Research / Structure / Validation / 04
Prototype the risky part
Choose fidelity by uncertainty. Prototype decisions and edge cases before decorating screens.
Before you begin
By the end, you can…
- Match prototype fidelity to uncertainty
- Include states beyond the happy path
- Write a task that tests behavior rather than opinion
01 / Understand
Prototype the question
A prototype is an instrument for learning. If the uncertainty is whether people understand the sequence, paper may be enough. If the uncertainty is keyboard behavior or perceived latency, a coded prototype may be necessary.
High visual polish can make a weak concept feel finished and discourage honest critique. Spend fidelity only where it helps answer the current question.
02 / Apply
States are part of the design
Include loading, empty, partial, invalid, permission-denied, offline, and success states when they affect the decision. A single golden path hides the moments where trust is won or lost.
Give participants a goal and context, not click instructions. Observe where they start, what they expect, and how they recover. Ask them to explain what they think happened after acting.
03 / Make
Your studio task
Make — Prototype only the least-certain moment in a flow, including loading, empty, and error states.
- Name one uncertain product decision.
- Choose the minimum fidelity required to test it.
- Prototype the core path plus two risky states.
- Run a task without coaching and record expectations, errors, and recovery.
A teammate can test the risky decision without being distracted by visual polish.
04 / Check