What is the difference between AEO, GEO and SEO?

Search Engine Optimization (SEO)

The work of earning a position in a ranked list of links. Measured in positions, clicks and sessions.

Answer Engine Optimization (AEO)

The work of being named inside a composed answer. Measured in mentions per check.

Generative Engine Optimization (GEO)

The same work under the name academic research gave it in 2023. GEO is more common in papers, AEO in the industry; treat the difference as vocabulary, not method.

What actually changed

Not the inputs. The scoreboard.

The signals that made a business findable a decade ago are the same signals a model leans on when it composes a recommendation: a consistent listing, real reviews, a site that answers questions in plain language, corroboration on pages you do not own. What changed is what happens after retrieval. A search engine hands you ten options and lets you choose. An answer engine chooses for you and hands you the result.

• Unit of success: a ranked position becomes a mention inside an answer.
• Denominator: impressions become checks (one question, asked of one engine, at one point in time).
• Feedback loop: a rank tracker reads a public results page; an answer has to be elicited, because there is no page sitting there to read.
• Volatility: the same question asked twice can return different answers, so one observation is noise. Rankings were far steadier than this.
• Attribution: a click leaves a referrer. A mention that ends in someone typing your name into their phone leaves nothing in your analytics at all.

Why measurement is the real break

You can read your position from a results page that exists whether or not you look at it. There is no equivalent artefact for an AI answer. The answer is generated on request, for one person, and then it is gone. The only way to know what assistants say about you is to ask them, on a schedule, and count what comes back.

This is why AEO work that is not paired with measurement drifts into superstition: something changed, the phone rang, a story gets told and repeated. A counted appearance rate, which Apex publishes as your AI Visibility Score, turns it back into work you can check, and occasionally into work you can abandon.

What carries over from SEO

• Technical health. A page a crawler cannot fetch is a page a model cannot retrieve.
• Entity consistency: name, address, phone, hours and categories, identical everywhere.
• Reviews: volume, recency, and whether anyone answers them.
• Content that answers the questions customers actually ask, in their words.
• Being cited on pages you do not control. This mattered before and matters more now.

What does not carry over

• Keyword density. Models read meaning; a padded page mostly signals that the page is padded.
• Position tracking as the headline number. There is no position to track.
• Click-through as proof of visibility. Many answers never produce a click, which does not mean they never produced a customer.
• The assumption that one canonical result exists. Different engines give different answers to the same question on the same morning, and routinely do.

So do you need three strategies?

No. You need one body of work and two scoreboards: the rankings you already track, and an appearance rate across the assistants your customers use. Apex measures the second and publishes it as your AI Visibility Score: ChatGPT, Gemini, Perplexity and Claude, your own customers' questions, on a fixed schedule, always reported with its counts.