Answer engine optimization is still a website problem.

What AEO means without the hype

Answer engine optimization is the work of making a page eligible to be quoted. Google’s May 2026 guide is blunt: generative AI features on Search are rooted in core ranking and quality systems. There is no special AEO schema, and llms.txt is not a ranking lever for Google.

A useful split: SEO asks whether the URL can be indexed and matched. AEO asks whether a model can lift one or two sentences without the surrounding pitch. GEO is the same job under another name, plus brand mentions on sources models already read.

What a production AEO audit looks at

Can a non-Googlebot crawler read the answer in the first HTML response, not only after client-side render? Is there a self-contained definition in the first 40–60 words of the relevant section? Is the organization named the same way on every page? Does structured data match the visible copy?

Citation trackers such as Profound, Peec, Otterly, and Scrunch tell you whether a prompt mentioned the brand yesterday. They do not tell you the page is blocked, thin, or only rendered in JavaScript. Siteory’s AI search audit is the site-side check. Pair them if you already buy prompt tracking. Do not buy tracking first if the HTML is empty.

Tactics Google told you to ignore

Do not chunk every page into 150-word cards, publish a second URL “for ChatGPT,” or add FAQPage markup after May 2026 to chase a rich result. Do not block GPTBot while hoping to appear in ChatGPT. Write the answer once, in crawlable HTML, and keep the date honest when you change the substance.