What to give
- Product and the question the interviews explored: [e.g. a grocery app; why first-time users do not place a second order]
- Notes from each interview: [paste each under its own label, P1, P2 and so on, with role or segment if known]
- Number of people interviewed: [e.g. 9]
- Constraints the team works within: [e.g. two engineers, one quarter, no change to pricing]
Check each interview
- List every problem, need or workaround mentioned, with the participant label.
- Keep the participant's own words for the strongest line.
- Mark whether it came up unprompted or only when asked.
- Note the segment, so a problem from one segment is not read as everyone's.
Then rank
- Group the problems into opportunities, each written as a user need, not a feature.
- Frequency: how many participants raised it, out of the total.
- Severity from 1 to 3: 1 is an annoyance, 2 costs time or money, 3 stops them using the product.
- Score as frequency multiplied by severity, and rank.
- For the top five, write a how-might-we question and one experiment the team could run in under two weeks.
Output format
| Rank | Opportunity | Raised by | Severity | Score | Two quotes with labels | How might we | Experiment |, then a short list of what the interviews did not cover and should next time.
Guardrails
- Use only the pasted notes; never invent a quote or a participant.
- Keep requests for specific features separate from the need behind them.






