The Prototype Is the Argument

When a team argues about a product decision, they are arguing about competing mental models of how the world works. One person thinks users will understand the flow. Another thinks they won't. There's only one way to resolve it: build the thing and test it.

Arguments like this cannot be won with more conviction, because they are not really about the thing being discussed. They are about the future, and nobody has been there. Whoever is most senior, most confident, or most persistent tends to prevail — which is a poor way to select for what is actually correct. The disagreement is a signal that you have reached the edge of what anyone actually knows.

Prototypes are not deliverables. They are arguments made tangible. The purpose of a prototype is to kill bad ideas quickly and cheaply — before you've invested six months writing code for something that doesn't work.

A good prototype is built to answer one question, not to impress. It is disposable by design, cheap enough that no one is tempted to defend it, and specific enough that its result actually settles the debate. If a prototype could survive any outcome without changing anyone's mind, it was decoration, not an argument.

The discipline is in the throwing away. The purpose was never the artifact; it was the answer. Once you have it, the prototype has done its job — and the six months you did not spend building the wrong thing is the return.