Buyers increasingly read an answer instead of a list of links. Here is what actually determines whether that answer mentions you.
For twenty years the goal of search marketing was a position on a page of links. That goal is quietly being replaced. A growing share of commercial questions are now answered directly — by an AI overview, by an assistant, by a chat interface that never shows a list at all. The click you were competing for often does not happen.
This is not the end of search marketing. It is a change in what winning looks like. You are no longer competing for a rank; you are competing to be the source an answer is assembled from. The practice has a name now — answer engine optimisation, or generative engine optimisation — and while the label is new, most of what works is recognisable to anyone who has done technical and editorial SEO properly.
What actually gets quoted
Answer engines favour content they can parse without ambiguity. A clear question addressed in a clear heading, followed by a direct answer in the first sentence or two, before the context and the caveats. Burying the answer three paragraphs into a narrative is how good content gets passed over for worse content that is easier to extract.
There is measurement behind this rather than just intuition. AirOps analysed AI Overview citations and found 55% came from the first 30% of the cited page — the extractor reads the opening of a section to decide whether it answers the query, and rarely reads on to find out. A conclusion held back for the end is a conclusion nothing will ever quote.
Structure matters for the same reason. Proper heading hierarchy, tables where you are comparing things, structured data that states what a page is about rather than leaving it to be inferred, and internal links that make the relationships between your pages explicit. None of this is new advice. What is new is that the penalty for ignoring it is now invisibility rather than a slightly lower rank.
Specific numbers beat confident prose
A joint study from Princeton, Georgia Tech and IIT Delhi tested which content changes actually moved visibility inside generative answers. The techniques that produced the largest gains — up to 40% more visibility — were adding statistics, citing sources and including direct quotations. Rewriting for keyword density did close to nothing.
That finding is worth taking literally. If your page says "site speed affects conversion", it is one of ten thousand pages saying that. If it says "we reduced time to interactive from 4.1s to 1.8s and organic enquiry rate rose from 1.9% to 2.6%", it is a page with something quotable in it. The model has a reason to reach for you rather than anyone else.
Freshness compounds this. AirOps found 83% of AI citations on commercial and evaluation-stage queries came from pages updated within the previous twelve months, and more than 60% from pages updated within six. An accurate page from 2022 is competing against an equally accurate page from this quarter, and losing.
Entity, not just keyword
Answer engines reason about entities — companies, products, people, places — and what is reliably said about them across the web. If your business is only described on your own website, you are a claim. If it is described consistently across directories, review platforms, industry press and your own pages, you are a fact.
Practically: make sure your name, category, locations and specialisms are stated the same way everywhere, and invest in being referenced by sources that are themselves trusted. This is what digital PR is for, and it now pays back twice — once as a link, and once as corroboration that you are who you say you are.
This is also why reputation work has stopped being a defensive activity. Review platforms are heavily represented in the sources these systems draw on, so what is written about you in public is now an input to how you are described in private conversations you never see.
What to measure instead of rankings
Rank tracking is becoming a poor proxy. Better questions: are we mentioned in the answer for the queries that precede a purchase in our market? Is branded search growing? Is qualified traffic holding even as total impressions fall?
That last one matters more than it first appears. Many businesses will see impressions drop while revenue holds or rises, because the traffic that disappears is the traffic that was never going to buy — the definitional query, the idle question, the student. Judging the channel on volume alone will lead you to cut something that is working.
Build the habit of asking the assistants directly. Take the twenty questions that precede a purchase in your category, ask each one monthly, and record whether you appear and how you are described. It is manual and it is the only measurement that reflects what a buyer actually sees.
The things that get in the way
Content that cannot be reached cannot be cited. If a page renders only after JavaScript executes, or sits behind a form, or is blocked in robots.txt to the crawlers these systems use, none of the advice above applies to it. Check what your robots file actually permits before assuming the content is the problem.
The other common blocker is a site that says everything and asserts nothing. Pages written to avoid committing to a position — hedged, general, applicable to anyone — are exactly the pages a model has no reason to quote, because quoting them adds nothing to its answer.
Where to start
Take the ten questions your best customers ask before they buy. Check what an assistant currently answers for each, and note who it cites. You now have a content brief grounded in real commercial intent rather than a keyword tool, and a list of the competitors and publications you need to be mentioned alongside.
Then fix the two cheapest things first: put the answer at the top of every page that has one, and add the numbers you already know to the pages that currently gesture at them. Both are edits to existing content rather than new production, and both are the changes the research says move the needle furthest.
Common questions
- What is answer engine optimisation (AEO)?
- Answer engine optimisation is the practice of making content likely to be used and cited when an AI system answers a question directly, rather than optimising purely for a position in a list of links. It overlaps heavily with good SEO — crawlability, clear structure, authority — but places more weight on answering the question immediately, supporting claims with specific sourced figures, and being described consistently across the web.
- Does AEO replace SEO?
- No. Answer engines draw from the same index and rely on the same signals — a page has to be crawlable, indexable and credible before it can be cited. AEO is best understood as an extra set of requirements layered on top of technical and editorial SEO, not a replacement for it. A site with unresolved crawl or speed problems will not be cited regardless of how its content is written.
- How do you measure whether AI answers mention your brand?
- Ask the assistants the questions your buyers ask, on a regular schedule, and record whether you appear and how you are described. Supplement that with branded search volume, direct traffic and the ratio of qualified enquiries to total sessions. Conventional rank tracking will not capture it, because the citation often produces no click at all.
- How long does it take to appear in AI answers?
- Faster than traditional rankings on pages that already have authority — restructuring an existing well-linked page to answer a question directly can show up within weeks. Establishing a new site or a new entity takes considerably longer, because the corroborating references across other sources have to accumulate first.