A SaaS client of ours, mid-sized, decent budget, came to us in 2025 wanting to go all in on AEO. Not “add it to the plan” — replace the plan with it. New tools bought before we’d even audited what they had. Their entire link building process redirected to chase citations instead of the keywords that were actually driving revenue. A push to pause existing content work and pour that budget into new content built specifically to get quoted by ChatGPT.
Within a few months the problems showed up. Growth that should have compounded stalled instead. The site had real structural issues sitting untouched — thin pages, cannibalised terms, an internal linking mess — because every hour and every rand had gone toward “the AI thing.” Had they spent that same budget fixing what already existed, we’d have been talking about a genuine turnaround. Instead we spent the next few months doing the SEO work that should have happened first, before circling back to citations at all.
That client is the reason I wanted to write this. Not because AEO isn’t real. It is. But because the conversation around it has turned into term soup – AEO, GEO, AIO, all used interchangeably, all wrapped in urgency that doesn’t hold up under scrutiny. And somewhere in that noise, people are making the same mistake that client made: treating answer engine optimization as a replacement for SEO instead of what it actually is, which is an extension of it.
What Answer Engine Optimization Actually Is
Answer engines – ChatGPT, Perplexity, Gemini, Google’s AI Overviews – don’t hand back a list of links. They synthesise an answer and, sometimes, cite where it came from. Answer engine optimization is the practice of structuring your content so those systems can find it, understand it, and cite it as that answer.
That’s it. One paragraph.
Everything past this point is about what that actually requires, and what it doesn’t.
The Argument Every Vendor Guide Is Getting Wrong
Search “answer engine optimization” and the page that ranks first organically is a Reddit thread. After that: a monitoring platform’s resource article, a SaaS product page, a Forbes contributor column, an enterprise content platform’s guide. Every one of them is selling something — a citation-tracking dashboard, a “monitoring gap” you need their tool to close, a 90-day rollout plan that happens to require their subscription.
They also share the same argument, stated or implied: AEO is a new discipline. New team. New tools. New budget line. Rip out what you were doing and build a parallel operation to chase citations.
I’ve watched what happens when a client believes that. It’s the story above. It’s not hypothetical — it’s the direct, measurable cost of treating AEO as a category instead of a layer.
Here’s the plainer version of what I tell clients in that meeting, every time: the foundation of SEO is helping someone get found, and that doesn’t change based on which engine is doing the finding. There’s a structural path — site architecture, content quality, technical access, authority — that gets your content found regardless of surface. What shifts with AEO isn’t that foundation. It’s a specific set of tactics layered on top of it: how you structure an article, how clearly you state things, how you mix content types on a page. That’s optimisation within SEO, not instead of it.
Where AEO and SEO Are Actually the Same
The goal hasn’t changed
If you’re not helping somebody solve a problem, you’re not doing any form of optimisation. Doesn’t matter whether that’s playing out in a Google results page, a ChatGPT thread, or Gemini answering inside Google itself. SEO was always supposed to do two things at once – get the site found, and get the person searching an actual answer to their problem. Rank a page that solves nothing and you’ll get the click and lose the trust. That was true before AI search existed. Still true now.
Link building, authority, and structure still do the job
Entity authority, brand mentions, PR, topical clusters that reinforce each other — none of that disappears because an AI system is doing the reading instead of a person scanning a results page. If anything, it matters more, because answer engines are trying to establish which sources to trust before they’ll cite one.
Where AEO Actually Requires Something Different
This is the part most explainers skip past on their way to the FAQ schema pitch. There are real differences — they’re just narrower than the hype suggests.
Content structure
Answer engines reward clear, concise, extractable statements. Chunking your content so a single paragraph can stand alone as an answer. Mixing in more content types on one page — tables, comparisons, data — instead of one long wall of prose. None of this replaces good writing. It’s a formatting discipline layered on top of it.
Technical access
This is genuinely different from classic SEO, and it’s mostly about how these systems read a page rather than what the page says. Markdown formats and llms.txt files (still debated, still evolving) exist because AI crawlers can’t parse JavaScript the way Google’s crawler can. If your content is rendered client-side with nothing readable underneath, an AI system may simply never see it — regardless of how good the content is.
Measurement – and this is where AEO earns its own toolkit
Here’s where I’ll actually agree that something separate is warranted: tracking. Ranking and being cited are not the same metric, and conflating them will waste your time. You can rank #1 in Google and never get cited by an LLM. You can be cited three times in a week and get zero referral traffic from it. Makes sense doesn’t it, once you say it out loud? Traditional rank tracking doesn’t capture citation behaviour, so some separate visibility into what’s being cited is worth having.
Where they converge again is the metric that actually pays the bills: traffic and conversions. As of May 2026, Google Analytics 4 added a native “AI Assistant” channel, separate from Organic Search, that automatically tags sessions referred from tools like ChatGPT and Gemini. That’s now sitting inside a platform most of you already have open. Search Console and Microsoft Clarity are moving the same direction. You don’t need new infrastructure to see this — you need to look at the infrastructure you already have.
Do You Need to Buy a Monitoring Tool?
Probably not yet.
And I say that as someone who’d benefit from telling you otherwise.
Citation tracking right now is genuinely difficult to do well. The queries that trigger a citation shift constantly. The models get updated without notice. What counts as “being cited” isn’t standardised — is it a link, a brand mention, an attribution with no click at all? Different tools answer that differently, which means you’re paying for a number whose definition changes under you.
