Answer engine optimization (AEO) is the work of writing and structuring your pages so an AI answer engine can lift a clear answer from them and name your brand as the source. An answer engine is a tool like ChatGPT, Google AI Overviews or Perplexity that reads web pages and writes one answer, and AEO is the content and citation part of the larger job called GEO.
AEO, GEO, AI SEO and LLMO all point at the same field of work, and Google's own guidance says its AI search needs no special optimization beyond good SEO. The names differ in how much of the job a vendor sells, so the steps in a proposal tell you what you buy. My view: the cheapest AEO edit is one sentence, the direct answer at the top of the section that targets a buyer question, and most pages lose at that line.
No engine publishes how it scores passages, so everything here comes from public tests and from our own tracking of more than 15,000 prompts, and the scoring rules can change with the next model.
This guide follows one buyer question from the search to the cited answer, then shows how to measure the result and where to start. In this guide I cover:
What is AEO?
How answer engine optimization works
How a brand gets cited
AEO compared with GEO and SEO
How AEO is measured
What the research says about AEO
Where to start with AEO
What is AEO?
AEO is the part of AI search work that happens on your pages: each section opens with a direct answer to one buyer question, with a number or a named source inside it. ChatGPT, Google AI Overviews, Perplexity, Gemini and Copilot all work as answer engines, writing one answer with a few sources under it.
An answer gives a brand two things to win:
A mention: the answer names your brand in its text.
A citation: the answer lists one of your pages as a source.
AEO works mainly on the citation, because a cited page is the one the engine took its words from.
Vendors use much the same words. HubSpot's guide to AEO calls it "being the source those engines trust enough to name and cite", and the tracking vendor Profound describes structuring content so AI tools "can easily understand, trust, and cite it as direct answers".
Google's documentation settles the naming question in one paragraph:
Google treats AEO and GEO as two names for one job and calls both SEO. Its page on AI features adds that "there are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary." On Google, AEO work and SEO work are the same work.
Sidenote: AEO is spelled "answer engine optimisation" in the UK, and the term is older than AI answers. It grew up around Google's featured snippets, the short answers shown above the results.
Further reading: our guide to GEO covers the whole job that AEO is part of.
How answer engine optimization works
Answer engine optimization works by handing the engine a passage it can use as the answer, on a page its bots can reach. An engine moves through four stages:
Searching: it turns the question into several searches of its own.
Collecting: it gathers pages and cuts them into short passages.
Scoring: it rates every passage against the exact question and keeps the best.
Writing: it writes one answer from the best passages and credits some of them.
The scoring stage is the one to write for:
Because the engine scores passages, a single strong section can win the citation for the whole page.
Google calls the first stage a query fan-out: AI Overviews and AI Mode may issue "multiple related searches across subtopics and data sources" to build one response. Here are the searches ChatGPT ran for one real buyer question from our tracking:
One question became 3 searches. A page has to come back from one of them before it can be cited, and AirOps found that a third of the pages ChatGPT cited (32.9%) showed up only for one of these extra searches.
Those searches keep the buyer's own words, which decides the names you should use on a page:
We ask each buyer question every day. Over a month, one UK question about a "generative engine optimization agency" produced 33 searches across 19 answers, and every one used that exact term while only 2 added "AEO". So use every name your buyers use.
A public test measured the scoring stage, where most pages lose. Metehan Yesilyurt on the Peec AI blog ran the query "what are the best aeo tools" through twelve open rerankers, the models that score passages, and compared a product description with a one-sentence shortlist of tools. Ten of the twelve report their score in percent:
The shortlist scored 99.4% or more on all ten, and the product description scored under 1% on eight of them. One model, Ettin, scored the description 99.9% as well. Each model uses its own scale, so the useful comparison is the two passages inside one model.
Sidenote: these are public rerankers, and nothing says ChatGPT or Google runs any of them. Read the test as a pattern; each engine keeps its own scoring private.
How a brand gets cited
You get cited when one section of your page answers the buyer's exact question in its first two lines. The example below is our illustration of how the scoring works.
Take a finance manager who types: which accounting software works for a UK business with fewer than ten staff? Two pages in that search sell accounting software, and only one is built around her question:
The first page has a section headed with her exact question, and its opening sentence names the options, says who each one suits and gives a price band with a source, which is why the engine lifts it. The second calls itself "a modern accounting suite for growing businesses" and answers a question nobody asked.
