Claude for Product Managers
Lesson 1
The Research and Interview Synthesis Problem
The Research Synthesis Problem
The bottleneck in user research has never been running interviews. It's converting twenty hours of transcripts into two pages of insights that drive decisions. Most teams do this badly they synthesize from memory, they over-index on the most memorable quote, and they skip the step of connecting insights to product implications. Claude changes this. But the prompt matters enormously.
The Interview Synthesis Prompt
I ran [N] user interviews exploring [topic]. Here are the transcripts / notes:
[Paste transcripts — you can do 3–5 at a time if they're long]
Please analyze these and extract:
1. RECURRING PAIN POINTS: Problems mentioned by more than one user, with specific quotes as evidence. Note how many users mentioned each.
2. WORKAROUNDS: Any behavior users have developed to cope with the current experience. These are high-value signals — workarounds mean pain strong enough to motivate behavior change.
3. UNMET EXPECTATIONS: Moments where a user expected something to work differently than it does. Quote the exact moment.
4. SURPRISING FINDINGS: Anything that doesn't fit the hypothesis we were testing. If you don't know our hypothesis, flag this and I'll tell you.
5. WHAT THE RESEARCH DIDN'T ANSWER: Based on the questions users were asked, what do we still not know?
Format: For each item, include (a) the finding, (b) the evidence (quote + user identifier), and (c) a tentative product implication. Label implications as tentative — I'll validate them.