What is Answer Engine Optimization (AEO)?

Answer engine

A system that replies to a question with a composed answer instead of a page of results. The ones Apex measures are ChatGPT · Gemini · Perplexity · Claude.

Answer Engine Optimization (AEO)

The work of becoming one of the businesses an answer engine names, and of being described accurately when it does.

Mention

One answer that names your business. It is the atomic unit of AEO the way a ranked position was the atomic unit of SEO.

Check

One question, asked of one engine, at one point in time. Checks are what you count; mentions are what you count inside them.

Why the unit of success changed

A search results page shows ten links and lets the person choose. An answer engine usually names one to five businesses and stops. There is no page two, no scroll, and often no click at all. If you are not in the shortlist the model composes, you were never in the running, and nothing in your analytics will tell you it happened.

Ask an assistant "best HVAC company near Austin" and you get a paragraph naming a handful of companies. The person asking now has a shortlist they did not build and cannot see the alternatives to. AEO is the work of being on it.

What an answer engine does with the question

The details differ by engine and change without notice, but the shape is stable:

• It reads the question and decides what kind of answer is wanted: a recommendation, a fact, a comparison.
• It retrieves supporting material: live search results, its own index, licensed data, and whatever the person has already said in the conversation.
• It composes one answer from that material, choosing which names to include and how to describe them.
• It may or may not show its sources, and what it shows is not always everything it used.

Two consequences follow. First, you are competing to be retrieved and then competing to be selected: two different problems with two different fixes. Second, the same question asked twice can produce two different answers, which is why a single check tells you almost nothing and a schedule of checks tells you a great deal.

What the work actually consists of

Two things, mostly. Being unambiguous about who you are and where you operate, so an engine can treat scattered mentions as one business rather than three uncertain ones. And being corroborated on pages you do not own, because what other sources say about you outweighs anything you say about yourself.

Everything after that is sequencing: which contradiction to fix first, which source is doing the most work for the engines in your area, and which answer is worth rewriting this month rather than next. The order is different for every business, and getting it wrong is how a year of effort produces no movement anyone can point to.

How you know whether it worked

The measurable outcome of AEO is appearance rate: the share of checks in which an engine names you. Apex publishes it as your AI Visibility Score, always beside its counts. A percentage without its denominator is a claim, not a measurement.

Apex puts a business's own customer questions to ChatGPT, Gemini, Perplexity and Claude on a fixed schedule and counts what comes back. Repetition is the whole point: it is what makes a movement in the number evidence rather than a coin landing the same way twice.

Where to start

Discover where you stand before you change anything. A free Apex scan puts real customer questions to every engine we measure and comes back with your AI Visibility Score and its counts, the questions we asked, and the names that came back instead of yours. The last of those is usually the most useful thing on the page.