Validate — Test demand before you build

Runs an idea through a stage-appropriate pressure test, designs the cheapest experiment that would settle it, scores the evidence, and produces a dated go/no-go brief with a revisit date.

The goal is to fail fast and cheap — not to confirm what the founder already believes.

Mental model

pressure-test  →  design       →  run           →  verdict
(is the idea      (what's the      (the founder     (score the
 coherent?)        cheapest test    does this)       evidence,
                   of the gap?)                      make the call)

Most founders arrive wanting the verdict without the evidence. The job is to route them back to the earliest incomplete stage, not to score optimism.

Step 0 — Load context