AEO, GEO, LLMO and AI SEO: which of these are real

Four acronyms arrived in about eighteen months for what is largely one job. Here is what each one means, where they genuinely differ, and what the platform owner says about all of it.

The definitions

AEO, Answer Engine Optimisation. Structuring a page so an assistant can lift a direct answer out of it and attribute that answer to you. The unit is the page. In practice: the answer in the first paragraph rather than the eighth, headings that state a claim instead of teasing it, and something on the page that only you could have written.

GEO, Generative Engine Optimisation. Being one of the names a model produces when it composes an answer from several sources. The unit is the entity, not the page. This depends far more on consistency and corroboration across the web than on anything you can change inside your own site.

LLMO, LLM Optimisation. A fourth label for GEO’s territory. No meaningful difference in the work.

AI SEO. Usually a vendor’s umbrella for the above, and sometimes for something else entirely: using AI tools to do conventional SEO faster. Worth asking which is meant.

Where AEO and GEO actually differ

This is the one distinction worth keeping, because it predicts which half of a project will fail.

AEO gets a page quoted. GEO gets a company recommended.

You can do AEO perfectly and still fail GEO, and most businesses do. A measurement published in September 2026 puts a number on the gap. Across roughly 35,000 prompt and engine observations, when neither a brand nor its domain appeared in the retrieval path, the brand was mentioned about 3% of the time. When its own domain was cited, mention rose to around 49% on one engine and 58% on another. The page mattering is real, but what decides whether the page is in front of the model at all is retrieval, which is mostly a function of everything outside your site.

The author of that study disclosed that he runs an AI-search measurement company, and labelled the model predictive rather than causal. Both caveats are worth carrying.

What Google says about the category

Google published a guide, Optimizing your website for generative AI features on Google Search, last updated 10 July 2026. On AEO and GEO by name:

From Google Search’s perspective, optimizing for generative AI search is optimizing for the search experience, and thus still SEO.

The paragraph then links to Google’s guidance on evaluating third-party SEO advice.

The same guide lists several popular tactics under things you can ignore. On machine-readable files:

You don’t need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search (including its generative AI capabilities), as Google Search itself doesn’t use them.

On markup:

Structured data isn’t required for generative AI search, and there’s no special schema.org markup you need to add.

And on chunking, a staple of AEO checklists:

There’s no requirement to break your content into tiny pieces for AI to better understand it.

This is Google’s own documentation describing Google’s own surfaces. It says nothing about how ChatGPT or Claude choose sources, which are different systems with different retrieval. But three of the four most-sold AEO tactics are explicitly disclaimed by the largest of these systems, and that is worth knowing before buying any of them.

What the terms are useful for

Despite all of the above, I use AEO and GEO when talking about projects, for one reason: they name the two halves of a budget.

The AEO half is work you can do, on a schedule, with a known cost: render without JavaScript, make the page crawlable and fast, answer the question early, keep the identity unambiguous. It is finite and it finishes.

The GEO half is slow, partly outside your control, and never finishes: being written about, listed, compared, reviewed and named by other people. It is also where most of the result comes from.

A proposal that treats these as one line item is hiding which one it is actually doing. That, rather than any dispute about terminology, is why the distinction is worth keeping.

A note on the acronym count

Four labels in eighteen months is a signal about the market rather than about the work. The number of names for a thing tends to track the number of people selling it.

The underlying job has not changed much: have something worth citing, make it readable by a machine, and be known elsewhere. Two of those are old, and the third has always been the hard one.