Timedora

Prototyping to Learn

Learn to frame a design question, choose a prototype that can answer it, and use observed evidence to guide what to try next.

Choose a That Answers the Question

to learn, not merely to build

A is a representation of an idea made to answer a design question. It might be a sketch, model, role-play, mock-up, or working system; it does not have to be an early version of the finished product. The guiding principle is to make the simplest representation that can provide useful evidence about a particular .

Begin by stating what you need to learn. Make the question specific enough to guide both what you represent and what you observe. For instance, “Can people find the right bus stop quickly?” can become “Does a route map help first-time riders identify the correct stop?” A paper map showing a few routes and landmarks would let you observe where riders look, which stop they choose, and what they misunderstand. Building a complete transit interface would add details that this question does not require.

Match the prototype to the question
Design questionUseful prototypeWhat it can reveal
Do people understand or value the idea?Storyboard, concept card, or short explanationInterpretation, interest, and concerns
Can someone complete a task or understand a layout?Paper screens, wireframe, or clickable mock-upNavigation, missing information, and task flow
Does a physical form fit, feel comfortable, or work in context?Quick physical mock-upSize, reach, grip, posture, and physical constraints
How might a service or interaction feel?Role-play, staged scene, or experience prototypeHandoffs, timing, emotional responses, and overlooked steps
Will a technical mechanism work?Focused functional experiment or technical modelWhether a mechanism or component performs as expected
Examples of matching a prototype form to the design question and evidence sought.

Choose what to represent—and how much detail

Represent the part of the idea that bears on the question, and leave unrelated details unfinished. forms are options, not rigid categories: they can be combined, and physical objects, digital experiences, workflows, and services can all be prototyped. For example, a cardboard kiosk could include paper screens while a person acts as the attendant.

Low- prototypes—such as sketches, paper screens, simple models, or improvised role-play—are quick ways to explore ideas and broad assumptions. Their roughness can also signal that the design remains open to change. Higher- prototypes can help investigate detailed questions about visual design, interaction, or realistic use, but take more effort and may draw attention to details that are not being tested. need not be all-or-nothing: increase detail only where the question requires it.

Work through the question, then learn from the test

Suppose a team is unsure whether a reusable container’s one-handed lid is easy to open. The concerns grip and movement, not appearance or manufacturing cost. A quick foam or cardboard lid-and-container mock-up can be enough to let people try the opening action. Invite people with different hand sizes or levels of dexterity, and observe whether they can open and close it without assistance. The evidence comes from what they do; avoid explaining how the is supposed to work.

Change the question and the appropriate changes too. To learn whether the container fits in a lunch bag, the team needs a rough model with accurate dimensions. A highly detailed lid mechanism would not answer that fit question better. Similarly, if the is whether people want a service, a brief story, conversation, or acted-out encounter may reveal more than a detailed interface.

A useful cycle is to name the , choose what to represent, make the simplest testable version, plan what to observe, and record what happened, what surprised you, and what remains uncertain. Then revise the representation or the question and test again. A confusing concept or failed mechanism can still be useful if the result changes what the team knows and tries next. A is a tool for learning; its detail should serve the question, not substitute for one.