What is prototyping?
A prototype is a question you can hold. Instead of describing how a page, an app or a flow would work, a team makes a version of it, just real enough to try, and watches what happens. Prototypes range in fidelity. A low-fidelity prototype might be pencil sketches on paper that someone taps through while a designer swaps the pages. A mid-fidelity one is a set of gray boxes linked together in a design tool. A high-fidelity prototype looks and behaves almost like the finished product, with real content, motion and data.
Most digital prototypes today are built in Figma or similar tools, which let designers link screens, add transitions and share a working link with anyone. Some ideas only reveal themselves in code, such as a complex animation, a gesture or a page that depends on live data, so those are prototyped in the browser.
Why it matters
Why prototyping matters
Changes are cheap on paper and expensive in code. A prototype finds the confusing step, the missing screen and the button nobody sees while fixing them costs an afternoon instead of a sprint. It also settles arguments: a team debating two navigation ideas can put both in front of five people and stop debating. And it makes ideas legible to clients and stakeholders who cannot read a wireframe, because clicking through something is a different experience from being shown a picture of it. Many of the worst product decisions were made by people who approved a slide and never touched the thing.
In practice
How prototyping works in practice
- 01
We prototype the riskiest part first. Before designing every screen, we build the flow the whole idea depends on, the sign-up, the booking, the player, and put it in front of real people with a task to complete. We watch, say little and note where they hesitate. Five or six sessions usually surface the biggest problems. Then we fix them and test again, raising the fidelity as the questions get finer.
- 02
Fidelity is matched to the question. If the question is whether people understand the structure, gray boxes are better than polished screens, because people comment on the colors instead of the logic when they see a finished look. If the question is whether a motion feels right, it has to be prototyped at full fidelity, on the device it will run on. Most prototypes are thrown away once they have answered their question; that is what they are for.
Watch out
What people get wrong about prototyping
Polishing the prototype instead of testing it. A prototype that took three weeks to perfect is a design, not a prototype, and the team is now too attached to change it. Second, showing it to colleagues instead of users; colleagues already know how it is meant to work. Third, letting a high-fidelity prototype set expectations the build cannot meet. A prototype can fake data, speed and edge cases that the real product will have to handle honestly. And skipping it because the deadline is tight, which is precisely when a mistake found late costs the most.
