More insights.
Subscribe to our newsletter.
Deep dives into design thinking, creative process, and the intersection of business and aesthetics.
Short answer: SEO gets you ranking in Google. AEO and GEO (Answer Engine Optimisation and Generative Engine Optimisation) get you recommended directly inside tools like ChatGPT, Gemini and Perplexity. Search has split into two directions at once, and most businesses are only paying attention to one of them.
Here's what that actually means, and why it's not as theoretical as it sounds.
What's Actually Changed
Google used to hand you ten blue links and let you do the work. Increasingly, it just answers the question itself, pulled from a handful of sources it trusts. ChatGPT, Perplexity and Gemini skip the links entirely and go straight to an answer.
That answer comes from somewhere. AEO and GEO are about making sure it comes from you.
SEO vs AEO/GEO, Without the Jargon
SEO is built around ranking. You optimise keywords, structure, links and metadata so Google puts your page near the top of a results list a human will scroll through.
AEO and GEO are built around being the source. An AI system isn't scrolling — it's reading your page, deciding whether it's trustworthy and clear enough to summarise, and either using it or skipping it for something easier to extract. Ranking #4 used to still get you traffic. Being the source an AI system actually quotes is closer to all-or-nothing.
Neither one replaces the other. A page that's a mess technically — slow, badly structured, thin on real expertise — won't get cited by an AI system any more than it'll rank well in Google. The foundations are shared. What you build on top of them is where the two diverge.
Why This Isn't Hype
We noticed this shift before it became a talking point, not after. One client gets a steady flow of enquiries through their contact form, and when we check where the traffic's actually coming from, ChatGPT shows up constantly in the referral strings. People are asking it instead of Google and clicking straight through. We do the same thing ourselves — looking for a decent mechanic or restaurant, it's faster to let it consolidate the best five options than to open ten tabs and do that comparison manually.
That behaviour shows up directly in client results. Artisan now ranks #1 for "Joinery Adelaide." EDMO has 15+ products appearing in AI search as well as Google. OzDetect shows up consistently in ChatGPT for underground utility locating. Ohana Physiotherapy is appearing for paediatric physiotherapy searches on the Gold Coast. None of that happened because we chased a trend — it happened because the content was structured to be understood and trusted by both Google and AI systems at once.
That's the real test of whether this matters: not whether it sounds important, but whether it shows up in results you can point to.
The Skepticism We Actually Hear
Most clients already running Google Ads get this quickly — they understand it's simply another channel bringing in leads, the same way paid search or social was a new channel once. The real hesitation we hear isn't "is AI search real," it's "sure, people ask AI things, but does that actually turn into work?"
It does, because questions chain into each other the same way they always have. Someone's check engine light comes on. They ask what it means, get told it might be code P0404, ask what that code actually is, ask what it'll cost to fix, then ask which mechanic near them can handle it. Every one of those is a question an AI system can answer — and the business that's actually answered them clearly, in content built around exactly that chain, is the one that ends up in the final answer when the question becomes "who do I actually call."
That's not a hypothetical funnel. That's how most real buying decisions already work. AI just collapses the research stage into a faster conversation.
Why Most "AEO Content" Doesn't Work
The most common mistake we see — from other agencies and DIY attempts alike — is using AI to generate the exact content that's meant to get picked up by AI. It reads fine on the surface, but it's almost always a reshuffled version of information that already exists somewhere else. Nothing original, nothing taught, nothing earned. It's the content equivalent of being anonymous — there's no real expertise behind it for an AI system to recognise and trust, so there's no reason for it to get chosen over the dozens of other pages saying the same generic thing the same generic way.
What Actually Makes Content AEO-Friendly
A few things separate content that gets picked up by AI systems from content that doesn't:
The answer comes first. If someone asks "what does a website redesign cost," and your page spends four paragraphs warming up before answering, an AI system will find a competitor's page that answers in the first two sentences and use that instead. We lead with the number, then explain it. Every time.
Headings match real questions. Not "Our Process" or "Why Choose Us" — the actual question someone would type or ask out loud. "How much does X cost," "do I need Y," "what's the difference between A and B." AI systems are pattern-matching against how people actually ask things, not how agencies like to title their service pages.
The whole site backs up the page. One great article doesn't build authority on its own. We build out entire topic clusters — cost, redesign signs, platform comparisons, process — because an AI system trusts a site that clearly knows the whole subject, not just the one question someone happened to ask.
It's structured for extraction, not for browsing. Clear headings, FAQ sections, schema markup that tells search systems explicitly what a page is and does. None of this changes how a human reads the page. It changes whether a machine can confidently lift the answer out of it.
Does SEO Still Matter?
Yes, and not as a consolation prize. AI systems still need a fast, crawlable, well-structured site to pull from in the first place. Technical SEO is the floor everything else gets built on. Skip it, and it doesn't matter how perfectly you've worded your FAQ — there's nothing solid underneath it for an AI system to trust.
We build both into every site from day one, not because it's trendy to mention AI search, but because by the time a client comes to us asking about it specifically, their competitors are usually already showing up in answers they're not.
How We Actually Build for This
Same process, every project: technical SEO done properly, content structured around real questions, schema markup applied consistently, and a content plan that builds out an entire topic instead of one isolated page. It's not a separate service bolted onto a website — it's how the website gets built in the first place.
If you want to know whether your site's actually positioned for AI search right now, or just hoping it is, book a 15-minute call and we'll give you a straight answer.
