What is generative engine optimisation?
Generative engine optimisation is the work of getting a business named in the answers AI assistants write. Where search returns a page of links to choose from, an assistant writes a paragraph and names a handful of businesses in it. GEO is about being one of them.
How is that different from being on Google?
A search engine gives you ten links and lets you choose. An assistant reads the sources itself and hands back one answer naming two or three businesses. There is no second page, no position four to settle for, and nothing to click through.
That compression is the whole thing. Ranking eleventh on Google still puts you in front of somebody who scrolls. Being the eleventh most quotable business for a question an assistant answers in eighty words puts you nowhere at all.
It also means the old measurement does not transfer. There is no rank to look up, because there is no ranking -- only whether you were named this time, and how often across many times.
Does anyone actually ask AI for local recommendations?
Enough to matter and not yet enough to panic. Around 65% of Americans have used AI search, and about 15% say they trust it a lot. The honest position is that this is early, growing, and cheap to be early on.
We would rather say that plainly than quote a number designed to frighten you. The argument for doing this now is not that AI has replaced search. It is that the businesses being named today were chosen by systems reading pages that already exist, and changing what those systems find is not a race that anybody in most local markets has started.
What does the work actually consist of?
Three things, roughly in order: making the site readable by machines, saying plainly what the business is and where, and existing on the sources assistants quote. None of it is exotic. Most of it is unglamorous and most sites have not done it.
There is no trick in the list and there is not supposed to be. The advantage is not secret knowledge; it is that this is measurable, most businesses have not done it, and doing it properly takes an unglamorous couple of months.
- Readable: the words are in the HTML, and the crawlers are not blocked from fetching them.
- Stated: structured data declaring what the business is, where it is, and what it sells -- rather than leaving a model to infer it from marketing copy.
- Answer-shaped: pages about one question each, opening with the answer, because retrieval works on passages rather than pages.
- Present: the directories, profiles and round-up pages that AI answers about local businesses are actually built from.
Is GEO just SEO with a new name?
No, though they overlap. Both care whether a machine can read your site. They diverge on what gets measured, what wins, and which sources matter -- a business can rank first on Google and never be named by ChatGPT, and the reverse happens too.
The clearest single difference: Google renders JavaScript before indexing a page, and the AI crawlers do not run it at all. A site that builds its content in the browser can be perfectly indexed and effectively blank to an assistant.
Common questions
Do I need to do GEO if my SEO is already good?
Good SEO helps and does not cover it. The most common gap we measure is a site that Google reads perfectly and the AI crawlers see as almost empty, because those crawlers do not execute JavaScript.
How long does GEO take to show a change?
Pages need to be crawled and indexed before they can change an answer, which is usually two to six weeks. Anyone re-measuring the week after the work goes live is measuring nothing.
Can anyone guarantee my business gets named?
No, and nobody honest will try. What can be done is measuring before and after with enough repeats to tell a real change from random variation, and reporting the range rather than one flattering number.
Sources
- Vercel and MERJ, AI crawler behaviour across 500M+ fetches
- Aggarwal et al., 'GEO: Generative Engine Optimization' (2023)
Where a figure here comes from one study we say so, because one study is one study. Where something rests on practice rather than published work, we say that instead.
If this answered one question, these answer the next ones.
GEO vs SEO: what actually differs?
SEO competes for a position on a page of links. GEO competes to be quoted inside a written answer. They share the groundwork -- a readable, well-structured site -- and diverge on what is measured, which sources matter, and what counts as winning.
How does an AI assistant decide which businesses to name?
It searches, reads a handful of pages, and writes an answer from what those pages say. So the question is not 'how do I rank' but 'which pages does it read, and is my business on them'. Usually those pages are directories, reviews and round-ups rather than your own site.
Why does AI name a different business every time you ask?
Because the answer is generated, not looked up. Ask the same question four times and you often get four overlapping but different lists. That variation is not a bug to be explained away -- it is the reason a single answer tells you almost nothing.