Guides
Why AIs recommend your competitors (and not you)
You ask ChatGPT the question one of your customers would ask, and the answer names three competitors — not you. Frustrating, but it is neither luck nor a conspiracy: the way an AI selects brands follows a mechanism. And what follows a mechanism can be worked on.
How an AI picks the brands it names
When asked for a recommendation, an AI draws on what it has read: your website, comparison articles, customer reviews, press, reference directories, forums. The more clearly a brand is described, in the places AIs consult, the more likely it is to appear in an answer.
It is not about size. A study presented at the KDD 2024 conference by researchers from Princeton and Georgia Tech shows that better-structured, better-sourced content can gain up to 41% visibility in AI answers. Modest but well-documented companies overtake bigger but unreadable players.
The most frequent causes of absence
Across the pre-diagnoses we run, the same causes come up. The website is hard for AI crawlers to read: key information locked in images, pages that only load through scripts, access accidentally blocked. Information is scattered or contradictory: the activity described differently across pages, prices nowhere to be found, an address that differs between the website and directories.
Another major cause: no third-party sources. If nobody but you talks about you — no reviews, no articles, no comparisons — the AI has no reason to prefer you over a competitor documented by ten sources. Finally, staleness: content untouched for years signals a dormant brand.
What recommended brands do
Brands that AIs treat well share simple habits. Each of their pages answers a real customer question, with a direct answer in the first lines. Their key information — activity, offers, prices, service area — is stated in plain text, identical everywhere. They maintain reviews and appear in the sources AIs read for their sector.
And they keep things current: recent content is picked up by AIs far more than old content. Freshness is not cosmetic — it is one of the strongest signals.
Where to start
Before fixing anything, measure. Without a baseline you optimise blindly — and you will never know whether the effort paid off. Our free pre-diagnosis gives you that starting point: your presence rate in AI answers on your market, the competitors named instead of you, and what AIs say when asked about you.
Frequently asked questions
Can you pay to be named by ChatGPT?
No. Organic AI answers are not for sale like ad space. Presence in answers is built through the quality and consistency of what is published about you — which is good news: the playing field is fairer than advertising.
My competitor is smaller than me yet gets named. Why?
Because AIs do not measure company size, they read what is published. A smaller competitor described clearly, with reviews, comparisons and consistent information, is easier to pick up in an answer than a discreet market leader.
How long does it take to appear in answers?
Web-search answers can shift within weeks after you update your content and sources. The models' memory updates with each new version — several times a year. It is groundwork, measured through regular re-testing.