One evening, I was scrolling LinkedIn after a long day of client calls. Three posts, one after another, told me everything about where our industry stands right now.
The first post was celebrating. Google had just published its guide on appearing in AI features in Search, and one line from it was doing the rounds: optimizing for AI search is still SEO. The post read like a victory lap. Nothing changes. Carry on as before.
The second post declared SEO dead. Same week, same Google line, opposite reading. It ended with a pitch for an AEO package, because apparently everything I have done for over ten years stopped working on a Tuesday.
The third was a comment war under someone else’s post. Both camps, shouting past each other, each quoting the same document.
I have spent more than ten years in SEO. I started when ranking meant one simple thing: your page shows up when someone types a phrase into Google. Today, at Adtric, I spend my days helping brands show up in a world where the answer often appears before any website link does. I read Google’s guide carefully, more than once. I have also watched what is happening inside our own client accounts this year.
Here is where I stand: both posts were wrong. And they were wrong in the same way. They read one sentence and skipped everything around it.
“Still SEO” describes the entry ticket. It says nothing about the game.
What Google actually said, in plain words
Strip away the technical language and Google’s guide says something simple. When its AI answers a question, whether in the answer box you now see at the top of many searches (called an AI Overview) or in the chat-style AI Mode, it does not invent the answer from nowhere. It first searches its own regular index, the same one behind the classic blue links. It picks pages it already trusts, and it writes its answer from them, with links back to the sources.
Two plain conclusions follow. First, if your pages are not indexed and not ranking anywhere for a topic, the machine cannot find you, so it cannot mention you. That is why SEO is the foundation. Second, Google says you need no special code or hidden trick to appear in these answers. If your page can appear in normal search, it can appear here.
One more detail is worth knowing, because it matters later. When you ask one question, the system quietly runs several related searches behind the scenes. Ask it for the best 3 BHK under ₹2 crore in Noida Extension, ready to move, and it may separately look up projects, prices, possession status and locality reviews, then combine what it finds into one answer. One question from the buyer. Many searches underneath.
The part that has not changed
My first answer is to the SEO-is-dead post, and it is not just my opinion. The independent evidence published this year points in one direction. A widely shared review of 54 studies on AI citations found that the two strongest factors behind getting mentioned in AI answers are the most old-fashioned things imaginable: whether your page is easily accessible, and how well it ranks. An Ahrefs analysis in the same review found that 38 percent of AI Overview citations come from the top ten classic results. And a survey of 500 marketers by Neil Patel’s agency found that original research, real first-hand material, is the content most likely to get picked up.
My reading of all of it fits in one line: do SEO well, and you have already done most of the work of AEO. The machine reads the same web we have been optimizing for a decade. Anyone telling you to throw that away is selling you the upper floors of a building with no foundation under it.
The part I am watching change in our own accounts
Now my answer to the nothing-changes post, and this one comes from a dashboard, not a debate.
On one education client account in Delhi NCR, our rankings held steady through the past year. Positions barely moved. Impressions kept climbing. And clicks kept falling, quarter after quarter, starting around the time AI answers began appearing on more of their queries. We were not doing worse SEO. The click was simply being answered before it could happen.
The published data tells me our account is not an exception. A June study of clickstream data found that around 68 percent of Google searches now end without a click to any website. Ahrefs measured that the top-ranking page loses more than half its expected clicks when an AI Overview sits above it. Rank number one, get half the reward. That is the new arithmetic.
And there is a second change, the one I believe is bigger. How people search has changed in front of us. Five years ago, a buyer typed “flats in Noida Extension”. Today that same buyer types, or speaks, “best 3 BHK under ₹2 crore in Noida Extension for a family, ready to move, near a metro station”. Longer. More specific. More local. More personal. A parent no longer searches “schools in Greater Noida West”; she asks which school has transport for her sector, activity-based learning and open registrations. In India, where roughly three-quarters of web traffic is on mobile, much of this now happens by voice and in mixed Hindi-English phrasing.
