Guide
Why Doesn't ChatGPT Mention Your Brand?
Six causes why competitors appear in the answer and you don't — and the order in which to fix them.
Missing from ChatGPT
Diagnosis before therapy: measure first, then fix
You ask ChatGPT for the best vendors in your category — and the answer names three competitors, but not you. That is neither random nor a black box: in almost every case one of six causes is behind it, and each has a different fix. This guide walks through the diagnosis in the right order — from "is the finding even real?" to the target page that should win a buying-intent question.
Step 0: Verify the finding is real
Generative answers vary between sessions. Before deriving measures: ask the same question several times in fresh sessions, without history or personal context. If your brand is missing from one answer in five, you have a variability issue. If it is missing from all of them, you have a visibility problem — and the following causes are the candidates.
The question itself matters too: buying-intent phrasings ("Which agency for X is recommended and what does it cost?") produce vendor lists. Purely informational questions do not — being absent there is normal and no cause for alarm.
Cause 1: Your brand is not an entity for the model
Without web access, ChatGPT answers from model knowledge. Brands only exist there if they appeared often enough in the right context in training material: industry articles, directories, review platforms, Wikipedia, trade press. A brand that only exists on its own website is a foreign word to the model.
Fix: build presence on the third-party pages that define your category — industry directories, comparison articles, expert posts, review platforms. This is slow work with a long lever: it acts on model knowledge and on the sources of web search at the same time.
Cause 2: The cited sources belong to your competitors
With web access, ChatGPT assembles the answer from concrete pages — very often comparison and "best vendors" articles. If those lists feature your competitors and not you, the answer inherits exactly that.
Fix: find out which sources are actually cited for your buying-intent questions (a live check shows them), and work specifically on getting into those lists — instead of doing generic "more PR".
Cause 3: No page answers the question directly
ChatGPT cites pages from which an answer can be extracted: a clear statement in the first sentences, criteria, prices, comparisons. A homepage full of brand poetry ("We reimagine communication") answers no buying-intent question at all.
Fix: assign one target page per buying-intent question and structure it answer-first — the direct answer first, then proof, criteria and internal links. One question, one page, one clear answer beats ten diffuse subpages.
Cause 4: Technology blocks the crawlers
If AI crawlers cannot load your pages, you do not exist for web search — no matter how good the content is. Common blockers: blanket bot rules in robots.txt or in CDN/firewall rulesets, content that only renders via JavaScript, and pages behind logins or cookie walls.
Fix: audit crawlability (our crawlability guide walks through the checklist), allow the relevant bots deliberately, and make sure the core content is delivered server-rendered.
Cause 5: You are visible — just not for the questions that make money
A brand is often present for informational questions ("What is X?") but missing from the buying-intent ones ("Who offers X and what does it cost?"). That feels like visibility but does not convert.
Fix: measure the Money Prompts of your category specifically — the questions immediately before a buying decision. Only that portfolio shows whether your visibility is revenue-relevant or just reach cosmetics.
Cause 6: It changed a while ago — and nobody noticed
AI answers are not static: models get updated, sources change, competitors move in. Whoever checks manually once a quarter notices a loss weeks too late.
Fix: measure recurringly instead of spot-checking — the same prompts, daily, with stored answers and an alert when a mention is lost. The guide "How to track your brand in ChatGPT" shows how to build that series.
The right order
Not everything at once: first the multi-run check (is the finding real?), then look at the sources of real answers (where does the recommendation come from?), then assign one target page per Money Prompt and make it technically and editorially citable. Third-party presence runs in parallel as the long-term lever. And measure from day one — otherwise you will never know which fix worked.
The same approach applies to the Google AI Overview: it cites sources too and prefers pages with extractable answers. Setting up cleanly for ChatGPT almost always improves both surfaces.
Cluster links
Frequently Asked Questions (FAQ)
Can I pay to appear in ChatGPT answers?
No. Organic ChatGPT answers have no ad slot you could buy. The path runs through citable pages, presence in the actually cited sources, and entity building.
How long until something changes?
Source-based fixes (target page, inclusion in cited comparisons) can show up in web-grounded answers within weeks. Model knowledge only changes with new model versions — which is why third-party presence is a long-term lever.
Does an llms.txt guarantee I get mentioned?
No. It is a signpost for crawlers, not a recommendation guarantee. Without citable content and third-party signals, an llms.txt alone changes nothing about the answer.
How do I know which cause applies to me?
With one live check per buying-intent question: it shows whether you are named, who appears instead, and which sources are cited. Those three pieces of information determine which of the six causes applies.
Next step