A polished prototype is not always the best first test. Choose the level of detail that exposes the assumption under review without making the result look more settled than it is.
Stanford d.school's Design Thinking Bootleg includes prototyping and testing as modes of design work. Design Thinking Bootleg.
Isolate one uncertainty
For a hypothetical physical character, a rough volume model might explore proportions while a graphic mock-up explores a placement idea. Neither should be presented as a fully specified product. State which question the prototype can and cannot answer.

Record what changed after the test
Keep the observation, interpretation and next design action separate. A positive reaction to appearance is not confirmation of construction or performance. If the prototype's limitations affected the response, preserve that context.
The useful output is learning that informs a decision. Increasing fidelity should serve the next question, not become an automatic way to make an unresolved idea look ready for production.

Source reviewed 25 October 2026. Practical interpretation and original illustrations by PLUSHIP / Qunhe; examples are illustrative.
Related reading
A Small-Scale Trial Should Answer the Next Delivery Question
A Storyboard and a Physical Prototype Test Different Assumptions
