Most product briefs focus on features. We start with behaviour. What do people currently do? What do you need them to do instead? What are the behavioural barriers stopping them? What would make the new behaviour easier than the old one?

We spend time here because getting this wrong means building the wrong thing. We use behavioural science frameworks to understand the gap between current behaviour and desired behaviour.

We research how the target behaviour works in context. Not through surveys asking people what they'd hypothetically do. Through observation, interviews, and studying how people actually behave in the real world.

We look for: When and where the behaviour currently happens? What triggers it or prevents it? What habits or routines it's competing with? What cognitive load is involved? What environmental factors help or hinder it?

This research phase often reveals that the problem you think you have isn't actually the problem. That's valuable to know before you build anything.

We develop concepts grounded in different behavioural science principles. Maybe one uses implementation intentions to help people plan when they'll use the product. Another uses social proof to leverage what others are doing. A third uses friction reduction to make the desired behaviour absurdly easy.

We sketch, prototype cheaply, and test concepts with real users quickly. We're looking for what resonates behaviourally, not what looks pretty in a presentation.

Once we have a concept that works in theory, we build it and test whether it works in practice. We measure actual behaviour change, not just whether people say they like it.

Does it reduce friction? Does it create the habit loop we designed for? Do people use it once or keep using it? Does it work in the messy real world or only in ideal conditions? We iterate based on what we learn.

Most products go through multiple rounds of testing and refinement before they're ready.

We launch small before we launch big. Pilot with a limited group, measure behaviour change rigorously, refine based on what we learn, then scale. This approach means you're not betting the entire budget on an untested product. You're building evidence that it works before you commit fully.

Even after launch, we keep measuring and refining. Behaviour change products need ongoing optimisation.

What works for early adopters might not work for late adopters. Habits that stick in month one might fade by month three. We design for long term behaviour change, not just initial uptake.