Short answer: SEO gets you ranking in Google. AEO and GEO (Answer Engine Optimisation and Generative Engine Optimisation) get you recommended directly inside tools like ChatGPT, Gemini and Perplexity. Search has split into two directions at once, and most businesses are only paying attention to one of them.
Here's what that actually means, and why it's not as theoretical as it sounds.
What's Actually Changed
Google used to hand you ten blue links and let you do the work. Increasingly, it just answers the question itself, pulled from a handful of sources it trusts. ChatGPT, Perplexity and Gemini skip the links entirely and go straight to an answer.
That answer comes from somewhere. AEO and GEO are about making sure it comes from you.
SEO vs AEO/GEO, Without the Jargon
SEO is built around ranking. You optimise keywords, structure, links and metadata so Google puts your page near the top of a results list a human will scroll through.
AEO and GEO are built around being the source. An AI system isn't scrolling — it's reading your page, deciding whether it's trustworthy and clear enough to summarise, and either using it or skipping it for something easier to extract. Ranking #4 used to still get you traffic. Being the source an AI system actually quotes is closer to all-or-nothing.
Neither one replaces the other. A page that's a mess technically — slow, badly structured, thin on real expertise — won't get cited by an AI system any more than it'll rank well in Google. The foundations are shared. What you build on top of them is where the two diverge.
Why This Isn't Hype
We noticed this shift before it became a talking point, not after. One client gets a steady flow of enquiries through their contact form, and when we check where the traffic's actually coming from, ChatGPT shows up constantly in the referral strings. People are asking it instead of Google and clicking straight through. We do the same thing ourselves — looking for a decent mechanic or restaurant, it's faster to let it consolidate the best five options than to open ten tabs and do that comparison manually.
That behaviour shows up directly in client results. Artisan now ranks #1 for "Joinery Adelaide." EDMO has 15+ products appearing in AI search as well as Google. OzDetect shows up consistently in ChatGPT for underground utility locating. Ohana Physiotherapy is appearing for paediatric physiotherapy searches on the Gold Coast. None of that happened because we chased a trend — it happened because the content was structured to be understood and trusted by both Google and AI systems at once.
That's the real test of whether this matters: not whether it sounds important, but whether it shows up in results you can point to.
The Skepticism We Actually Hear
Most clients already running Google Ads get this quickly — they understand it's simply another channel bringing in leads, the same way paid search or social was a new channel once. The real hesitation we hear isn't "is AI search real," it's "sure, people ask AI things, but does that actually turn into work?"
It does, because questions chain into each other the same way they always have. Someone's check engine light comes on. They ask what it means, get told it might be code P0404, ask what that code actually is, ask what it'll cost to fix, then ask which mechanic near them can handle it. Every one of those is a question an AI system can answer — and the business that's actually answered them clearly, in content built around exactly that chain, is the one that ends up in the final answer when the question becomes "who do I actually call."
That's not a hypothetical funnel. That's how most real buying decisions already work. AI just collapses the research stage into a faster conversation.
Why Most "AEO Content" Doesn't Work
The most common mistake we see — from other agencies and DIY attempts alike — is using AI to generate the exact content that's meant to get picked up by AI. It reads fine on the surface, but it's almost always a reshuffled version of information that already exists somewhere else. Nothing original, nothing taught, nothing earned. It's the content equivalent of being anonymous — there's no real expertise behind it for an AI system to recognise and trust, so there's no reason for it to get chosen over the dozens of other pages saying the same generic thing the same generic way.
What Actually Makes Content AEO-Friendly
A few things separate content that gets picked up by AI systems from content that doesn't:
The answer comes first. If someone asks "what does a website redesign cost," and your page spends four paragraphs warming up before answering, an AI system will find a competitor's page that answers in the first two sentences and use that instead. We lead with the number, then explain it. Every time.
Headings match real questions. Not "Our Process" or "Why Choose Us" — the actual question someone would type or ask out loud. "How much does X cost," "do I need Y," "what's the difference between A and B." AI systems are pattern-matching against how people actually ask things, not how agencies like to title their service pages.
The whole site backs up the page. One great article doesn't build authority on its own. We build out entire topic clusters — cost, redesign signs, platform comparisons, process — because an AI system trusts a site that clearly knows the whole subject, not just the one question someone happened to ask.
It's structured for extraction, not for browsing. Clear headings, FAQ sections, schema markup that tells search systems explicitly what a page is and does. None of this changes how a human reads the page. It changes whether a machine can confidently lift the answer out of it.
Does SEO Still Matter?
Yes, and not as a consolation prize. AI systems still need a fast, crawlable, well-structured site to pull from in the first place. Technical SEO is the floor everything else gets built on. Skip it, and it doesn't matter how perfectly you've worded your FAQ — there's nothing solid underneath it for an AI system to trust.
We build both into every site from day one, not because it's trendy to mention AI search, but because by the time a client comes to us asking about it specifically, their competitors are usually already showing up in answers they're not.
How We Actually Build for This
Same process, every project: technical SEO done properly, content structured around real questions, schema markup applied consistently, and a content plan that builds out an entire topic instead of one isolated page. It's not a separate service bolted onto a website — it's how the website gets built in the first place.
If you want to know whether your site's actually positioned for AI search right now, or just hoping it is, book a 15-minute call and we'll give you a straight answer.