Product Design Services

Research is only worth the time if it moves a decision. I start by finding out what people are actually doing today — the spreadsheet they keep alongside your product, the step they skip, the field they fill in wrong every time. That work is interviews, session recordings where they exist, and reading whatever support tickets you already have.

On the DAE Search project, employee interviews were the thing that reframed the problem: the search box worked, but nobody trusted the results enough to stop asking a colleague. No amount of interface polish would have surfaced that. It changed what got built.

If you have no users yet, research looks different and I say so rather than running a study that produces nothing. There are ways to validate a direction before anyone has signed up, and there are ways that only produce the answer you wanted.

Screens, states, and the paths between them. The part that takes the time is not the happy path — it is the empty state, the error, the half-filled form, the record that arrived with a field missing. Those are where products feel broken, and they are usually what got skipped.

I design in a system rather than as a pile of screens: type scale, spacing steps, colour roles, component states. That is what keeps the fortieth screen consistent with the first, and what makes the build fast instead of bespoke each time.

Accessibility is part of the design pass, not an audit afterwards. Keyboard paths, visible focus, and contrast ratios measured with a contrast tool rather than judged by eye — hand-picked accent colours fail AA far more often than they look like they should. I work to the WCAG 2.2 AA criteria; that is a description of how I build, not a certification.

A prototype earns its cost when there is a real disagreement to settle or a flow nobody can picture. Then it is the cheapest way to find out. When the team already agrees, a prototype is a rehearsal of work you could simply do — and building it in React often takes less time than faking it.

That judgement is most of what you are buying. Generating screens is close to free now; deciding which four to delete is the work.

No slide deck as the final artefact. The output is the thing itself, plus enough written reasoning that the next person can tell why a decision went the way it did.

Every engagement starts with a call and a written scope before any money changes hands. I don't publish a day rate, because a two-week design pass and a six-month build aren't the same conversation — the scope decides the number, and you see it in writing first.

After that it is short cycles with something visible at the end of each one. You get a URL early and it stays current — reviewing a real page beats reviewing a screenshot, and it surfaces the problems that only appear at a real width on a real phone.

I work remotely with teams anywhere. Async by default, with a call when a decision needs a conversation instead of a thread.

It suits a team that needs one person to own product design through to a shipped front end — a founder without an in-house designer, or a product team that needs a specific surface built properly without adding headcount.

It suits regulated work. Fintech, healthcare and pharma carry constraints that shape the interface rather than decorate it: approval gates, audit trails, the fields you may not pre-fill. I have spent fifteen years in those constraints.

It does not suit a project that needs a five-person team in parallel, or one where the design is already finished and only needs pixel-pushing. I will say so on the call rather than take it.