AI answers push this even further, because they are personalized. Two people can ask the same question and get different answers, shaped by their location and history. Researchers who study this now advise covering a topic across many intents and buyer situations, rather than chasing one phrase. This is my view as well: the fixed keyword list, checked once a month, describes less and less of what real buyers actually do. Optimization has to follow the buyer, and the buyer has moved.
The uncomfortable part: smaller brands have it harder
Here is the opinion that earns me the most pushback, so let me say it plainly. AEO is real, it is already getting popular, and I believe it is the future of how search works. But the hype around it hides a hard truth: right now, it is more difficult for a lesser-known brand to become visible in AI answers than it ever was in classic search results.
In classic search, a smaller brand with sharp content and patience could out-rank a bigger name on merit. I have done it for clients many times. In AI answers, so far, the field tilts toward brands the machine already recognizes and trusts. One analysis of 177 brands found 90 percent had zero mentions across AI search. Another found that only 2 percent of cited pages appear across all three major assistants, while 91 percent appear in just one. The machine cites who it recognizes. And recognition is branding.
So my advice to founders has changed this year. Your press coverage, your reviews, your presence on the platforms where your category gets discussed: these now feed your search visibility directly. That was always half-true in SEO. It is very true in AEO. A brand nobody talks about is a brand the machines stay quiet about too.
How we scope this at Adtric now
When a CMO asks me whether keywords still matter, my answer is yes, and here is the frame I use in every scoping conversation. Keywords serve the classic results, which have not disappeared and still carry most of the clicks that do happen. Clusters, priority pages for your services, products and locations, and buyer prompts serve the AI answers. We scope both, because the buyer lives in both.
In practice, that looks like this. Keyword and ranking tracking continues in tools like Semrush and Ahrefs, the same discipline as before. Alongside it, we map every priority service, project or location to a cluster: one main page, supporting content around it, and clean structured data underneath, which is simply a machine-readable label stating plainly who you are, what you sell and where you operate. And every month we check the questions real buyers ask across ChatGPT, Gemini and Google’s AI surfaces, and record whether the brand shows up in the answers. Rankings and prompts sit side by side in the same report. Neither replaces the other.
Two honest limits, because I would rather lose a little shine than your trust. Some fashionable tactics have little evidence behind them; a much-discussed file called llms.txt scored near the bottom of that 54-study review, so we tell clients that plainly and skip it. And this field is young. What we measure today, we will measure better next quarter, and some of it will change. Anyone promising you certainty in AEO right now is selling the hype I opened this piece with.
So, was Google right? Yes. At the base, it is still SEO. Were the celebrating posts right? No. The game above the base has changed: the clicks, the queries, the surfaces, the growing weight of brand. Were the SEO-is-dead posts right? Also no. They are selling floors without a foundation.
SEO is the foundation of AEO. AEO is more than SEO. Both sentences are true at the same time, and an agency that cannot hold both together will either mis-scope your work or oversell it.
That is the standard we hold ourselves to at Adtric: classical craft underneath, new measurement on top, and a straight answer about what the evidence does and does not support. Performance, simplified.
Frequently asked questions
SEO is the foundation and AEO builds on top of it. The strongest known factors behind AI citations are classic SEO basics like accessibility and rank, but AEO adds new work: cluster coverage, monthly prompt tracking across AI assistants, and brand signals that machines can recognize.
Yes. Classic search results have not disappeared, and they still carry most of the clicks that do happen. What has changed is that keywords alone no longer define the scope of the work; clusters, priority pages and buyer prompts now sit alongside them.
No. Around 38 percent of AI Overview citations come from the top ten classic results, so ranking helps a lot. But only 2 percent of cited pages appear across all three major AI assistants, so each one needs its own visibility check.
By tracking the actual questions buyers ask, every month, across ChatGPT, Gemini, AI Overviews and AI Mode, and recording whether the brand appears in the answers. This sits alongside classic rank tracking in tools like Semrush and Ahrefs, not in place of it.
Right now, yes. Studies show most brands get zero AI mentions, and machines lean toward names they already recognize. Building brand visibility through press, reviews and community presence now directly supports search visibility.