The hook
We almost spent a quarter building a feature nobody had asked to pay for. Instead we spent a weekend faking it — and the fake taught us more than the roadmap ever could.
The thesis in one move
The expensive question in product isn't can we build it — it's will anyone use it. Pretotyping answers the second question first, with a convincing fake, before you commit engineering time. You are not testing whether the product works; you are testing whether the demand is real.
The body
Here is the weekend test we run. Friday: write the landing page for the feature as if it already shipped — the headline, the one screenshot, the call to action. Saturday: put a button where the feature would live and wire it to a "we're rolling this out — want in?" capture instead of real code. Sunday: drive a trickle of real traffic to it and watch what fraction of people click the button that leads nowhere.
That click is the signal. It costs you a weekend and a few hours of copywriting, and it replaces a roomful of opinions with a number. In our case the number was brutal and useful: a 2% click-through against a feature three engineers were ready to spend six weeks on. We killed it on Monday and redirected the team to the thing that did pull — a humble export button that quietly tested at 19%.
The discipline that makes this work is honesty about what the fake is measuring. A pretotype measures intent, not satisfaction. It tells you people wanted to try it, not that they'll love it. That's fine — wanting to try is the gate that kills most bad ideas, and it's the cheapest gate to build.
So what / what to try
Pick the next feature on your roadmap that you're least sure about. Before you scope it, give it the weekend test: a page, a button, a number. Let the fake earn the build. The worst case is you spend a weekend; the best case is you save a quarter.