SEO vs GEO: where search is going

Classic SEO optimised for a ranked list of links. Generative engines optimise for an answer they can assemble and attribute. The first rewards keyword coverage; the second rewards clear, quotable statements of fact.

The good news is that they are not in conflict. A page that states plainly what a business does, for whom, where and at what price does well in both. The vague, adjective-heavy homepage does badly in both.

We put concrete claims in short, self-contained sentences, because that is the unit an engine lifts. We keep prices, service areas, timelines and contact details as text rather than baked into images. We use headings that read like questions people ask.

We also mark up the obvious things (organisation, services, FAQs), not as a growth trick, but because structured data is the cheapest way to be unambiguous about facts. We do not mark up reviews on our own pages either way: Google only credits genuine third-party review markup, never a business rating its own work.

Google's own guidance sets the ceiling here: there is no special schema.org markup that makes a page more likely to appear in AI Overviews or AI Mode (Google Search Central, developers.google.com/search/docs/appearance/ai-features). Structured data is table-stakes hygiene, not a ranking lever, in AI search the same as in classic search.

Word-count targets, keyword density, and any tactic whose only justification is that a crawler might like it. If a paragraph exists for the machine and reads badly to a person, it hurts you on the metric that matters, which is whether the visitor gets in touch.