SEO spend versus what the engines answer

If you already invest in being found — service pages, content programmes, a named marketing partner — that spend was built for pages and rankings. Rankings still matter. They are not the same surface as the one a buyer now opens first.

A mid-market buyer who needs a firm like yours increasingly asks ChatGPT or Perplexity before they ask a referrer. The engine does not return a results page of ten blue links. It returns a short answer and a shortlist of names. Being findable in Google and being named in that answer are related jobs. They are not identical.

Your SEO spend was designed for the first job. Service pages target the terms partners care about. Content is written to earn links and sit under those terms. A marketing partner reports rankings, traffic, and conversions from the site. That system is coherent. It assumes the buyer starts on a search results page and clicks through to you.

Models work differently enough that the two can diverge. They trust entities — a firm the model can treat as one thing, with a stable name, places, and services. They prefer extractable claims: plain sentences a page can quote without interpretation. They lean on corroborating sources: directories, press, client mentions, and other sites that repeat the same claim in language a model can reuse. A practice can own the terms it cares about in Google and still be absent when a buyer asks an engine who to hire. The reverse also happens: a firm with a messy site gets named because someone else described them cleanly, repeatedly, in places the model already trusts.

None of this is an argument to stop SEO. Rankings still put you in front of buyers who search. Site structure, technical hygiene, and clear service pages still help models as well as crawlers. Cutting SEO because generative answers exist would be the wrong move. The right move is to stop assuming that ranking spend automatically shows up where buyers now start.

The gap shows up in partner meetings as anecdote. Someone pastes a ChatGPT screenshot. Another partner says they still get Google leads. Both can be true in the same week. Without a locked prompt set and a re-run on a fixed date, you cannot tell whether your SEO investment is translating into being named — or whether another practice is sitting in the slot your rankings never touch.

That other surface is what a Geo Audit measures. Geolab Studio writes the buyer prompts in the language a serious buyer actually types. We run the engines we can score. We write down who is named instead of you, by prompt, and whether your ranking presence and your generative presence line up. The Audit does not replace the site. It does not rewrite your content calendar. It tells you whether the site, and everything around it, is what the engines are using when the question is the category, not your brand.

For a UK mid-market accountancy, for example, the cost of the divergence is quiet. You keep paying for pages that perform in search. A buyer who never searches your brand asks an engine for a firm like yours. Someone else is named. Your pipeline never sees the miss, because there was no click to miss.

Measure both surfaces. Keep the SEO you already buy. Add a clear read on who the engines answer with when the prompt is the job. Then decide what to fix first — on the site, off the site, or in how clearly the practice can be quoted at all.