Journal How AI engines choose who to name.
No engine publishes exactly how it chooses. But every answer is built from what the engine has read, so three inputs are yours to change: whether your own pages answer the question plainly, whether your facts agree everywhere they appear, and whether other sites the engine trusts mention you.
The guide
Where answers come from
An engine draws on two kinds of reading. Some of what it knows comes from the text it was trained on. For many questions it also searches the web at the moment you ask, reads a few pages and writes an answer from them, often linking those pages as sources. Perplexity, ChatGPT search and Google AI Overviews all show their sources this way.
Either way, the engine can only say what something it has read supports. That is the opening for AEO.
1. Your own pages answer the question
When an engine searches for “best payroll tool for a 50-person team”, it looks for pages that answer that question directly. A page whose heading is the question and whose first lines are the answer is easy to quote. A page that opens with a slogan and puts the facts three clicks away is not.
State what a buyer needs to shortlist you: what you cost, who you are for, how you compare, how long setup takes. Mark those facts up with schema so they are machine-readable as well as human-readable.
2. Your facts agree everywhere
Engines read your site, your directory profiles, your LinkedIn page and articles that mention you. When your price, your category or your customer size reads differently on each of them, an answer has no single fact to repeat. Keep one version of each fact and use it everywhere.
3. Other sites mention you
Answers to shortlist questions lean on pages that compare options: comparison articles, “best of” lists, directories, review sites and community threads in your category. A brand that appears on none of them gives the engine little to cite beyond its own claims. Earning a few honest mentions on the pages the engines already read is part of the work.
Why the answer changes from one ask to the next
Answers are written fresh each time. The wording of the question, the person asking, their location and the date can all change who is named. So one screenshot proves very little, good or bad. Measure a fixed set of questions across the same engines every month, and watch the trend.
A ten-minute check you can run today
Pick five questions your buyers ask. Ask each one in ChatGPT, Perplexity, Gemini and Google. Note who is named, and whether you are. Then open the page on your site that should answer each question and read its first two sentences: do they answer it? That gap is where AEO starts. Our sample report shows what the full version looks like.
By Rouge Flamingo · Published · 3 min read
Questions
Does schema markup get us cited?
Schema helps engines read your facts accurately. On its own it does not get you named; the content and the mentions still have to be there.
How often should we check?
Monthly, with the same set of questions, so one month compares with the next. Change the set only at the start of a quarter.
Which engines should we track?
The ones your buyers use. We track four by default: ChatGPT, Perplexity, Gemini and Google AI Overviews.
Read next
What is AEO?
AEO, Answer Engine Optimisation, is the work of making your brand the one AI assistants name when buyers ask about your category. It is also called GEO, Generative Engine Optimisation. SEO earns your page a place in a list of links; AEO earns your brand a mention inside the answer itself.
AEO vs SEO
SEO gets your page ranked in a list of links. AEO gets your brand named inside the answer an AI assistant writes. They share a foundation, pages that are easy to crawl and read, but AEO is judged by whether the answer names you, not by where a page ranks.