SEO and GEO are merging. AI tools still rely on search, so clear content, sound site structure, reviews and brand authority still matter. Strengthen those foundations, answer buyer questions plainly and track which sources AI cites.
Platform Intelligence · Part 2 of 5
SEO is changing.
It's merging with GEO, generative engine optimisation, and the two are becoming one job and that's good news for your store. Every few months we read that the whole thing is finished - the AI overview ate it, the blue links are gone, no point optimising for a search engine when no one clicks anymore. Good headlines. But not what's happening.
And if you've spent years and money getting your fundamentals right, most of that work still counts. Probably more of it than you'd think. It's feeding a different machine now, that's all.
Generative engine optimisation. Strip the jargon and it simply means getting your store set up so the AI tools see you and recommend you when someone asks them what to buy, whether that's ChatGPT, Google's AI overview, Perplexity or Copilot. It's the same principle as SEO, just looking at a new kind of result.
So what's actually new about it?
To be honest, not a lot once you know how these tools build an answer.
When you ask an AI for the best products in a category it needs to start somewhere.
It goes and searches. It breaks down your question into a handful of smaller searches behind the scenes, runs them on the web and the search indexes, reads what comes back, and stitches the lot into a tidy answer. Shopify and others call this a query fan-out. One question, several quiet searches and back with one reply. Now that changes how we think about whether SEO matters, because the AI is still leaning on search to build its answer.
The things that helped you rank pre-Ai are the same things that get you pulled into the answer post-Ai. Clear pages, good structure, citations and real brand authority. Strong search visibility raises your odds of being cited. Neil Patel makes the same point, Shopify makes it in their own GEO guide, and it lines up with how these models work.
The fundamentals are the foundation and AI sits on top.
Neil Patel has a phrase I like for the shift: search everywhere optimisation. Your customer isn't searching in one box anymore. She's asking ChatGPT, scrolling Google's AI answers, reading Reddit, watching a creator on her phone while she waits for her coffee. Your job is to be present and consistent across all of it. SEO is one room in that house now rather than the whole house.
Most of it.
Clear, descriptive content that answers the questions your buyers are asking. Page structure and titles, category metafields and FAQs so the machine knows what a page is about. A clean, crawlable site, because there are more crawlers from more places trying to read you than there have ever been. And your reviews and your ratings, the words real customers use which are a bigger trust signal for AI than they ever were for search. None of this is new work. It's the same work we've been doing all along.
The minute a new channel appears an industry springs up to sell you tools for it, and there's a wave of "AI visibility" trackers about at the moment promising all sorts of things. Rand Fishkin reckons AI-mention tracking on its own is becoming a couple-hundred-million-dollar market. He's also pretty blunt that most of it has next to no return, because AI answers are a bit of a lottery by design. Ask the same question twice, get two different answers.
Is that worth a monthly subscription?
I'd go into this carefully. Run the free test, run it often, and note whether you're showing up and whether what it says about you is accurate. Paying to obsess over a number that changes every time a model updates is a different thing altogether, and nobody can tell you which of those trackers will still be around next year.
Put that money into building your fundamentals and your brand authority instead. You own it. Its value compounds. A tracking dashboard won't. New channel, same rule as before.
Take your five most important pages and read each one as if you're the machine trying to recommend it. That's harder than it sounds, and it usually takes a couple of goes.
Make sure your site answers the questions buyers are asking before they buy, in plain words. Sizing, materials, shipping, what happens if it doesn't fit.
Then go and run your brand test and note which sources the AI cites.
That list will tell you more than any tracker will.
