Validate a Business Idea
6 weeks · 3 milestones
Run experiments to confirm real demand before building.
Milestone map
Milestone map
3 milestones
Conduct at least 5 structured problem interviews with people who match your target customer profile. A problem interview is NOT a pitch — you are not describing your solution; you are asking about the customer's current life, their frustrations, and what they have already tried. The interview must explore whether the problem you are planning to solve is real, frequent, and important enough for the customer to pay to fix. Keep a written record of each interview: who you spoke to, what they said about the problem, and how they rated its severity. The key output is an honest assessment of whether the problem is validated — including a clear-eyed view of any evidence that contradicts your assumption.
Proof required
Submit: interview notes or transcripts from at least 5 problem interviews, each identifying the interviewee's role and context (no full names needed), the specific problem statements they described, how they currently address the problem, and their approximate rating of how painful or important it is. Include a 100+ word synthesis assessing whether the problem is validated, partially validated, or invalidated by the interviews.
What gets checked
- At least 5 interviews with distinct individuals who match the target customer profile — 5 interviews with the same person, with close friends or family who will say yes to anything, or with people who do not match the customer profile do not pass; the interviewee's role or context must be described for each
- Interview notes capture the customer's own language, not the founder's framing — a note that says 'they confirmed the problem exists' is a summary; a note that says 'they described spending 2 hours per week manually reconciling spreadsheets and called it their biggest operational headache' is evidence
- Synthesis must include any contradictory evidence — if even one interviewee said the problem is not serious, not frequent, or already solved by an existing tool, that must appear in the synthesis; a synthesis that only reports positive signals is incomplete
Resources
Enroll free to unlock learning resources →Mastery
Lenny Rachitsky — Free Substack on Product Strategy
Free tier — search the archive for 'problem validation' and 'early customers'; the real founder examples show how to interpret ambiguous interview signals
Unlocks after completing Foundation + Depth