Examine the interaction
before expanding the idea.

A sequence of screens, a paper sketch, or a small interactive model can represent an idea without implementing a complete application. Identify where the task starts, what choices the person makes, and what indicates that they have finished.

For a hypothetical shared reading-list tool, a prototype might explore how someone adds a reference and understands who can see it. The model need not include a real account system to support that discussion, but its simulated access behavior must be made clear.

If a review involves people trying a task, the invitation, information handling, and purpose should be appropriate to the activity. Use material suitable for the setting and avoid introducing unnecessary personal information.

Notice where a person hesitates, asks for clarification, or understands a label differently. Record what happened separately from your interpretation of why it happened. A polished screen or a favorable comment does not, on its own, establish that the task is understood.

Explain which parts were interactive, which were simulated, and what the review did not cover. A prototype may reveal a confusing choice without establishing performance, production security, or compatibility with another service.

Use the observations to identify a focused revision or a further question. Keep enough context to compare the next version meaningfully. Do not turn a limited review into a claim about every future user. Examples on this page illustrate a research topic; they do not describe a prototype or participant study delivered by the company.