What is answer engine optimisation (AEO)?

Last updated 1 September 2026

TLDR

Answer engine optimisation is the practice of structuring content so that AI assistants such as ChatGPT, Claude, Gemini and Perplexity quote or name your business when someone asks a question in your category. It shares its foundations with SEO — crawlability, credibility, speed — but rewards answer-first formatting, explicit definitions and attributable facts rather than keyword placement.

Key facts

AEO targets AI assistants; SEO targets ranked lists of links. Most of the technical groundwork is shared.
Assistants favour content that answers a question in its first sentence and defines its terms explicitly.
Being quoted depends partly on third-party sources the assistant retrieves, not only on your own site.
Assistant answers are non-deterministic, so measurement uses mention rate across repeated runs, not a fixed position.
Blocking AI crawlers in robots.txt removes you from the retrieval pool for assistants that honour it.

How is AEO different from SEO?

SEO optimises for a page appearing in a ranked list of links, where the click is the goal. AEO optimises for a passage being extracted and repeated inside a generated answer, where the mention is the goal and the click may never happen.

In practice the technical requirements overlap almost entirely. A page that is uncrawlable, slow or thin will fail at both. Where they diverge is structure: AEO rewards a direct answer in the first sentence under a question-shaped heading, clearly labelled facts, and named sources, because those are the units a language model can lift with confidence.

What does an AEO-formatted page look like?

A short summary block at the top that answers the question outright. A list of key facts stated plainly. Headings phrased as the questions a person would ask, each answered in its opening sentence before any elaboration. Definitions given explicitly rather than assumed. Sources named, with what each supports.

It also means structured data: FAQPage or QAPage markup where the content genuinely is a question and answer, and speakable markup on the summary. Markup does not make a weak page quotable, but it makes a good one easier to parse.

Do I have to choose between SEO and AEO?

No, and treating them as separate programmes usually wastes money. The same article can rank in classic search and be quoted by an assistant; the changes AEO asks for — get to the point, define your terms, cite your facts — tend to improve conversion and readability as well.

The one genuine trade-off is length padding. Content written to hit a word count with introductory throat-clearing performs worse for extraction. Cutting it does not hurt rankings.

How do you measure whether AEO is working?

Run a fixed set of buyer-style prompts against each assistant on a schedule and record how often your business is named and which of your pages is cited. Track the mention rate over time rather than a single result, because the same prompt can return different names on different days.

Secondary signals worth watching: referral traffic with assistant user agents or referrers in your analytics, and direct or branded search lifting after an assistant starts naming you.

Does AEO replace traditional search traffic?

Not today. Classic search still sends the large majority of discovery traffic for most small businesses. What has changed is that a meaningful slice of high-intent research now happens inside an assistant, and that slice is invisible to standard analytics unless you measure it deliberately.

The sensible position is to treat AEO as an addition to search work rather than a replacement for it.

Definitions

Answer engine: A system that responds to a question with a synthesised answer rather than a list of links — for example ChatGPT, Claude, Gemini or Perplexity.
Mention rate: The proportion of repeated runs of the same prompt in which a given brand is named.
Citation: A specific URL an assistant attributes part of its answer to.

Sources

Schema.org FAQPage and QAPage specifications — Definitions of the structured data types used for answer content.
Published crawler documentation from OpenAI, Anthropic, Perplexity and Google — User agent names and robots.txt directives referenced above.

Want this checked on your own site?

The free scan shows your technical health and whether assistants mention you.