What is AEO, or answer engine optimisation?

Answer engine optimisation, AEO for short, is the work of preparing your page so that a direct answer to a question is lifted from it, by a search engine or by an AI engine.

An answer engine is any engine that replies to a question with an answer rather than a list of links. So the question in AEO is not "what position is my page in?" but "did the sentence the asker reads come from my page?"

The same work, and the work next to it, goes by many names in Arabic and English. This is what each name means when you meet it:

• AEO, answer engine optimisation: the focus is the direct answer lifted from a page. In Arabic, «تحسين محركات الإجابة».
• GEO, optimising for AI answers: the focus is your page being cited inside an answer an AI engine writes in its own words. Some call it "AI SEO", in Arabic «سيو الذكاء الاصطناعي».
• SEO, search engine optimisation: the focus is your page's rank among the links on a results page. Many Arabic speakers write it «السيو».
• Answer engine, or Q&A engine: the engine that writes the answer, not the work you do on your page.

From the direct answers that came before generative AI, the best known being Google's featured snippet: a box that puts the text quoted from a page above its title and link, the reverse of a normal result.

Google says its systems decide whether a page would make a good featured snippet, and that a site owner cannot mark a page for one. AEO inherited that: you prepare the page, and the engine chooses.

Then came answers that AI writes in its own words, and the name was used for them too. The name GEO comes from a research paper, "GEO: Generative Engine Optimization", first posted on 16 November 2023 and accepted at KDD 2024, which describes engines that gather and summarise information with generative models to answer a query.

AEO is usually said of the direct answer lifted from a page, and GEO of the answer an AI engine writes in its own words while citing its sources. The goal behind both names is the same: your page as the source.

Google's guide to its generative AI features says optimising for them is optimising for the search experience, and so still SEO, and mentions AEO and GEO as terms you may see for work focused on AI search. The three names are compared, with Arabic examples, in SEO vs GEO vs AEO.

An illustrative example: falak-dental.sa is a fictional dental clinic in Jeddah, and a patient types «كم سعر تقويم الأسنان الشفاف في جدة؟» ("how much are clear aligners in Jeddah?"). The engine looks for a sentence that answers both the price and the city.

The clinic's orthodontics page opens with its founding story and team photos, and the price sits in a table inside a closed tab at the bottom. The engine finds no sentence to lift, so it lifts one from a doctors' directory that gives the price in its first line.

After the edit, the section's heading is the patient's own question, and under it one sentence gives the price as a number, the district, and what moves the price up or down, then the detail. That sentence is the one fit to be lifted.

It changes the page's order, not its subject: the question becomes a heading, the answer comes in the first sentence under it, and the detail follows, so the engine finds a whole sentence to lift without reading the whole page.

Four changes to start with:

• Write your customers' questions as headings, in their words, and answer each in one or two sentences that make sense on their own.
• Put question and answer pairs in an FAQ section: Microsoft says assistants can lift these pairs word for word into their answers.
• Do not hide a key answer in a tab, a collapsing menu or an image: Microsoft says AI systems may not render hidden content, and that pulling text out of images reduces its accuracy.
• Put the number, the city and the condition in the same sentence: a sentence that needs the paragraph before it for context cannot be lifted.