Behaviour change is a UX problem dressed up in a lab coat.

Spend ten minutes inside any company trying to change human behaviour — health, finance, productivity, climate — and you will hear the same incantations. Nudge. Habit loop. Intention–action gap. BJ Fogg. Atomic Habits. The vocabulary is borrowed. The lab coats are rented. The science, mostly, is being used as decor.

Behavioural science is a real discipline and it has real insights. But what most product teams do with it is not science. It is interior design. They wrap a perfectly normal UX problem in academic-sounding language and call it strategy.

"If your behavioural model fits on a slide but your onboarding doesn't fit on a phone, you don't have a behaviour problem. You have a UX problem with a fancy name."

The lab coat tell

You can spot the lab coat by what it lets the team avoid. Every time a real product question comes up — what does the first screen do, what happens after the first failure, why does anyone come back on day three — the team retreats into the framework. They debate cue and reward. They redraw the loop. They do not redesign the screen.

Meanwhile, the user is staring at a notification at 9:14am on a Tuesday and deciding, in less than a second, whether your product is worth their attention. No model survives that moment. Only the interface does.

What I learned building Droplet

Droplet is, at its core, a behaviour-change product. So I read the books. I sat in the workshops. I drew the loops. Then I shipped a first version that obeyed every principle and helped almost nobody. It was theoretically perfect. It was practically lifeless.

What changed things was not better theory. It was making the product easier to open, faster to respond, and impossible to misunderstand in three seconds. The behavioural science didn't get more sophisticated. The UX got more honest.

Why teams stay in the lab coat

A test you can run on Monday

Take your behaviour-change product. Open it on a borrowed phone, in a room you don't usually work in, when you are slightly tired. Try to do the one thing it is supposed to make easier. Notice how many of the frictions you hit are about your model of human behaviour, and how many are about a button being in the wrong place, a label being unclear, or a load time being half a second too long.

If the answer is overwhelmingly the second category — and it almost always is — then the lab coat has been doing you a disservice. Take it off. Open Figma. Fix the screen.

"Behaviour changes when products get easier to use. Everything else is content for the conference talk."

None of this means behavioural science is useless. It means it is a small ingredient in a much larger craft, and treating it as the main course is how teams end up with elegant models and unloved products. The science earns its place when it sharpens the interface. The rest of the time, it's a costume.