GLP-1s have been the biggest story in wellness for a few years now, and the numbers have stopped being abstract. Roughly one in eight American adults is on one — Ozempic, Wegovy, Mounjaro, whatever's next. Serena Williams is injecting herself on camera in a Super Bowl ad. And for the first time in a generation, the US obesity rate has actually started to come down. Whatever you think of the drugs, something real is happening.

So when we sat down to pick this quarter's internal project, the GLP-1 boom was the obvious place to look.

Every quarter, the team and I pick a category, give ourselves a deadline, and build a full brand for no one — strategy, identity, product, campaign, the film, the whole thing. It runs like client work, minus the client: a real brief, a real ship date, the lot. It's one of the most fun things we do, and we always learn something along the way.

The point isn't freedom for its own sake. The point is pressure-testing. Can we take an underserved category, build something that belongs on the shelf, and ship it in a month? If we can't do it for ourselves, we've got no business saying we can do it for you.

The other part, which matters more than the portfolio line: client work can quietly sand down your taste. You stop pitching the thing you actually want to make and start pitching the thing you know will clear seven stakeholders. You can do that for years before you notice. Internal projects are the thing that reminds us what our work looks like when we're the ones deciding.

This one's called VERVE. A GLP-1 brand for no one — except the longer we spent on it, the more we caught ourselves wondering whether we should just build the thing for real.

The consumer experience around one of the biggest shifts in how bodies work in a generation is genuinely embarrassing. Every app we looked at was a diet app with new paint on it: calorie counters, streaks, cartoon vegetables, design that whispers "lose weight" while the voice shouts "you can do it bestie" and the underlying logic amounts to "tell us what you ate and we'll grade you on it." None of which maps to what the drug is actually doing to the person using it.

The drug kills appetite, which sounds fine until you realise people stop eating protein and start losing muscle along with the fat — up to forty percent of the weight lost can be lean mass if no one intervenes. Energy flattens. Nausea shows up at two in the morning and sends people to Reddit for answers because their app is still busy logging breakfast. A Tuesday injection day feels nothing like a Friday, and not one of the existing products seems to know that.

Thirty million people willing to inject themselves every week, and zero premium brands had been built natively for any of it. That was the opening. We took it.

The whole category runs on subtraction — less appetite, less food, less weight, less body — and every screen, tagline and downward-arrow graphic reinforces it. But nobody chose that. They inherited it. The drug itself is subtractive: it removes appetite, and the apps built around it quietly copied its logic without noticing. The colours went pale because the mood was careful. The type went thin because the feeling was delicate. The interfaces counted down because the mental model was depletion. The design language of the category is just the pharmacology, reflected back.

Which would be fine, except a companion product's job is the opposite of the drug's job. The drug takes away. The thing that helps you live with it has to put back. On a GLP-1 your requirements actually go up — more protein to protect muscle, deliberate strength work since lean mass is leaving either way, better information faster because symptoms don't wait till Monday. Restriction stays the organising logic long after the drug has made restriction irrelevant.

That became the filter for every decision, every screen, every line of copy. Reductive or constructive? If it was reductive it wasn't ours — the category can have the subtraction, VERVE owns the construction: the stronger body, the preserved muscle, the person who comes out of this more capable than she went in.

The positioning fell out of that filter without much argument: VERVE is the premium companion for GLP-1 users who want to lose fat without losing strength — protein-first nutrition, muscle-preserving training, coaching that reacts to symptoms instead of pretending they don't exist. But there was a second decision underneath it that mattered more. We refused to compete in weight loss at all.

We named a new category instead — GLP-1 lifestyle optimisation — because in a category of one, you set the standard by default. The existing players split into two camps: legacy weight-loss apps bolting a GLP-1 toggle onto a willpower model, and trackers that log what you did without ever telling you what to do. The premium, build-don't-subtract position was sitting empty.

The whole category sells less. We went the other way:

"More you. Every day."

Before anything looked like anything, we wrote the brief. And the brief had to be honest about something the category refuses to say out loud: these drugs don't come with instructions. Not real ones.

You get a prescription, a pen, maybe a ten-minute appointment. The drug handles appetite — that part works. But up to forty percent of the weight lost can be lean mass, not fat. Nobody tells you that at the pharmacy. Nobody tells you that you need over a hundred grams of protein a day to protect your muscle, which is nearly impossible when the drug has killed your hunger. Nobody tells you that nausea peaks on days one and two post-injection, that hair thinning at month four is common and usually temporary, that the reason you look "deflated" despite the scale going down is muscle loss, not your imagination.

Right now a GLP-1 user's daily toolkit looks like this: MyFitnessPal for calories, a Reddit thread for side effects, a YouTube trainer for workouts, their GP for ten minutes every three months. Five sources. None of them talking to each other. None of them built for this.

