Search. Google also builds its answers with “query fan-out,” running several related searches across subtopics, so your detail pages count, not just your homepage. Check in Search Console that your key pages are indexed and free of those tags.

Cause 4: the sources AI reads don't talk about your brand

Even if your site is perfectly readable, AI doesn't build its answers only from what your website says. In an in-house test in September 2026, we asked Google's AI Mode for the best digital banks in Argentina: it built the ranking from just four sources (Rankia, Ikiwi, a YouTube video, and a news portal), and none was a bank's own site. If those sources don't mention you, it doesn't matter how good your site is. Lumen AI's Authority Sources panel shows which domains each engine cites when answering about your category, so you know where you need to appear.

Cause 5: your test is biased

Many business owners ask ChatGPT about their brand from their own account and draw a conclusion. But the answer can vary with the account, the memory of past conversations, the location, and even between one query and the next. For a first check, use a temporary chat and repeat the question several times. To really know how you're seen, you need to measure systematically and across more than one engine, because ChatGPT, Gemini, and Perplexity can answer the same question differently. That's what Lumen AI does every day for your category, as part of an AI visibility strategy.

Checklist: how to check in 10 minutes whether ChatGPT can read you

1. 1
   Open your robots.txt: Look for “Disallow: /” under OAI-SearchBot, GPTBot, or “User-agent: *”. If there is one, talk to whoever manages your site.
2. 2
   View the page source: Right-click, “View page source,” and search for a sentence from your copy. If it isn’t there, your content depends on JavaScript.
3. 3
   Verify indexing in Search Console: Confirm your key pages are indexed and carry no noindex or nosnippet.
4. 4
   Note which sources talk about your category: Ask AI about your category and see which media, comparison sites, and directories it cites. That is where you need to appear.
5. 5
   Repeat the test in a temporary chat: Run it several times and on more than one engine, and save the answers to compare.