What is answer engine optimisation?
In short
Answer engine optimisation (AEO) is the practice of making a company's information retrievable and citable by AI systems that answer questions directly rather than returning a list of links — ChatGPT, Perplexity, Google's AI Overviews, Gemini and Claude. It is also called generative engine optimisation (GEO); the terms mean the same thing. Because these systems retrieve passages rather than whole pages, each section of a page has to make sense read entirely on its own, and the organisation behind it has to be identifiable as an entity before any of it can be attributed correctly.
What is answer engine optimisation?
Answer engine optimisation, usually shortened to AEO, is the practice of making a company's information retrievable and citable by AI systems that answer questions directly instead of returning a list of links. Those systems include ChatGPT, Perplexity, Google's AI Overviews, Gemini and Claude. It is also called generative engine optimisation (GEO), and the two terms mean the same thing.
How does an answer engine choose which sources to cite?
Answer engines use retrieval-augmented generation. The user's question is converted into a vector, semantically similar passages are retrieved from an index, the top-ranked passages are assembled into a context window, and the model writes an answer with citation markers attached to the passages it used. Crucially, the unit being retrieved is a passage, not a page — documents are split into chunks and each chunk is embedded independently. The model never reads your page the way a person does.
Two consequences follow directly. A section that only makes sense in the context of the section above it will be retrieved without that context and will read as incoherent. And a page can rank well in classic search while none of its individual passages are ever selected.
Is AEO different from SEO?
Yes, and it depends on it. Search engine optimisation makes a page indexable and ranked. Answer engine optimisation makes a passage retrievable, quotable and attributable, and makes the organisation behind it identifiable as an entity. The overlap is real — a site that classic crawlers cannot read fails at both — but the unit of work is different.
The evidence that they have come apart is public. Ahrefs studied Google AI Overviews in March 2026 across 863,000 keywords and 4 million cited URLs and found that 38% of cited pages also appeared in the top ten organic results for the same query. The same study in July 2025 put that figure at 76%.
What actually influences whether a company gets cited?
| Factor | What the evidence says | Source |
|---|---|---|
| Sources you control | 86% of AI citations came from sources a brand can manage or influence — 44% first-party sites, 42% listings and profiles | Yext, October 2025, 6.8m citations across 1.6m answers |
| Citing your own sources | Citing sources, adding quotations and adding statistics produced a 30–40% relative improvement on the study's visibility metric | Aggarwal et al., GEO, ACM SIGKDD 2024, GEO-Bench (10,000 queries) |
| JavaScript rendering | No major AI crawler executes JavaScript; GPTBot fetched JS files 11.5% of the time and ClaudeBot 23.8% without running them | Vercel and MERJ, December 2024 |
| Organic ranking | 38% of AI Overview citations also ranked top ten, down from 76% eight months earlier | Ahrefs, March 2026 |
Does answer engine optimisation actually produce results?
Published results from agencies working in this category show what a functioning AEO programme produces. Every figure below is another company's engagement, reported by the agency that ran it and attributed to that agency by name. None of it is SAEOGO's work, none of it has been independently audited, and none of it is a forecast of what any particular business should expect.
These figures and the section further down on what cannot be guaranteed are not in tension, and it is worth saying why rather than leaving it to the reader. The spread here runs from a 6× change in seven weeks to a 49× change over fourteen months, across different categories, different starting points and different definitions of the metric. A range that wide is precisely why a specific outcome cannot honestly be promised in advance — the variance between companies is larger than anything a method controls. What can be committed to is the measurement: a fixed prompt set, per-engine reporting, and a named source behind every citation, so you can see what moved and what did not instead of being handed a multiple.
| Reported outcome | Reported by |
|---|---|
| 49× growth in LLM referral revenue over 14 months for a B2B technology client, alongside 26× LLM referral traffic | Optimist, published case studies |
| 8× LLM conversions in 8 months for a fintech client, with 25× LLM referral traffic over the same period | Optimist, published case studies |
| 13× LLM-sourced revenue year on year for a retail brand | Optimist, published case studies |
| AI-referred trials from 575 to 3,500+ in seven weeks (6×), with a reported 600% citation uplift across ChatGPT, Claude and Perplexity | Discovered Labs, B2B SaaS case study |
| ChatGPT referral traffic converted at 15.9% against 1.76% for Google organic, October 2024 – April 2025 | Seer Interactive, case study |
Two caveats worth carrying with those numbers. They are self-reported by the agencies that ran the work, which is the normal standard for agency case studies and is not the same as independent verification. And the Seer conversion figure comes with a volume caveat stated in the study itself: AI traffic accounted for roughly 0.07% of organic sessions over the period — a much higher conversion rate on a much smaller base. Deliberately not averaged into a single range, because averaging results across different companies, categories and time periods would produce a number that describes nobody.
