Getting found when buyers ask AI
In short
Answer engine optimisation, generative engine optimisation and SEO are one discipline under three names, and this is that one service. It has two layers: the entity layer, so an AI system can establish which company you are and attribute what it finds to you, and the passage layer, so individual sections of your pages survive being lifted out and read alone. It is measured on a fixed set of buyer questions re-run across ChatGPT, Perplexity, Gemini, Claude and Google's AI Overviews — answer share and citation rate reported per engine rather than blended into one score, and tracked through to enquiries rather than reported as traffic.
What does this service actually do?
It makes your company one of the sources an AI assistant reaches for when a buyer asks it a question in your category. That splits into two jobs. The first is the entity layer: making it unambiguous which company you are, so a system can tell you apart from every similarly named business and attribute what it finds to the right organisation. The second is the passage layer: making individual sections of your pages survive being lifted out and read on their own, because these systems retrieve passages, not whole pages.
Everything below is one engagement with one measurement model. It is not three products.
Are AEO, GEO and SEO three different services?
No. Answer engine optimisation and generative engine optimisation are two names for the same discipline — GEO is the term that came out of the academic literature, AEO is the term buyers and agencies use. "AI search optimisation" and "AI search optimization" are the same thing again, in plainer words and in both spellings. Search engine optimisation sits underneath all of them: it is a prerequisite, not a rival.
The name SAEOGO is built from those acronyms — SEO, AEO and GEO — precisely because they are one job. Anyone quoting you separately for each is charging three times for one piece of work. If you want the terminology history rather than the service, that is set out on what answer engine optimisation is, and there are longer treatments on generative engine optimisation, on AI search optimisation and on why SEO is still the floor.
Why does ranking well no longer guarantee being cited?
Because being retrievable and being ranked have come apart. Ahrefs found in March 2026 that 38% of pages cited in Google's AI Overviews also ranked in the top ten for the same query — down from 76% in July 2025. Six months earlier, a strong ranking was a reasonable proxy for being quoted. It no longer is.
That is the whole commercial argument for this work. If your existing search performance still bought you visibility inside AI answers, there would be nothing here worth paying for. The measurable gap between the two is what there is to close.
How do I know whether I have a problem at all?
You look, before you buy anything. I run a free AI visibility check: a fixed set of buyer questions put to ChatGPT, Perplexity, Google's AI Overviews, Gemini and Claude, recorded as a walkthrough showing who was named, who was cited, which source earned the citation, and who appeared instead of you.
It also includes a rendering test, because the most common single cause of invisibility is mechanical rather than editorial. Vercel and MERJ tracked over 500 million GPTBot requests in December 2024 and found no evidence of JavaScript ever being executed — GPTBot requested JavaScript files 11.5% of the time and ClaudeBot 23.8%, and neither ran them, with Gemini and Applebot the exceptions. If your pages assemble themselves in the visitor's browser, most of these systems are reading an empty document.
What does the work cover?
Five stages, run in this order. The order is the part that matters: each stage is what makes the next one worth doing. The full reasoning is on the method page.
| Stage | What it fixes | What you end up with |
|---|---|---|
| Entity resolution | An engine cannot reliably tell which company you are, so it attributes your work elsewhere or declines to name you | One canonical description used identically everywhere, Organisation structured data with a stable identifier, and a complete set of real profile links |
| Prompt-space mapping | Nobody has written down what a buyer actually asks, so nothing can be measured | A fixed, agreed prompt set in writing, kept unchanged between runs so results are comparable |
| Passage architecture | Your pages only make sense read top to bottom, and retrieval does not read them that way | Question-shaped headings and self-contained sections that still answer the question when lifted out alone |
| Corroboration | The third-party sources engines cross-check describe you inaccurately, or not at all | Listings, profiles and directory entries that agree with each other and with your own site |
| Measurement | You cannot tell whether any of it worked | Answer share and citation rate per engine, re-run on a schedule, tracked through to enquiries |
How is progress reported?
Per engine, never blended. ChatGPT, Perplexity, Gemini, Claude and Google's AI Overviews retrieve and cite differently enough that a result on one predicts very little about the others, so a single averaged score would hide the only pattern worth acting on. Each report names which source earned each citation, and which competitor was named where you were not.
Technical fixes are reported as a before-and-after against the specific check that failed, with the measurement method stated. There is no projected traffic figure and no forecast ranking, because those would be numbers I made up.
What cannot be promised?
That you will be cited. Answer engines select sources at the moment they assemble a response, their operators change retrieval behaviour without notice, and the same question can return different answers to two people on the same day. What can be controlled is whether the business is identifiable, whether its content exists in a retrievable form, and whether the third-party sources these systems lean on describe it accurately. Anyone offering a guaranteed placement in an AI answer is describing something they do not control.
Who is this not for?
- Companies whose buyers do not research before making contact. If the purchase is impulsive or purely price-led, this is the wrong spend.
- Companies with no clear proposition yet. Retrieval amplifies whatever description exists; it does not invent one.
- Anyone who wants a guaranteed number of citations by a fixed date, because that is not a promise I can keep honestly.
What happens next?
Start with the free check rather than a proposal. If it shows your visibility is already fine, that is what the recording will say and you should not hire anyone.
Visibility audits start at €2,500, and implementation projects typically run €4,500–€9,500. What moves that number, and how engagements are paid for, is set out in full on the pricing page. See pricing
Frequently asked questions
Is answer engine optimisation the same as SEO?
No, but it sits on top of SEO and cannot replace it. Search engine optimisation gets a page indexed and ranked. Answer engine optimisation makes a specific passage on that page retrievable and attributable by a system that answers in prose. A site that classic crawlers cannot read will fail at both, which is why the SEO foundation gets checked first.
Do I need answer engine optimisation if my rankings are good?
Possibly, and the only way to know is to look. Ahrefs found in March 2026 that 38% of pages cited in Google AI Overviews also ranked in the top ten for the same query, down from 76% in July 2025. Strong rankings and strong citation rates have visibly come apart, so the question is now an empirical one about your category rather than an assumption.
Why does acting on this now matter?
Because entity authority compounds. An engine that has repeatedly retrieved and cited a source builds confidence in it, and displacing an answer a model has already settled on is harder than being the source it settles on. The window is defined by how quickly your category consolidates, not by any vendor's timeline.
What does the engagement actually produce?
A written prompt set you keep, structural and entity changes made on your own site and profiles, and a recurring measurement report showing answer share, citation rate, which source earned each citation and who was named instead of you. Everything produced is yours and repeatable without me.
Can you guarantee my company will be cited?
No. Answer engines select sources at the moment they assemble a response, and their operators change retrieval behaviour without notice. What can be controlled is whether the business is identifiable, whether its content exists in a retrievable form, and whether the third-party sources engines lean on describe it accurately.
Find out what the engines currently say about you
Send me your company name and website. I run a set of buyer questions across ChatGPT, Perplexity, Gemini, Claude and Google's AI Overviews, and send back a recorded walkthrough of what came out: where you appeared, where you did not, who was named instead, and the two or three structural reasons why. No charge, no obligation, and you keep the findings whether or not you decide to hire me.
I run these myself, so there is a queue. Expect a few working days rather than an instant report.