Answer · checked 31 August 2026
Why do AI assistants never mention my business, and how do I fix it?
Usually for one of two reasons, and they need opposite fixes. Either the assistant cannot read your site — the page renders empty without JavaScript, the entity is ambiguous, and there is no description it can quote or action it can name — or it can read the site fine and there is simply nothing on it worth citing. The first is an engineering defect, it is measurable in about ten seconds, and it is fixable in days. The second is a content and evidence problem and no amount of markup solves it. Measure first, because the fixes do not overlap.
Colour International LLC · 31 August 2026 · colourintl.com/learn/get-recommended-by-ai-assistants
Why you cannot see this yourself
Your site looks fine to you because you are looking at it in a browser that runs JavaScript, follows redirects, and already knows who you are. A retrieval crawler does none of that reliably. The failure is silent on both ends: nothing errors, no analytics event fires, and the assistant simply answers with somebody else. There is no notification that you were not considered.
This is why we lead with a scan rather than an opinion. Eight unauthenticated GET requests, the same ones an assistant makes, and the transcript of what each returned is printed on the page.
The failure that is almost always present
Across the samples we have published — ten major booking platforms at a median of 54 out of 100, and nineteen independent salon sites at a median of 56 — the constant is the missing capability manifest: not one site in either sample published one. So even where the content was readable, no assistant could learn that an appointment could be booked, or how.
The second constant is entity ambiguity. If several organisations share your name and your site never states which one you are in machine-readable form, a model that is trying not to be wrong will decline to name you rather than risk naming the wrong company.
What you can do this week, without buying anything
Scan the domain and read the transcript. Then, in rough order of what it usually returns: make the homepage's substance present in the server response; publish an unambiguous identity block naming the legal entity; state your access policy in robots.txt deliberately rather than by default; publish a plain-text summary of what you do; and add a capability manifest describing what can actually be done.
Everything the scan finds is yours to implement at no cost, and the scanner stays free however many times you re-run it. What we charge for is doing the work and keeping the result true afterwards — not for the findings.
Related
What is /.well-known/agent.json and should I publish one?
A capability manifest at /.well-known/agent.json is a small JSON file at a predictable address that states, in machine-readable form, who a site belongs to, what it offers, what an agent is permitted to do, and which operations exist — the thing a site has no other way to say.
Agent readiness vs SEO, AEO and GEO — what is actually different?
SEO, AEO and GEO are all about winning a contest for attention: ranking a page, owning a featured answer, or getting mentioned by a model.
Why do agent-readiness scanners give my site different scores?
Two agent-readiness scanners can score the same site 100 and 31 without either being broken, because they are measuring different things.
Check your own site
Scan any domain at colourintl.com — sixteen weighted checks, free, no account, and the transcript of all eight requests is printed on the result page. Everything it finds is yours to implement at no cost.
This page is also available as Markdown: curl -H "Accept: text/markdown" https://colourintl.com/learn/get-recommended-by-ai-assistants