Writing that brief radicalised the team a little. By the time we got to design, we weren't trying to make a nice health app. We were trying to make the one that should have already existed.

Every project starts with a moodboard that's really an argument. Ours was about two things: what VERVE should feel like, and what it should refuse to feel like.

We pulled Oura, Function of Beauty, Liftus, calm data-dense interfaces, warm neutrals, geometric type systems, editorial photography with real bodies in real light. The reference wasn't "here's what health apps look like." It was closer to: here's the cultural register we want to operate in. Premium but not cold. Precise but not clinical. The kind of brand you'd trust with your body because it clearly took your intelligence seriously.

The moodboard had a clear centre of gravity: depth over brightness, structure over decoration, warmth through proportion and spacing rather than colour. That sounds abstract until you look at what we were rejecting — the pastel-and-gradient wellness category, the clean-blue clinical category, the neon-green fitness category. The visual direction was built as much from refusals as from references.

The palette was the first real creative fight.

We had a direction — deep, authoritative, warm — and we had a rule: nothing that codes as diet, clinical, or trend-driven. What we landed on was a "warm dark mode" as the primary register. Midnight Slate (#1A2121) as the foundation: almost black, with just enough warmth in the undertone to feel alive. Warm Parchment (#EBE7E0) as the light surface: an off-white that reads premium the way heavy stock paper does. Warm Stone (#DCCCBA) bridging the two. And Oxidized Amber as the only accent — used specifically to celebrate the things VERVE celebrates: protein targets hit, lean mass preserved, strength gained.

The amber is the one we kept questioning. It sits close enough to coral to make us nervous. But coral reads femtech, and amber reads differently — it reads earned, analogue, warm-but-weighted. The distinction matters because the whole system lives or dies on whether the brand feels like it was built for strength or built to soothe. Amber stays.

Typography followed the same logic as the palette: substantial over delicate. Clean geometric sans-serif, medium and semi-bold weights, generous letter spacing. The brief was explicit about what we were avoiding — ultra-light, fragile, or overly technical fonts. The category defaults to hair-thin type because thin type signals effortlessness, and effortlessness is the promise the category is built on. VERVE's promise is the opposite. The type needed to carry weight, because the brand does.

This is where it either becomes real or stays theoretical.

The first question in onboarding isn't your goal weight. It's which medication you're on, what dose, and when you last injected — because that starts the clock everything else runs on. By the time you reach the dashboard, the app already knows more about your situation than every other product in the category combined.

Every other health app asks: what did you do today? Then it logs it. VERVE asks: how do you feel today? Then it tells you what to do. That inversion is the entire product.

The daily check-in — Gentle, Okay, Good, Strong — changes everything downstream. On a gentle day the plan softens: nausea-friendly meals, lighter training, the copy telling you this is a peak medication day. On a strong day it's a different product entirely. Same app, same user, two completely different experiences depending on where she is in her week. No other product in the category knows the difference between a Tuesday and a Saturday.

The weight screen doesn't say "you lost 22 lbs." It says: here's what you've actually built — fat and lean split, strength gains alongside. Goblet squat up 20 lbs. Dumbbell row doubled. The insight the category won't say out loud is the thing VERVE leads with.

Four weeks in, the category brief we started with had become something bigger. We set out to build a GLP-1 companion app. We ended up building an argument about what health brands owe the people they're designing for.

The diet-app category has spent decades optimising for one thing: making you feel like the problem. The goal is always smaller, the metric is always a deficit, and the design language — the pastels, the streaks, the cartoon vegetables — is built to make that deficit feel aspirational. The GLP-1 drugs broke the model. The appetite suppression the whole category was built around managing through willpower is now handled pharmacologically. So the design problem underneath it is completely different. What do you build for someone whose body is changing faster than anyone is helping them manage?

Our answer was a system where warmth and precision aren't in tension — they're the same thing. Where the most caring thing the interface can do is tell you exactly what's happening, in plain language, at the moment you need it. Where leading with protein instead of weight isn't a brand decision, it's a clinical one that just happens to also be the right design decision.

The best design decisions in this project were refusals. Refusing the obvious palette. Refusing to open with weight. Refusing the category's own assumptions about what the user needs to see first.

VERVE might never ship. The thinking it forced is already inside the next real brief we write, which was always the point. Except the group chat hasn't stopped — three weeks on, still quietly arguing about whether we should just build the app ourselves. We probably won't. Probably.

Next week: we made the film. A 60-second hero and two supporting cuts, all built around the same strategic idea. The brief was simple — show what the drug doesn't handle. How we answered it is its own story.