Skip to content

Answer engine optimisation

Answer engine optimisation is optimising to be cited by AI answers. How AEO, GEO and SEO differ, and how much of it is genuinely new work.

Definition

Answer engine optimisation is the practice of structuring content so AI-generated answers can extract and cite it. Generative engine optimisation describes the same work under a different name. Neither is a separate discipline from SEO: both depend on being crawlable, indexed and useful.

Illustration representing answer engine optimisation

Why it matters

A growing share of informational queries are answered on the results page or inside a chat interface. If a reader never reaches a site, ranking is not the thing being competed for. Being the source the answer was built from is.

The term matters less than the work. What follows is the honest account of which parts are new and which are the same job renamed, because the gap between those two is where most of the money in this category is currently being spent.

How it works

AEO and GEO are used interchangeably in practice. AEO is the older term and covers being the cited answer in any answer surface, including featured snippets. GEO is the newer one and usually refers specifically to generated answers such as AI Overviews and assistants. There is no meaningful methodological difference, and nobody should be charged twice for them.

SEO is the parent. Google's guidance on AI features in Search is explicit that there is no special markup for inclusion and that the standard guidance for Search applies. A page that is not crawlable and indexed, in the sense Search Essentials sets out, is not a candidate for anything.

So the practical split:

WorkNew or not
Crawlable, indexable, fastNot new. Ordinary technical SEO
Genuinely useful, specific contentNot new
Answering the question early and plainlyEmphasis shifted, not new
Short self-contained passages per questionEmphasis shifted
Facts that are checkable and datedEmphasis shifted
Measuring citation rather than rankingNew, and poorly served by tooling

Worked example

The structural pattern that makes a passage extractable is the same one this glossary uses throughout:

What is link equity?

Link equity is the ranking value that passes from one page to another through a link.

then the mechanism, the caveats, the example

A three-drawer chest with a shallow top drawer holding one object and deeper drawers beneath it
Answer first and elaborate after. The shallow drawer on top is what gets lifted out.

A heading stating the question, an answer beneath it that survives being lifted out, then the depth. The reason it works for machines is the same reason it works for a reader in a hurry.

Common mistakes

  • Buying it as a separate service.
    If the proposal does not include crawlability, indexing and content quality, it is not a strategy. If it does, it is SEO.
  • Looking for the markup.
    Google says there is none. Schema markup is worth adding for accuracy, not as an entry ticket.
  • Writing for machines.
    Passages padded with restated questions read badly to people, and people are still the ones who buy.
  • Ignoring measurement.
    Citation is hard to track. Impressions and clicks per query in Search Console remain the measurable part.
  • Blocking AI features in reflex.
    It removes snippet features too.

What we actually do

Answer the question in the first paragraph under a heading that matches it. Cite primary sources and link them. Date anything that can rot. Keep the page fast and crawlable. Build E-E-A-T through work that demonstrates experience rather than asserting it. That is the whole of our answer engine optimisation practice, and the reason it looks like good SEO is that it is.

Let's build

Got something to build?

Tell us what you're making. We'll tell you - honestly - if we're the right team, and how we'd do it.