Being quotable: how AI is changing corporate visibility
When people stop clicking results and read a generated answer, the question is no longer whether they find you. It's whether they cite you.
For two decades digital visibility was measured in positions: first result, third place, second page. The assumption was that someone read a list and chose. That assumption is weakening. A growing share of searches now ends on a synthesised answer that cites two or three sources and ignores everything else.
The shift is more radical than it looks. In traditional search you competed to be findable. Now you compete to be quotable, which is a different property and much harder to buy.
What makes a company quotable
Proprietary data. A generative system cites whoever brings a number that doesn't exist elsewhere. An annual sector survey, an observatory, a time series: assets built once and reused for years.
Attributable positions. Generic statements don't get cited because they add nothing. Positions traceable to a person with a verifiable role do.
Consistency across sources. If your website, your company profile, your public bio and your press coverage say slightly different things, the system doesn't know which version to report — and when in doubt it reports the competitor who is consistent.
Presence in credible third-party sources. One mention in a recognised outlet is worth more than twenty posts on your own blog. Not new: it's the old logic of authority, applied to an automated reader.
What doesn't work
Filling the site with optimised text doesn't produce quotability, it produces noise. Volume was a lever when production was expensive and signalled investment; now that anyone can generate five hundred pages in an afternoon, it signals nothing.
How to check, concretely
Define eight to ten questions a prospect would genuinely ask about your category — not about your name, about the category. Put them to the main generative answer systems. Note who gets cited, with what description, and whether the information about your company is correct. Repeat quarterly with the same questions.
It's a crude but honest measurement, and in most cases it produces an uncomfortable discovery: it isn't that the company is absent, it's that it appears badly described, with old figures, or filed under the wrong sector.
The question is no longer "how do we rank". It's "what is said about us when we're not in the room" — and now the room is automated too.
