How I work
Each project has its own constraints and unknowns, so the process adapts. The shape doesn't: a loose double diamond, in four steps.
- ①Look closely. Find out what's happening before deciding what to change. Gather raw signals from people, the product and the systems around it. Stay long enough to know which problem is worth solving.
- ②Make sense. Structure the findings into one clear problem statement and decide where to focus. This is where most of the direction gets set.
- ③Explore solutions. Produce several approaches and test them cheaply before committing to any. More options at this stage make the final choice easier to defend.
- ④Make real. Narrow down to one solution and build it properly, often all the way to a working prototype. Use it, cut what doesn't work, and ship when it's right, even if it already worked a week earlier. AI speeds up the building; what stays and what goes is decided by hand.
How I judge the work
Eight principles, each with the question I ask to test it.
Usefulness
Does it solve the right problem?
Design exists to change something real, not to decorate. A well-executed answer to the wrong question is still a failure, so check the problem before judging the craft.
Contrast
Is it different for a reason?
Good work departs from what came before where the old way falls short, and the difference is visible. Novelty for its own sake doesn't count.
Order
Is the structure obvious?
Hierarchy is clear and the relationships between elements are intentional. Form follows logic, so people understand the structure without being told, and it doesn't date with trends.
Subtraction
Is everything here for a reason?
Removing takes more discipline than adding: a thousand no's protect the one meaningful yes. Every element should justify its presence. If removing it doesn't feel like a loss, it was noise.
Clarity
Does it work without explanation?
People can see what's possible, what's happening and what to do next. If someone needs a walkthrough to understand what they're looking at, the design isn't done.
Speed
Does it feel fast?
Speed is a quality of the design, not only of the code. Fewer steps, feedback the moment you act, the next thing already where you expect it. People should move through it without waiting or wading.
Calm
Does it respect attention?
It shouldn't demand more focus than the task requires. If the interface is the most noticeable thing in the interaction, something is wrong.
Inevitability
Does it feel inevitable?
Good solutions look obvious in hindsight. If one still feels like one of several options, it needs more work. It's done when changing anything would make it worse.