Answer engine optimization: getting named in the answer, not ranked in a list.

An answer engine is any tool that replies with a written answer instead of a list: ChatGPT, Perplexity, Google AI Mode, Gemini, Copilot. Answer engine optimization is the work of becoming one of the sources that answer is built from, and one of the names it prints.

SEO aims at a rank. AEO aims at a sentence. That difference runs through everything. A page that ranks well can be long, hedged and full of everything you offer. A page that gets quoted has to state one thing plainly: what the service costs, how long it takes, who it is for, in text, on a page about that one thing.

The second difference is scarcity. Bain found that 60 percent of searches now end without a click to any website (Bain & Company, Feb 2025), and Gartner expects a 25 percent drop in traditional search engine volume by 2026 as assistants take over answering (Gartner, Feb 2024). The clicks are not being redistributed. They are being replaced by an answer, and being in it is now the whole of the job.

An answer engine has to do three things before it will put your name in front of a stranger: read what you do, verify it somewhere it did not get from you, and judge you safe to recommend. Each of the three tools weights that differently.

• ChatGPT looks up the live web through Bing, so a business with no Bing Places listing and a site Bing has not indexed properly is missing from the source material before the answer is even written.
• Google AI Mode is built on Google's index and Google Business Profile, so categories, hours, service areas, review volume and structured data on your own pages all feed what it says about you.
• Perplexity writes short answers with citations attached, which rewards pages that state a fact in plain text and punishes pages that bury it in a brochure paragraph or an image.

None of that is guesswork about model internals. It is what the tools do in the open, and it is why the same fixes move all three. Our guide on how ChatGPT search runs on Bing walks through the first of those three in detail.

Most local businesses fail the readable test before anything else. The homepage says everything and answers nothing, the services page lists nine things in nine sentences, prices are absent, and there is no structured data at all. There is nothing on the site an engine could quote even if it wanted to.

• Readable: one page per service, plain answers to real questions, prices stated as ranges in text, LocalBusiness, Service and FAQ structured data, fast mobile pages.
• Verified: the same name, address, phone, hours and services on Google Business Profile, Bing Places, Apple Maps and the directories AI cites for your trade.
• Trusted: reviews that keep arriving and can be read, real detail about the work, and mentions on the local sources an engine already pulls from.

That is the entire method, repeated every month and measured against the same 10 questions. There is no trick underneath it and no leverage that expires. The businesses that get named are the ones a machine can describe correctly without guessing.

Local service businesses where a buyer asks a question before they pick a name, and where one new customer is worth enough that a handful of them pays for the year. That is most of the trades and most of the clinics.

• Med spas and aesthetics clinics: our deepest market, with published first-party data city by city.
• Every industry we work in: the buyer question we target in each, and what one customer is worth.
• Not a fit: national ecommerce, and any city where a hospital system or national chain already owns the answer. We check first and decline when we cannot win.