Discovery is the process of reducing uncertainty before expensive delivery starts. Delivery builds the solution. Discovery proves you are solving the right problem.
Start with problem framing:
- Who is the target user?
- What job are they trying to do?
- What pain is frequent and meaningful?
- What business outcome does solving it create?
Then list assumptions explicitly. Example assumptions: users face this pain weekly, current workaround is costly, and they would adopt a simpler flow.
Customer interview basics:
- Use open-ended prompts about recent behavior
- Ask for concrete examples, not opinions about hypothetical features
- Avoid leading questions that validate your own idea
Common interview pitfalls:
- Pitching solutions too early
- Sampling only friendly users
- Treating stated preference as behavioral truth
After interviews, synthesize signals into themes: repeated pain points, context triggers, and outcome expectations. Opportunity mapping helps compare where user value and business leverage overlap.
Write hypotheses in a testable format: “If we [change], then [segment] will [behavior], improving [metric] by [range].”
Lightweight experiments:
- Concierge tests
- Clickable prototypes
- Fake-door tests
- Wizard-of-Oz process trials
Evidence quality rubric:
- Weak: internal opinions
- Moderate: qualitative patterns from real users
- Strong: behavior + metric movement from controlled tests
Discovery handoff to execution should include: validated problem statement, prioritized opportunity, risk list, and initial metric framework. This gives Engineering and Design clarity before building.