I'm writing this from a coffee shop in Sagres, Portugal. Except it's not really a coffee shop. It's a coffee shop come surf shop come hangout come content opportunity. There's a smoothie bowl menu. There are stickers on the wall from brands I half recognise. Someone nearby is editing drone footage on a MacBook.

Sagres has drawn surfers since the seventies. Great waves, year-round consistency, the kind of place that earned its reputation through word of mouth and salt water. When I first came here, that's what it still felt like. Raw. Unfinished. A few surfers, a couple of restaurants that didn't try too hard, and a kind of silence that made you pay attention.

What's changed isn't the discovery. Sagres was never a secret. What's changed is the layer that's been built on top. The curation. The branding. The smoothie bowls. The place has been "Californicated", not by new arrivals but by a new kind of arrival. People who come for the lifestyle, not the waves.

I don't say that with total disdain, because I'm sat here drinking good coffee and enjoying the wifi. But something has shifted. The original pull, that raw, edge-of-the-world magnetism, has been packaged and polished until it's almost indistinguishable from the packaging itself.

Here's what interests me, though. Nobody marketed Sagres into this.

There was no campaign. No brand partnership. No destination strategy that I know of. What happened was simpler and more powerful. A feeling travelled. And over time, a place rebuilt itself around a question that enough people were asking without quite knowing how to phrase it.

That question was something like: what does it feel like to be at the edge of something?

And that, I think, is what curiosity actually is. Not a tactic or a content format, but a pull. The thing that makes someone lean forward before you've even made your case.

Most marketing gets this backwards

The default mode in marketing is persuasion. We build campaigns to convince. We optimise for conversion. We write copy that tries to close the gap between where someone is and where we want them to be.

Curiosity works the other way round. It opens a gap. George Loewenstein called it the "information gap" theory back in the nineties. Curiosity fires when people sense a distance between what they know and what they want to know. The job isn't to give people answers. It's to make them feel the shape of a question.

Most brands try to be the coffee shop. They get the font right. Curate the playlist. Nail the visual identity. But they skip the bit that made people show up in the first place. The thing you can't quite google. The feeling that there's something here worth finding out.

Being interesting is not the same as making people interested. The difference between those two things is the entire game.

Curiosity built every brand I admire

Think about the brands that actually stay with you. Not the ones with the biggest budgets, but the ones that made you lean in.

Patagonia didn't start by selling jackets. They made you curious about what it means to actually give a damn about the planet and still run a business. The tension in that question is what pulls people in. The jacket is almost incidental.

Monzo didn't launch with a campaign about banking features. They made you wonder what it would feel like if your bank actually talked to you like a human. That single question, the gap between what banking was and what it could be, did more work than any media spend.

I recognise these patterns because I've spent twenty years building them. At their best, the brand experiences I've created work the same way: they start with a question that nobody in the room had thought to ask yet, and then build everything, the touchpoints, the content, the whole journey, around the pull of that question. Around the wanting to know.

The challenge I've been solving

I've spent the best part of a year working at the intersection of two things that most people treat as separate problems. Making a brand visible to AI, structuring content so that LLMs can find it, parse it, and serve it back to people who never visit the website. And making that same content worth a damn when a human actually reads it.

The plumbing matters. If your brand doesn't exist in the training data and the retrieval layer, you're invisible in a way that would have been unthinkable five years ago. The structure, the schema, the way you frame expertise so a machine can recognise it. All of that is real and necessary. I've built the frameworks, the content architecture, the governance models to make it work.

But here's what I've learned in the process. The more you optimise content for machines, the easier it is to lose the thing that makes humans care. You end up writing for parsability. Clear answers. Defined structures. Everything labelled, everything retrievable. That's fine, except that curiosity doesn't live in clean answers. It lives in the gaps between them. In the half-formed thought. In the detail that makes you stop and think, wait, what?

I've seen it happen in real time. You build a content hub that's beautifully optimised for AI discovery, and it works. The brand shows up in ChatGPT, in Perplexity, in the AI overviews. Job done. Except the content itself has become so efficient, so answer-shaped, that there's nothing left to be curious about. The machine got what it needed. The human moved on.

So the challenge I've been solving, and I don't think many people are talking about this honestly yet, is how to do both. How to build content that's structured enough for machines to find and interesting enough for humans to actually want. That's not a technical problem. It's a creative one. It requires you to hold two contradictory things in your head at once: how an LLM reads, and how a person feels.

The answer I've arrived at is that you need both layers, designed together from the start. A machine-readable structure that serves the AI discovery layer, and a human-facing layer that does what the best brand content has always done: makes you curious, pulls you in, gives you a reason to stay. It's an execution of what Neil Barrie's calling "priming and proving." Content that primes the machine to recommend you, and proves to the human that you're worth their attention. Built as one thing, not bolted together.

The brands that will win in an AI-mediated world aren't the ones with the best answers. They're the ones asking questions that make people want to go deeper. That make someone pause mid-scroll and think, actually, I want to read this myself, not the summary, the real thing. That's the experience I'm building. And it's the most interesting brand problem I've worked on in twenty years.

The thing about Sagres

The irony of Sagres is that the thing that made it magnetic is slowly being eroded by the magnetism itself. The rawness gets polished. The silence gets filled. The edge gets rounded off.

There's a lesson in that for brands too. Curiosity is not a one-time trick. It's a practice. The moment you stop asking questions and start resting on the answers you've already found, you become wallpaper. You become the place people have already been. The best brand leaders I know, and the standard I hold myself to, are the ones who keep finding new edges even when the old ones have been smoothed over by success.

The strongest force in marketing isn't persuasion or data or even creativity, though that helps.

It's the ability to make someone feel the edge of a question they didn't know they had. And to keep finding new edges, long after the first ones have done their work.

I'm going to finish my coffee now. The wifi's good, the waves look clean, and I'm curious about what's round the next headland.

Some things still work.

Asked at the counter

What is the information gap theory of curiosity? George Loewenstein's theory from the nineties: curiosity fires when people sense a distance between what they know and what they want to know. Persuasion tries to close gaps; curiosity opens them. The job isn't to give people answers, it's to make them feel the shape of a question.

Can content be optimised for AI and still be interesting to humans? Yes, but only if both layers are designed together from the start. The more you optimise for parsability, the easier it is to lose the thing that makes humans care, because curiosity doesn't live in clean answers, it lives in the gaps between them. It's not a technical problem. It's a creative one: holding how an LLM reads and how a person feels in your head at once.

What is priming and proving? Neil Barrie's phrase for content that primes the machine to recommend you and proves to the human that you're worth their attention. The crucial bit is that it's built as one thing, not bolted together: a machine-readable structure for the AI discovery layer and a human-facing layer that pulls people in, designed from the same starting point.