The customers who said "ChatGPT sent me"

Founder's Notes

by: Tim Holt

A few months back, two of our clients told me almost the same thing in the same week. One runs a primary care practice up in North County. The other runs a home-services company - the kind of business you call when something in your house is backing up at 9pm. Different worlds. Same sentence, more or less: we're getting people who say they found us through ChatGPT. One of them had a patient mention Claude by name.

I've been doing this long enough to be suspicious of anything that sounds like a trend, so my first instinct was to discount it. Self-reported attribution is the softest data there is. But it kept happening, and it kept being specific, and after a while "I don't believe it" stops being skepticism and starts being denial. So I want to write down what I actually think is going on, because I think it matters more than most of the SEO conversation I see.

SEO answers "where do I rank." GEO answers "do I exist."

Here's the cleanest way I've found to explain the difference to a client.

Traditional SEO is a competition for position. You're trying to be the link someone clicks on a page of ten blue links. The whole game assumes the person is going to leave the search engine and come to your site to make up their own mind.

Generative engine optimization - GEO - is a competition for citation. When someone asks ChatGPT or Claude "who's a good primary care doctor in Vista who takes my insurance" or "how do I know if I need to repipe my house," the AI doesn't hand them ten links to sort through. It gives them an answer. Maybe it names two or three businesses. If you're not one of them, you don't lose by a little. You're not in the room at all. There's no scrolling to page two. There's no page two.

That's the part that took me a minute to feel in my gut. In old search, being ranked #6 still meant a trickle of traffic. In AI answers, being "almost recommended" is worth exactly nothing. It's binary in a way Google never was.

What I keep getting wrong, and what I'm now sure of

Let me be honest about the thing people overhype: GEO is not a replacement for SEO, and anyone selling it that way is selling you something. The work that gets you cited by an AI is, to a frustrating degree, the same fundamentals good SEO always rewarded. Clean structure. Real expertise on the page. Clear, plain answers to the actual questions people ask. Schema markup so a machine can understand what your business is, where it is, and who's behind it. None of that is new. AI didn't invent the value of being clear and trustworthy.

What is new is where the payoff lands, and how lopsided the early-mover advantage is right now.

When we rebuilt the primary care client's service pages - proper medical schema, a real FAQ section answering the questions patients actually type, the doctor's credentials made legible to a machine - one of those pages was surfacing as a top AI result for local health searches almost immediately. Not months later. Fast. And the reason it was fast is the uncomfortable, motivating part: almost nobody else in that local market had done the work. The bar to be the cited answer in a specific local niche is currently low, because most businesses are still optimizing for a search experience that's quietly being replaced underneath them.

We saw the same shape on the home-services side. Full structured-data buildout, a real authority foundation for the founder as a named person and not just a faceless brand, the site actually fast enough for a crawler to love. Then, a few weeks on: customers showing up who'd been handed the business's name by an AI before they ever touched Google.

The quiet reason this is worth your attention

Here's what makes me care beyond novelty. The handful of people coming through AI right now convert better than the average click, and it makes sense once you sit with it. By the time someone arrives having asked an assistant a long, specific question and gotten your name back as the answer, they've already been recommended. They're not comparison-shopping ten tabs. They show up halfway sold. Industry data this year points the same direction - AI-referred visitors converting at multiples of regular search traffic - but I trust the thing I can see in my own clients' booking calendars more than any chart.

The volume is still small. I won't pretend otherwise, and I'd be lying to you if I dressed up a handful of customers as a flood. But I've been doing this long enough to know the difference between a small number that's noise and a small number that's a leading edge. This one has the feel of an edge. The overlap between who ranks on Google and who gets cited by AI is already shrinking fast, which means the two systems are drifting apart - and the businesses that treat them as one thing are going to wake up invisible in the one that's growing.

So when a client asks me whether GEO is "more important than SEO," I tell them that's the wrong frame. SEO is the foundation; you don't get to skip it. But the next customer is increasingly going to meet your business for the first time inside an answer, not on a results page - and the question that actually matters is whether you'll be in that answer or watching a competitor get named instead.

Two of my clients already know which side of that they're on. That's the whole reason I'm writing this down.

Best,

Tim