A buyer asked ChatGPT for AEO providers in Lithuania, and the real answer shows what an engine picks:
ChatGPT named 8 companies, and its 10 sources were two agency lists, seven agency service pages and homepages, and this guide. It also warned that AEO, GEO, AI SEO and LLM SEO "are often used almost as synonyms", so the service name alone tells a buyer little.
New posts moved one client's numbers on two topics of buyer questions:
In month two of the work, when the new posts went live, Outcraft went from 0% to 23.8% of answers on its SaaS free-trial conversion questions and from 4.4% to 28.6% on mobile checkout abandonment. The full story is in our B2B SaaS case study.
Sidenote: that result is one client and two topics of buyer questions. It shows how fast a well-placed answer can move, and it is no forecast for another category.
AEO compared with GEO and SEO
GEO is the name we use for the whole job, and it runs as four steps:
Open the door so the AI companies' bots can reach and read your site.
Give the engines something worth citing, on your site and on the sites they trust. The writing on your own site is the AEO part.
Watch what gets cited, page by page and question by question. This citation count belongs to AEO too.
Count the recommendations, meaning how often the engines name your brand and what it brings in.
On those four steps, the AEO part looks like this:
An AEO contract usually covers the writing on your own site and the citation count. Bot access, the work on other sites and the count of recommendations sit outside it, and a page can carry the best answer on the internet and still never be read when the server turns the bot away. We run all four steps as one program.
SEO is a different job with a different result, and AEO and GEO differ in how much of the job they cover:
SEO
What you try to win: a place in Google's list of links.
What wins it: pages that match what people search for, and links from other sites.
How you know it worked: your position for a keyword and the clicks from Google.
What backfires: keyword stuffing, meaning a page padded with one phrase.
AEO
What you try to win: the passage the engine reads out, with your page in the source list.
What wins it: a section that answers the exact question in its first two lines, with a number or a source inside.
How you know it worked: citations of your pages and the share of answers that name you.
What backfires: keyword stuffing again, which scored below the untouched page in the original GEO paper.
GEO
What you try to win: the same passages, plus how often the engines recommend you by name.
What wins it: AEO's pages, a site the bots can read and mentions on other sites the engines trust.
How you know it worked: citations, answer share per engine and the leads that come from AI answers.
What backfires: good pages behind a blocked crawler.
SEO and AEO also differ in speed. Ethan Smith, CEO of the SEO agency Graphite, argues in his guide to AEO on Lenny's Newsletter that early-stage startups can show up in AI answers quickly, while SEO takes years.
Buyers search for the comparison itself:
In the US, "aeo vs seo" gets 1,900 Google searches a month and "aeo vs geo" 1,600, more than any other AEO term we checked, so the comparison is the question most buyers bring to a first call. As a name, "ai seo" leads with 6,600 a month in the US.
Further reading: our guide to AI SEO, our guide to LLM SEO and LLMO and what AIO means.
How AEO is measured
AEO is measured by two numbers:
Citations: how often an engine used one of your pages as a source in an answer, read per page and per question.
Answer share: the share of answers to your buyer questions that name your brand at all. Tracking tools call it visibility, and Peec AI's documentation defines it as "the percentage of AI responses that mention your brand".
For the accounting software company, citations show which of its pages the engines use, and answer share shows how often it is named across its buyer questions.
Being read and being cited are different:
AirOps found that ChatGPT left 85% of the 548,534 pages it retrieved uncited. Retrieval means the engine read your page. Citations show it used it.
A brand can also be named while its own site goes uncited:
In our 2026 AI search visibility study, 34 of the 87 brands AI engines named, 39%, never had their own website among the top 30 cited sources. They were named on the strength of other people's pages, and that gap is what AEO works on.
The engines also lean on different sources:
Across 109 business projects in our study of which sources AI engines cite, ChatGPT, Google AI Overviews and Perplexity shared a median of 2 of their top 10 cited sources. A brand can be cited often in ChatGPT and rarely in Google AI Overviews, so read each engine on its own.
Citations also gather on a few pages, as they do on our own site:
Five of our 53 cited pages carry 63% of our AI citations over 30 days, and the homepage alone carries 21%, which is why we read citations page by page.
Both numbers move from day to day, since the same question can get a different answer an hour later. Run the questions on a schedule and read a four-week trend per engine, the routine in our guide to prompt tracking.
