Guide · AEO and SEO
They share a technical foundation and almost nothing else. Here is what changes, concretely, when the prize is a citation rather than a ranking.
Updated 2026-08-29
Search engine optimisation earns you a position on a results page. Answer engine optimisation earns you a citation inside a generated answer. Those sound like the same job described twice. They are not, and the differences are specific enough to change what you build.
The evidence that they are separate channels rather than two views of one: studies report that roughly 28% of the pages ChatGPT cites most often have zero organic visibility in Google. Ranking well does not automatically get you cited, and being invisible in Google does not exclude you.
Google ranks pages. Answer engines retrieve passages, typically a heading and the text beneath it, then assemble an answer from several of them, often from different sites.
The practical consequence: a page that only makes sense read top to bottom performs badly, because the engine never reads it top to bottom. Every section has to stand alone.
The short versionWrite sections that survive being read in isolation, because they will be.
Googlebot renders JavaScript. GPTBot, OAI-SearchBot, ClaudeBot, Claude-SearchBot and PerplexityBot do not.
A client-side rendered React or Vue application that ranks perfectly well in Google can be completely invisible to every AI assistant, and nothing in your analytics will tell you. This is an engineering problem, not a content one, and it is the single most common cause of an otherwise good site being absent from AI answers. See the full crawler reference.
The short versionServer-render, statically generate or prerender. There is no content fix for this.
The foundational peer-reviewed work in this field (the GEO paper presented at KDD 2024 by researchers from Princeton and Georgia Tech) tested which page changes lift visibility in generative engines.
Three tactics each produced gains of roughly 30 to 40%: adding citations to sources, adding quotations from credible sources, and adding statistics. Keyword stuffing performed at or below baseline.
The short versionCite, quote and quantify. The things that feel like academic habits are the ranking factors here.
The same research found the gains were largest for lower-ranked sites. Generative engines disproportionately reward smaller players who structure content well.
In classic SEO a larger competitor's decade of backlinks is a wall you cannot climb quickly. In AI search, a precise, well-evidenced answer from a small site can be cited ahead of a vague page from a large one.
The short versionThis is the rare channel where structure beats budget.
An SEO heading targets a phrase. An AEO heading matches the shape of what a person actually asked an assistant, because the retrieval step is comparing the question to your headings.
"Pricing" is an SEO heading. "How much does an AI answering service cost?" is an AEO heading. The second one gets retrieved.
The short versionPhrase headings as the question a buyer would type into an assistant.
Google removed FAQ rich results from Search in May 2026, so FAQ markup no longer earns the expandable snippet it once did. A lot of agencies quietly stopped implementing it.
That is the wrong conclusion. In an AEO context, structured data's job is telling a machine unambiguously which text is a question and which is its answer, and that is precisely the format answer engines retrieve.
The short versionKeep implementing schema. The audience changed; the value did not.
AI Overviews and AI Mode are controlled by ordinary Googlebot crawling. There is no separate bot to allow, and Google-Extended governs only Gemini training.
So classic SEO is not a competing investment. It is the entry ticket to Google's AI surfaces, and Google remains the largest of them.
The short versionAEO does not replace SEO. It runs on top of it.
Rank tracking does not answer the question "are we in the answer?". Measuring AEO means re-asking a fixed set of buyer questions across ChatGPT, Perplexity, Claude and AI Overviews on a schedule, and recording where you were and were not cited.
It is a noisier signal than rank tracking and it moves for reasons outside anyone's control, which is exactly why an honest agency reports it against a dated baseline rather than promising a number.
The short versionBaseline your target questions before changing anything, or you cannot prove the work did.
Frequently asked questions