How should the ChatGPT conversion figure be read?
As high intent, on small volume, growing quickly — all three at once, because any one of them alone is misleading. Seer Interactive reported ChatGPT referral traffic converting at 15.9% against 1.76% for Google organic between October 2024 and April 2025, and noted in the same study that AI traffic accounted for roughly 0.07% of organic sessions. Read only the first half and it looks like a channel that beats classic search by a factor of nine. Read only the second and it looks negligible.
Both readings are wrong on their own. What the pair actually describes is a small, unusually well-qualified stream of visitors: someone arriving from a generated answer has typically already had their question answered and their shortlist narrowed, which is a different person from someone clicking a tenth blue link. The volume caveat is real today, and it is the honest reason this is not yet a primary channel for most companies.
The direction is what decides whether to act now. 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 — the surface that answers buyers is detaching from the surface you already rank on, and it is doing so over months rather than years. Small and growing fast is a different investment case from small and static, and entering a category while it is still unsettled costs less than displacing an answer an engine has already settled on.
What does an AEO engagement typically involve?
Five things, in this order: establishing the organisation as a resolvable entity; writing down the questions buyers actually ask an assistant; restructuring content so each passage stands alone; making sure the pages exist for crawlers that do not run JavaScript; and correcting the third-party listings and profiles engines cross-check. Measurement runs continuously across all of it. The full method is published here.
What AEO cannot do
It cannot guarantee a citation. Answer engines select sources when they assemble a response, using retrieval systems whose operators change behaviour without notice, and the same prompt can return different answers to different users on the same day. Any offer of a guaranteed position in an AI answer describes something the seller does not control.
It also does not work on a business that is genuinely indistinguishable from its competitors. Retrieval rewards specificity. If nothing on your site says anything a competitor could not also say, the engine has no reason to prefer you, and no amount of structural work fixes that.
Frequently asked questions
What is answer engine optimisation in one sentence?
It is the practice of making a company's information retrievable and citable by AI systems that answer questions directly — ChatGPT, Perplexity, Google's AI Overviews, Gemini and Claude — rather than returning a list of links.
How do answer engines decide what to cite?
They use retrieval-augmented generation. The question is embedded as a vector, semantically similar passages are retrieved from an index, the top-ranked passages are assembled into a context window, and the model writes an answer with citations attached to the passages it used. The unit retrieved is a passage, not a page, which is why sections have to make sense read on their own.
Is AEO the same as GEO?
Yes. Answer engine optimisation and generative engine optimisation describe the same discipline. GEO is the term used in the academic literature, notably the 2024 KDD paper by Aggarwal and colleagues at Princeton and IIT Delhi, while AEO is more common among buyers and agencies.
Does AEO replace SEO?
No, it sits on top of it. Answer engines retrieve heavily from indexed web content, so a page classic crawlers cannot read is a page they cannot cite either. What has changed is that ranking no longer implies citation: Ahrefs found 38% of pages cited in Google AI Overviews also ranked top ten in March 2026, down from 76% in July 2025.
Does answer engine optimisation actually work?
Agencies working in the category publish results that suggest it does. Optimist reports 49 times growth in LLM referral revenue over 14 months for a B2B technology client, and Discovered Labs reports AI-referred trials rising from 575 to 3,500 in seven weeks for a B2B SaaS client. Both are self-reported by the agencies that ran the work rather than independently audited, and neither is a forecast for any other business.
Can anyone guarantee my company appears in AI answers?
No. Engines select sources at the moment they generate a response, their operators change retrieval behaviour without notice, and the same prompt can return different answers to different users on the same day. What can be influenced is whether the business is identifiable, whether its content is retrievable, and whether third-party sources 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.