100 topics · Read, check sources and follow your curiosity.

A little room to wonder

Products & Prototypes

A prototype is a testable representation of an idea. It should reproduce the features needed to answer a question without pretending to be a finished product.

Which parts of an idea need to be realistic for the next test to mean anything?

Several paperclips show repeated loops, curved ends and different surface coatings.
These are finished paperclips, not documented prototypes. Comparing their visible forms can help frame a prototype question, such as which opening or surface feature is worth testing.Dietmar Rabich · Source · CC BY-SA 4.0

Give the prototype one clear job

A prototype may explore appearance, movement, size or an interaction. A paper sketch can test whether a label is understood, but not whether a hinge survives repeated use. An attractive rendering may communicate a look while providing little evidence about how a material behaves.

Fidelity describes how closely a representation reproduces aspects of the intended product. It is more useful to ask “realistic in which respect?” than simply to demand high fidelity everywhere. A rough model can answer a well-chosen early question.

Define the success condition before testing

A useful test states its task and what would count as an informative result. “Does it work?” is too broad. “Can a seven-centimetre-wide paper card pass through this opening without bending?” gives a specific criterion.

A failed test can reveal a wrong assumption before more effort is spent on a finished object. Record the version, the conditions and the result. Changing several features at once may improve the object but makes it harder to explain which change caused the difference.

Worked example: a fictional storage sleeve

A learner sketches a sleeve for a paper card 7 centimetres wide. Version A gives it a straight opening 5 centimetres wide. In a flat paper model where the card must enter without bending or rotating, the opening fails the requirement because 5 is less than 7.

Version B uses an 8-centimetre opening, leaving 1 centimetre total clearance: half a centimetre on each side if centred. That answers one size question under the model’s assumptions. It does not establish seam allowance, material thickness, durability or how an actual hand grips the card. The next test could examine an opening tab, while keeping the chosen width constant.

A working model is not a release decision

A prototype can omit safeguards or use substitutes that would be unsuitable in a real product. Keep it clearly identified as a model. A test involving one fictional scenario or one person’s walkthrough is not proof that every intended user can use it.

Use folded paper, drawings or an existing safe object for this activity. No blades, hot glue, electrical parts or purchases are needed. The point is to choose an informative comparison and explain its limits, not to manufacture a polished object.

Only if you feel like trying

Choose the cheapest useful representation

  1. Pick an invented sleeve, storage tray or screen control and state one question about it.
  2. Choose a sketch, paper model or spoken walkthrough that can answer that question.
  3. Record one result it could support and two things it would leave untested.

Optional. Work on paper, point to the example or talk it through. No account, purchase or public sharing is needed.