What the research says about AEO
Evidence on the page wins more of an AI answer, and one controlled test measured how much. "GEO: Generative Engine Optimization", the KDD 2024 paper by Aggarwal et al. of IIT Delhi and Princeton that gave the field its name, rewrote the same pages in nine ways and measured how much of the answer each version won.
Two results from the paper matter most here:
Quotations from credible sources raised a page's share of the answer by 41%, statistics by about 30%, clearer writing by about 28% and citing sources by 27%. Keyword stuffing scored below the untouched page, and about 10% worse on Perplexity.
The gain went to the pages Google ranked lower. Citing sources raised the fifth-ranked page's share by 115%, while the first-ranked page lost 30% from the same change. Smaller sites have a way in here.
Sidenote: the paper ran on GPT-3.5 and a 10,000-question test set, and its authors expect the methods to change as the engines do. Read the figures as a direction for your own pages.
Our own studies are in our published research.
Where to start with AEO
Start with the door, then the measurement, then the writing. Measurement comes before writing because AEO is judged on citations, and a page rewritten before anyone counted its old citations can never be shown to have earned new ones.
Check that the bots can reach you. Read your robots.txt, the file that tells crawlers where to go, against our list of AI crawlers. OpenAI's crawler documentation lets you allow OAI-SearchBot for ChatGPT search and refuse GPTBot for training, and Perplexity's says PerplexityBot surfaces and links sites in its results. Then ask whoever runs your server whether a firewall blocks them.
Measure from day one. Pick ten questions your buyers ask, run them on each engine and record whether you are named and which page was credited. The free route is a logged-out ChatGPT window every week; a free AI visibility check with us runs them for you. A set that shows a trend starts at about fifty questions, chosen the way we describe in how we build a tracked prompt set.
Write the answers. Give each question its own section, answer it in the opening two lines and put a number, a quote or a source inside that answer, as the accounting page did. If writing is the bottleneck, our content service writes these pages.
We work in the same order. In our GEO service, month one is the audit and strategy, month two is execution and month three brings a before and after report.
I'd wait with AEO when:
Your buyers never ask an AI assistant before they buy, as with walk-in trade or emergency call-outs booked by phone.
You sell only through a marketplace, so the work sits on your listing and on pages the engines already read.
Nobody can publish or change a page for the next six months, so you would pay for measurement and watch a flat line.
Questions people ask about AEO
Is AEO the same as GEO?
They name the same field, and the difference is scope. A vendor selling AEO usually sells the writing and the citation tracking, while GEO should also cover crawler access, mentions on other sites and the count of recommendations. Ask which of the four steps a contract includes; all four are in our guide to GEO.
Do FAQ pages still work for AEO?
FAQ pages work for AEO when each answer stands alone and carries something checkable. Use the question a buyer would type as the heading, answer it in the opening two lines and put a number or a source inside the answer. Schema.org describes a FAQPage as a page "presenting one or more Frequently asked questions", and the same question-and-answer shape works inside an ordinary page, so a separate FAQ page is optional.
Does schema markup help with AEO?
Schema markup labels what is on a page, and on its own it does not earn citations. Ahrefs tracked 1,885 pages that added JSON-LD schema against 4,000 control pages: citations in Google AI Overviews fell 4.6%, a small but statistically significant drop, and the changes in AI Mode and ChatGPT were too small to tell apart from zero. Google's AI optimization guide says "structured data isn't required for generative AI search". Keep schema for SEO, and spend the AEO effort on the answers.
How do I know if AI engines are citing my website?
Ask them the same buyer questions on a schedule and record the answers. Type five of those questions into ChatGPT and Perplexity today in a logged-out session, so your chat history does not colour the result, and note the sources under each answer. For a trend you need the questions run daily across engines, which is what prompt tracking does.
What does AEO work cost?
Agenzy's GEO work, which covers everything the market calls AEO, starts from 5,000 EUR per month ex VAT, with 50 buyer questions tracked daily across ChatGPT, Gemini and Google AI Overviews. What moves the figure is how many markets and languages you sell in, how many buyer questions we track and how much content we publish. Published agency prices sit in our GEO agency cost guide.
Final thoughts
AEO is the work of giving an answer engine a passage it can lift and credit to you. It is the content and citation part of GEO, and it is won or lost in the first two lines of the section that targets a buyer question.
This week, pick the one page that should answer your buyers' most common question and rewrite its opening two lines as the answer, with a number or a source in it.
If you test this and the engines still skip your page, tell me on LinkedIn.
