Validate a startup idea before you build anything
Enthusiasm for an idea and evidence it's worth building are different things.
- Outcome
- A sharpened hypothesis about who has this problem and why, real conversations that test it, and an honest go/no-go instead of building on excitement alone.
- Time
- Spans 1-2 weeks of real conversations
This workflow is part of PromptKit Pro
A handful of prompts and workflows stay free for everyone — Pro unlocks the full library, including this one, plus the guided runner.
Startup ideas often go straight from excitement to building, skipping the step where the idea gets tested against real people who might actually have the problem. Enthusiasm is not evidence, and building first means finding out whether anyone wants it only after months of work. This playbook forces the validation step: sharpen who has this problem, have real conversations that could actually disprove the idea, and make an honest go/no-go call based on what you learn.
When not to use this
If you've already validated this specific idea with real customer conversations and evidence, skip straight to building. This is for an idea that's still mostly conviction and hasn't been tested yet.
Did this work?
You have real evidence from actual conversations, not just internal conviction, informing whether and how to build this.
Tested on claude-opus-5. Evidence status is draft; it moves to battle-tested only on recorded runs, never by hand.