04/08/2026
A few weeks back I put out a video explaining how AI is changing the way people actually search for and find businesses. It's the kind of thing that's easy to nod along to and then forget about, until one of my clients did something smarter than just taking my word for it.
They went and searched for themselves.
Not their own brand name, the generic terms real customers actually use for what they sell and they found their business showing up across multiple AI tools, multiple times, for searches they'd never specifically optimised for.
They were surprised. Seeing it laid out in front of you helps you see the theory match reality. So instead of leaving it as a nice story, I pulled the actual data.
Here's what it showed.
I built this client a dashboard that tracks AI-driven traffic as its own separate category, distinct from organic search, paid ads and direct visits, because lumping it in with regular organic traffic hides exactly the thing you actually want to see.
Over a 45-day window, nearly 150 people reached their site specifically through AI tools. Not Google's normal search results. AI systems reading the site, deciding it was worth surfacing, and sending someone through directly.
And it wasn't junk traffic either. People arriving that way tended to already understand what the business did and why it might suit them, because whatever the AI tool had pulled together to answer their question had already done some of that convincing before they ever clicked through. That's a genuinely different kind of visitor to someone who's clicked a cold ad.
Here's the bit I actually want you to take away from this, more than the numbers themselves. This client's website wasn't built with AI search in mind. Most of the content driving this result existed long before anyone was talking about AI Overviews or AI search traffic as a category. What actually drove the result was years of real content work already sitting on the site: genuine pages, genuine articles, real testing of what actually resonated with real customers, built up over a long time, ranking on Google the ordinary way the entire time.
That work didn't expire. It compounded. AI systems pulling together an answer to someone's question are drawing from exactly that kind of material, and good work done years ago is showing up in those answers today, often without the business even realising it's happening.
From what I've seen across client accounts, three things seem to determine whether AI treats a business as a genuine, citable expert, or just another generic listing that gets skipped over.
The first is clarity about exactly who you help. Vague, broad positioning gives an AI system nothing specific to match you against. The more precisely you can say who you're for and what problem you solve, the easier it is for something trying to make a confident recommendation to point at you specifically.
The second is a consistent identity that exists beyond just your own website. Reviews, mentions elsewhere, other people and other sites corroborating who you actually are, all of it helps confirm you're a real, established business rather than an unverified entry with nothing backing it up. Without that corroboration, you read as generic. With it, you read as an actual expert in your niche.
The third is genuine depth, not generic coverage of a topic. Content that goes deep on a specific service, a specific location, a specific type of problem, gives both a human reader and an AI system something concrete to point to. Shallow content covering the same ground as a hundred competitors gives neither of them a reason to pick you specifically.
Now here's the actual thesis underneath all of this, and it's the reason I wanted to write this up properly rather than just post the video and move on. If genuinely good work was done on your website in the past, you already have a real shot at being found through AI search today, even if nobody involved in making it had AI in mind at the time. That's exactly what happened here. It's not starting from zero.
And if that existing work gets revisited and deliberately updated to meet what AI systems and current search quality standards actually look for now, clearer expertise signals, more specific and current content, real corroboration beyond your own site, the chance of being found and actually recommended goes up again on top of that. You're not replacing what you built. You're making it work harder than it already is.
That's genuinely good news if you're sitting on years of content you've mentally filed away as "just old blog posts we should probably get around to updating one day." It might already be earning more than you think. And it's very likely worth a proper look before you assume it needs to be thrown out and started again.
I've written up the full version of this, including the exact three factors in more detail, on my website