product-discoverylisted

# De-risk product bets before building Discovery's job is to fail fast and cheaply — identify wrong assumptions before they're baked into shipped software. ## Opportunity Solution Tree (Teresa Torres) ``` Desired outcome (metric to move) └── Opportunity (unmet user need / pain / desire) └── Solution idea (intervention) └── Experiment (cheapest test) ``` Rules: - Opportunities come from user evidence — interviews, support tickets, analytics — not internal opinions. - One desired outcome per tree. Multiple outcomes = no prioritisation. - Solutions are hypotheses; experiments are the cheapest way to test each hypothesis. - The tree is a living document — update as evidence accumulates. ## Assumption mapping For each solution idea, map its assumptions across four risk dimensions: | Dimension | Question | Example assumption | |-----------|---------|-------------------| | **Desirability** | Do users want this? | "Users will pay $10/month for this feature" | | **Viability** | Does this create sustainable business value? | "This will reduce churn by 5%" | | **Feasibility** | Can we build this? | "The API supports the required event granularity" | | **Usability** | Can users use this without training? | "Users will understand the new onboarding flow without docs" | Score each assumption: **Risk** (1–3) × **Certainty** (1–3 inverse — low certainty = high score). Highest scores = test first. ## Validation methods (choose by cost) | Method | Cost | Validate