Answer engine optimisation is just SEO that admits the answer moved

The short version

Answer engine optimisation (AEO) is what you get when good SEO points at a different surface. The answer moved: instead of a list of blue links, it's now a 150-word AI summary at the top of the page. The inputs that earn that summary haven't changed much. Clear answers. Structured content. Entity authority. Direct, liftable prose. If your SEO was already solid, the incremental work to do AEO is smaller than the vendors charging for it would like you to believe.

What actually changed

Three things genuinely shifted when answer engines got good enough to matter.

The surface. Google's AI Overview, Perplexity, ChatGPT Search and similar don't serve a ranked list of pages. They synthesise an answer and attribute it. Getting into that answer is the new objective, and it looks different from ranking on page one. You can have a top-three ranking and still never be cited in the AI summary above it.

The metric. You used to measure rank and click. You now need to measure citation presence: does the AI mention you, quote you, link to you in its summary? That's a new measurement problem, and the tools for tracking it are still immature. But the metric changing doesn't mean the work that earns the metric has changed.

Verbatim extractability. This is the most practically important shift. Answer engines don't rewrite your content, they lift from it. A clean, self-contained 50-word answer to a question, early in the relevant section, is what gets quoted. A page that circles the answer across four paragraphs may still rank, but it rarely gets cited. This is a real marginal difference from how you optimised for a click.

That's it. Three things. The surface, the metric, and extractability. Everything else being sold as AEO is SEO with a new name on the invoice.

What didn't change

Authority signals. Structured content. Clear heading hierarchy. FAQ schema. Consistent entity signals across the web. These have been in every competent SEO checklist since at least 2019. They're in AEO guides now too, usually with slightly shinier language and a higher day rate attached.

"How is AEO different from SEO?" is one of the most-asked questions on this topic. The most honest answer is also the briefest: it's the same discipline targeting a different surface. If you've been doing SEO properly, you've been doing AEO, even if nobody called it that. If you haven't been doing SEO properly, AEO isn't going to fix it. You can't skip the fundamentals by rebranding them.

The vendors who package AEO as a new discipline you urgently need a new budget for are, in most cases, charging you twice for the same work. That's worth knowing before you sign anything.

The two things that are genuinely new

The metric is new but it's downstream of the work. The two things that actually change what you do day to day are:

Entity co-occurrence. For an AI to cite you by name, it needs to consistently associate your brand with your topic across many sources: your site, LinkedIn, third-party coverage, forum discussions, reputable directories. One well-optimised page isn't enough. The model needs to have encountered "Dog on the Table" and "AEO" and "GEO consultant" together enough times, in enough different places, to be confident associating them when a user asks. This is less of a technical SEO problem and more of a content distribution and presence problem. It's also the part most AEO guides skip entirely because it takes longer to sell than a keyword checklist.

Extractability at the passage level. Classic SEO optimised for the page. AEO optimises for the passage. If your answer to a question starts 300 words in, after a section intro and a definition, you're less likely to be cited than the source that opens with the answer. The discipline is the same (answer the question well), but the structural expectation is more specific: get the clean, standalone answer in the first 50-100 words of each section, phrased so a machine can lift it verbatim without losing meaning.

Is AEO different from GEO?

Related, but not the same. AEO focuses on getting your content into AI answer boxes. GEO (generative engine optimisation) is the broader frame: it includes AEO but also covers entity-level visibility, brand citation tracking, and shaping what AI systems say about your brand across every surface, not only in direct answers.

In practice, the work overlaps heavily. I treat them as part of the same offer rather than separate products (the GEO/AEO service page has the specifics). GEO is the more complete frame if you care about what the models actually say about you, not only whether you appear in the answer box for one query.

What this means if you're already doing SEO

If your SEO is in good shape: make your key answers extractable at the passage level, check whether your brand is consistently associated with your topic across enough external sources, and add citation tracking to your measurement stack. That's the incremental work. It's real and worth doing, but it doesn't require buying AEO as a separate service if you already have competent SEO in place.

If your SEO isn't in good shape: fix it first. No amount of AEO-specific work will earn you AI citations if your content doesn't have the authority and structure to deserve them. The answer engine has the same quality bar as Google, it just has a different way of showing who passed it.

And if someone is quoting you a significant fee to run "AEO as a separate programme" alongside your existing SEO, ask them to be specific about what they'd do differently. The honest answer will usually be a short list, mostly things your SEO should already be doing.

I run an LLM visibility audit that shows exactly where you stand across ChatGPT, Claude, Gemini and Perplexity right now. That's measurement, not a magic new methodology. The honest pitch: here's what the AIs actually say about your brand, here's the specific gap, here's the work to close it.