All articles
The blueprint

Experiments: how to actually test

Spotting a worry isn't enough. You have to test it. Every one names a way to test it and offers Build this experiment, which generates a ready-to-run kit.

The method, named (and explained)

Everyone hits a different wall: sometimes you don't know which test to run, sometimes you don't know the term. Each item names the cheapest believable method from a curated playbook (customer interview, fake door, landing-page smoke test, concierge, Wizard of Oz, pre-sale/LOI, pricing test, expert interview, and more), each with a plain-English definition, so a term like "fake door" stops being jargon.

The kit

Build the experiment and you get:

  • What it is and why it fits this worry.
  • Run it: clear, concrete steps.
  • Your materials: shaped to the method: an interview script, the fake-door copy, price points, and so on. Ready to copy and use.
  • Read the result: a clear Passes if… and Fails if… with a rough target and how many people you need, so you decide up front what the result will mean.

Once you've built one, the button flips to View experiment so you can tell at a glance which ones you've worked through.

Deciding your pass/fail line before you run the test is the one thing most people skip, and it's the thing that makes the result worth trusting.