From SEO to GEO

Generative engine optimization: getting cited by AI

When someone asks ChatGPT who can solve their problem, the answer is not a list of ten links: it is two or three names. We work so one of them is yours.

Request the AI audit Solutions

Search stopped returning a list

For twenty years the job was to rank. Now there is a second question: does it cite you? Answering that one is generative engine optimization, and it is a different job.

A classic search engine shows ten results and the visitor picks. A model reads the same sources and returns a paragraph with two or three names. There is no page two. What changed is not the channel, it is who reads your site: no longer a crawler that ranks, but a model that summarises. And a model cites what it can verify — a consistent entity, structured data, and content that answers the question without preamble.

What generative engine optimization actually is

AI search optimization is not a new layer on top of SEO. It is the same technical work, arranged so a model can read it and cite you.

Entity and consistencyOne name, one phone number and one address everywhere you appear: site, Google Business Profile, directories and profiles. A model that finds two versions cites neither.
Structured dataOrganization, LocalBusiness, Service and FAQ schema. It is the difference between a model interpreting your page and a model reading it.
llms.txt and crawler policyExplicit permission for GPTBot, Google-Extended, PerplexityBot and ClaudeBot, plus a plain-text index of what is on your site and what each page is for.
Citable contentFAQs, comparisons and pricing pages that answer in the first sentence. A model cites paragraphs that stand on their own, not text that depends on the paragraph before it.
Reviews and speedA real review cadence and a site that loads fast and crawls completely. The two signals that already mattered for SEO and now matter twice as much.
Bing Webmaster ToolsBing's index feeds Copilot and ChatGPT search. It is the channel almost nobody sets up, and it is free.

How to measure something that has no ranking

There is no Search Console for AI. So we build one: a fixed prompt set, measured every month, with the evidence kept.

1
Prompt setWe define 20 to 40 real questions from your market: "best electrician in Conroe TX", "how much does an A/C repair cost in Houston".
2
Monthly runThe same questions, every month, across ChatGPT, Gemini, Perplexity and AI Overviews. The set does not change: change the set and there is nothing to compare.
3
Mentions and citationsTwo different metrics. Being named is not the same as being linked, and the second one is what brings visits.
4
In your portalThe history lands in your client portal, next to Google Ads and GA4. One place, not four screenshots by email.

Questions we get before signing

Does this replace SEO?

No, it extends it. Eighty per cent of the technical work is the same — entity, schema, speed, reviews — and it still ranks on Google. What is added is the twenty per cent that lets a model cite you.

How long until I see it?

The first prompt run is your baseline and lands in week one. Changes in how models cite you show up between 60 and 90 days, because they depend on the sources being reindexed.

Can you guarantee ChatGPT will recommend me?

No, and be wary of anyone who does: there is no control panel and no bid for position. What can be done is measure where you stand today, move the signals the models use, and show you the change month by month.

What do I need before we start?

Your own site with real content and a verified Google Business Profile. If either is missing, that is the first job — and we do it.

Start by knowing where you stand today

The opening audit runs your prompt set across all four systems and hands you the picture: where you are named, where your competitors are named, and which signal is missing. No commitment to continue.

Request the audit