01 / Discover • 20 min + 3 conversations
Find a problem worth solving.
A specific person, a recurring problem, and evidence that it matters.
Demand brief
Understand the move.
Start a useful draft now.
Name one kind of person and one recent moment when their work became difficult. Mark anything you have not observed yet as an assumption.
Draft checkpoint: save the first honest note, then stop and do the observation. This checkpoint stays a draft; it does not complete the mission or verify research.
Start with a recent situation, not a feature idea. Ask someone to walk you through the last time they did the work. What triggered it? Where did it slow down? What did they do instead?
Keep observations separate from your interpretation. Three conversations can reveal a pattern worth testing; they cannot prove a market. Look deliberately for someone whose experience contradicts your idea.
Inspect a worked example (optional)
Worked example · a tool for small design studios
The handoff that keeps breaking
Two studio leads rebuilt a client brief from email before kickoff last week. Each lost about 45 minutes. A third studio avoids this with a shared checklist. Our next test is whether that checklist solves the problem for the first two.
The stronger account names observed behaviour, a cost and a counterexample. It leaves room for a simpler solution.
Compare with a weaker answer
Agencies need an AI project manager. Everyone we asked liked the idea.
Loading your saved work…