Diagnose First, Design Second

There's a pattern I see in every design team I've worked with. Someone identifies a problem — or what they think is a problem — and the immediate response is to open Figma.

Wireframes appear. Mockups get polished. Prototypes get tested. And three months later, the metrics haven't moved because the team solved the wrong problem beautifully.

The diagnostic gap

Most design processes are treatment-first. We jump to solutions because solutions are tangible. They feel like progress. A wireframe in Figma looks like work. Reading support tickets for two days does not.

But the best designers I've worked with — the ones who consistently ship work that moves numbers — they spend disproportionate time in diagnosis. They're reading data. Talking to users. Mapping journeys. Asking "why" five times before they ask "what if."

Julie Zhuo puts it simply: diagnose with data, treat with design.

What diagnosis looks like

Diagnosis isn't research for research's sake. It's directed investigation with a clear output: a reframe.

A reframe takes the stated problem and replaces it with the actual problem. "Users don't like our dashboard" becomes "users can't find the three workflows they use daily because the dashboard is organized by feature, not by task."

That reframe is the design. Everything after it — the wireframes, the prototypes, the visual polish — is execution.

The uncomfortable truth

Diagnosis is uncomfortable because it often reveals that the thing the team wants to build isn't the thing the users need. It creates tension with roadmaps, stakeholder expectations, and the sunk cost of previous decisions.

But that tension is the job. Design leadership means having the conviction to say "we're solving the wrong problem" and the evidence to back it up.

A practice, not a phase

Diagnosis isn't a phase at the beginning of a project. It's a continuous practice. Every design review should ask: are we still solving the right problem? Every metric review should ask: did our treatment work, or do we need a new diagnosis?

The best design work I've ever done started with the longest diagnosis. The solution was almost obvious by the time I got there. That's how you know the diagnosis was right.