Why ChatGPT recommends your competitor instead of you
The seven most common reasons AI assistants name a competitor and skip your brand, how to tell which one applies to you, and what to fix first.
Short answer
Usually for one of seven reasons: AI crawlers cannot read your site, your pages never say plainly what you are, your competitor appears on the pages the assistant reads and you do not, you use different words than your buyers, the facts about you disagree across the web, your brand is newer than the model, or the competitor has more independent proof. The sources an answer cites usually show you which one it is.
How do AI assistants decide which brands to name?
An assistant answers a buying question in one of two ways. Without web search, it answers from what it learned during training, which is a snapshot of the web up to a cutoff date. With web search, it runs one or more searches, reads a handful of pages and summarizes them. Most buying questions in ChatGPT, Gemini and Claude now trigger the second mode.
In both modes the logic is similar: the assistant repeats what many independent sources agree on. A brand that is described the same way on its own site, on review sites, in comparison articles and in community threads is easy to recommend. A brand that only describes itself, or that the assistant cannot read at all, is easy to leave out.
So when a competitor is named and you are not, the cause is almost always in one of the seven places below.
Cause 1: Can AI crawlers read your website?
If an assistant cannot fetch your pages, it cannot cite them. Three things block it more often than people expect:
- A robots.txt rule that disallows AI crawlers such as OAI-SearchBot, the crawler ChatGPT search uses (OpenAI documentation).
- A bot protection or firewall setting at your host or CDN that turns away AI crawlers, even when robots.txt allows them.
- Content that only appears after JavaScript runs. Many AI crawlers read the raw HTML and do not run scripts, so a page that looks complete in your browser can look empty to them.
This is the fastest cause to rule out. Our free AI crawler checker tests robots.txt access for the main AI crawlers and checks whether your content is in the HTML.
Cause 2: Does your website say plainly what you are?
Many homepages open with a slogan instead of a description. "Work, reimagined" tells a person very little and an assistant even less. When an assistant looks for "invoicing software for freelancers", it needs a sentence on your site that says almost exactly that.
Write one plain sentence that names what you are, who it is for and where you operate, and use it near the top of your homepage, in your page title and in your company profiles. For example: "[Your brand] is invoicing software for freelancers and small agencies in the EU."
Cause 3: Is your competitor on the pages the assistant reads?
When web search is on, the assistant builds its answer from a small set of pages, often "best tools for X" articles, review sites, directories and forum threads. If your competitor appears on those pages and you do not, the assistant has no material about you to work with.
Look at the sources cited in the answers. The 30-minute manual test shows how to collect them. A comparison article that is cited again and again, and does not mention you, is often the single most valuable page to get onto.
Cause 4: Do you use the same words as your buyers?
Brands often invent a category name for themselves. Buyers do not search for it. If you call yourself a "revenue intelligence workspace" and your buyers ask for "a CRM for a small sales team", the assistant has no reason to connect the two.
Check which words your buyers use in sales calls, reviews and support tickets, and make sure those exact words appear on your homepage and product pages. You can keep your own term, but explain it in the buyer's words.
Cause 5: Do the facts about you agree across the web?
Old prices on a directory, a former company name on a review site, or a feature list that changed two years ago all create conflicting facts. Assistants trust a brand less when sources disagree, and they sometimes repeat the outdated version.
Search for your brand name on the main review sites and directories in your category, and update every profile you control with the same name, description, price range and location.
Cause 6: Is your brand newer than the model?
Each model is trained on data up to a cutoff date. A brand that launched or rebranded after that date does not exist in the model's memory, so it cannot appear in answers given without web search.
You cannot change the training data, but you can win the answers that use web search. Those depend on Causes 1 to 5, which is why new brands should fix them first.
Cause 7: Does your competitor have more independent proof?
Assistants lean on evidence from other people: reviews, case studies, press mentions, comparison articles and community discussions. A competitor with hundreds of reviews and a dozen articles about them will usually beat a brand with a great website and no outside mentions.
Build proof steadily: ask happy customers for reviews on the sites your buyers read, publish real case studies with numbers, and take part in the communities where your buyers ask questions, as yourself and within each community's rules.
Which cause applies to you?
Start from what you see in the answers:
| What you see | Most likely cause | Check first |
|---|---|---|
| You are never named, even for your own niche | Causes 1 or 6 | Crawler access and your launch date |
| You are named, but described wrongly | Cause 5 | Your profiles on review sites and directories |
| You are named only when the question uses your own terms | Cause 4 | The words buyers use for your category |
| The same few articles are cited, and you are not in them | Cause 3 | Those articles and how to get included |
| A competitor wins most answers by a wide margin | Cause 7 | Their reviews and mentions compared with yours |
| Your homepage is never cited, only other sites | Cause 2 | Your homepage headline and first paragraph |
What should you avoid?
Do not try to shortcut proof. Fake reviews, paid placements dressed up as independent articles and mass-produced pages written only for search engines can get you removed from review sites and search results, and they do not create the independent agreement assistants rely on. Google describes what it treats as helpful, people-first content in its guidance for creators.
If you want to see which of the seven causes hold you back without testing by hand, the AnswerSeat free audit checks crawler access and your on-site signals, asks buyer questions across ChatGPT, Gemini and Claude, and shows which competitors are named instead of you.
Questions people also ask
Can I pay to be recommended by ChatGPT?
We are not aware of any way to buy a place in the regular answer. Assistants base it on what they read and learned, so the reliable route is to be easy to read, clearly described and mentioned by independent sources.
How long does it take before ChatGPT starts naming my brand?
Fixes that affect web search answers, such as unblocking crawlers or getting onto a cited article, can show up within weeks. Answers given from the model's memory only change when a newer model is trained, which can take many months.
Why does ChatGPT name my competitor but Gemini names me?
Each assistant uses its own search sources, its own model and its own way of choosing pages. Different answers across assistants are normal, which is why you should test all of them.
Is this the same as SEO?
It overlaps a lot. Crawlability, clear pages and outside mentions help both. The difference is that assistants give one answer with a few names instead of a page of links, so being among the first few sources matters more.
See which brands ChatGPT, Gemini and Claude name instead of you.
The free audit runs buyer questions for you and hands you a prioritised list of fixes.
Run my free auditNo credit card. Results in 3 to 5 minutes.
Comments
No comments yet. Questions and experiences are welcome.