Here is an uncomfortable test. Open your site, view the raw HTML source — not the rendered page, the actual source — and ask: if all you had were these characters, would you know who this company is, what it sells, and whether to recommend it?
For most brands, the honest answer is no. And that is precisely what an AI agent sees.
The JavaScript problem
Modern sites render content in the browser. A human waits a beat and sees a polished page. An AI crawler often gets a near-empty shell — <div id="root"></div> and a script tag — because it does not execute your JavaScript the way a browser does. Your beautiful homepage extracts to nothing.
The entity problem
Even when the text is readable, models need to know who is speaking. If your identity is inconsistent — a different company description on every page, no Wikidata entry, no sameAs links tying your profiles together — the model cannot confidently ground "you" as a real entity, so it hedges or omits you.
The citability problem
Assistants quote atomic claims: short, self-contained, verifiable sentences. Pages built from vague positioning ("we deliver world-class synergistic solutions") give a model nothing it can lift into an answer. No quotable claim, no citation.
How to tell in 60 seconds
You do not need to read your source by hand. Run your URL through the Legible Readiness Index — it scores how findable, citable, and machine-readable your site is across eight dimensions, and shows you the specific gaps. Most brands are surprised by how low they start.
The good news: invisibility to AI is a solvable, mostly technical problem. The brands that fix it now will be the ones AI recommends while everyone else wonders where their traffic went.