Before spending on a new platform, look at what you already have. If you’re running Ahrefs, Semrush, or Search Console, a decent chunk of the picture is already sitting in a dashboard you’re paying for. Show a client how fast the actual search results shift week to week, and the cost-to-signal ratio on a lot of these citation tools starts to look thin. In some cases the tools genuinely help. But weigh the cost of the subscription against what you actually learn from it — and remember the two metrics that matter most, traffic and conversion, are ones you can already see.
Why the “Search Is Dying” Predictions Keep Missing
In February 2024, Gartner predicted traditional search engine volume would drop 25% by 2026 because of AI chatbots and other agents. That’s now.
Google still holds roughly 90% of the search market. The prediction, as stated, didn’t happen.
The mistake wasn’t the direction — it’s the blanket framing. Traffic to websites genuinely has dropped for a category of search: informational queries. If someone’s question gets fully answered inside ChatGPT or an AI Overview, there was never much reason for them to click through in the first place. But commercial-intent search — the searches closer to a purchase decision — hasn’t moved the same way, and branded search is holding roughly steady. People still bounce between platforms doing this: Google to YouTube, ChatGPT to Google to TikTok, circling until they’ve narrowed things down, and only then landing on a site.
A single blanket statistic assumes everyone adopts a new tool and changes their entire behaviour overnight. That’s not how people search. It’s also worth naming plainly: AI-referred traffic tends to convert well. Adobe’s own analytics found generative AI referral traffic converting 31% higher than other channels over the 2025 holiday period, and Ahrefs’ first-party data found AI search traffic – under 1% of total visits – accounted for over 12% of signups. The exact multiplier varies by industry, and I wouldn’t hang a strategy on one number from one source. But directionally, less traffic isn’t automatically less business.
If your goal is raw traffic, a 25%-drop headline is genuinely alarming. If your goal is conversions, the picture looks a lot calmer — and a lot more workable.
The Moment the Panic Actually Stops
I’ve watched this shift happen in real meetings. It’s usually when someone finally puts traditional search traffic and AI search traffic side by side over time, instead of reacting to a single scary number. When traditional search hasn’t collapsed — and in a lot of cases has kept growing — and AI search traffic is rising alongside it rather than instead of it, the hype drains out of the room fast. It stops looking like an existential threat and starts looking like what it is: another channel to bring into the mix, not a reason to tear up the plan.
The Framework That Actually Works: Search Everywhere Optimization
If I had ten minutes with someone completely overwhelmed by AEO, GEO, and AIO, here’s what I’d tell them.
Fix your SEO fundamentals first. That’s where the biggest gains still live — for either kind of search. Then go back and optimise what you already have instead of rushing to build new. Refreshing existing content does double duty: it lifts your standing in traditional search and gives you a better shot at being cited in AI search, because both reward the same signal — content that’s current, specific, and genuinely useful.
The acronym soup – AEO, GEO, AIO – is mostly an artefact of an industry looking for a new thing to sell. What’s actually happening is that SEO is expanding into something closer to Search Everywhere Optimization: the same fundamentals, applied across more surfaces, with a layer of tactics specific to how AI systems read and cite content.
That SaaS client eventually got there too. Not by buying more tools.
By going back and fixing what they already had.
Where to from here? Probably not a new tool. Probably your existing content, looked at properly for the first time in a while.
FAQ
Is AEO the same as GEO?
No, though they’re closely related and often used interchangeably. GEO (Generative Engine Optimization) is the broader practice of getting your content synthesised and referenced inside AI-generated answers across platforms. AEO is usually used more narrowly, focused on being selected as the direct cited answer. In practice, most of the tactics overlap — this is more a labelling problem than a strategic one.
Do I need a separate team for AEO?
Not a separate team — an expanded one. The core fundamentals of SEO haven’t changed enough to justify a parallel department. What helps is adding capability: someone thinking about PR and citation-worthy distribution, someone who understands the technical access requirements (markdown, llms.txt), and a measurement layer that separates citations from rankings. That’s an extension of an SEO function, not a replacement for it.
What tools actually track AI citations?
Most citation-tracking tools are still immature, and what counts as a “citation” varies between them — a link, a brand mention, an attribution with no click. Before buying one, check what you’re already getting from Ahrefs, Semrush, or Search Console. Weigh the tool’s cost against the two metrics that actually matter: traffic and conversions, both of which are visible without a new subscription.
Has Google AI Overviews actually killed organic traffic?
It’s dropped traffic for a specific category: informational queries that AI Overviews answer directly on the page. Commercial-intent search and branded search haven’t moved the same way. Treating this as a blanket collapse in search traffic overstates what the data actually shows.
Does losing clicks to AI Overviews mean losing business?
Not necessarily. Traffic from AI-referred sources has shown strong conversion performance in independent studies from Adobe and Ahrefs, even where the total volume is small. Fewer clicks isn’t automatically fewer customers — it depends on whether you’re optimising for traffic or for the outcome traffic is supposed to produce.
What’s the first thing I should fix before worrying about AEO?
Your existing SEO fundamentals — site structure, thin or cannibalised content, technical access issues. Optimising what you already have gets you gains in both traditional and AI search at once. Chasing citations before that foundation is solid is the exact mistake that stalls growth instead of accelerating it.
No comment yet, add your voice below!