The skill: Turning messy signals — interview notes, support tickets, NPS comments, sales call recordings, feature requests — into clear patterns and actionable opportunity statements. The bridge between listening and deciding.
Make it a weekly ritual
Feedback synthesis that happens monthly is archaeology. Weekly synthesis catches trends while they're still actionable. Block 30-60 minutes every Friday. Pull from all sources: interview notes, support tickets, NPS comments, sales call recordings, feature requests, social media mentions.
Support tickets especially — every ticket is a user telling you where your product failed them. Tag by theme, track frequency over time, and watch for workarounds. Workarounds reveal unmet needs better than feature request forms ever will. A user who built a spreadsheet to work around your missing feature is telling you something more valuable than a user who clicked "upvote" on a request board.
Separate what you saw from what you think
The most important discipline in synthesis: keep observations and interpretations apart. "5 users mentioned they can't find the export button" is an observation. "Users don't need export" is an interpretation. Record both, but label them differently. Mixing them up is how teams make confident decisions based on conclusions nobody actually validated.
Never make a product decision based on a single data source. Two sources agreeing gives confidence. Two sources disagreeing means dig deeper. One source alone — no matter how compelling — is one perspective.
Weight by commitment, not volume
One customer who churned over a problem outweighs ten who mentioned it in a survey. Actions beat words. A user who built an elaborate workaround is telling you something more important than a user who upvoted a feature request.
When scoring priority, use frequency × severity:
- High frequency + low severity — many users mildly annoyed. Usually a UX fix: confusing labels, extra clicks, poor defaults.
- Low frequency + high severity — few users badly hurt. Usually a core capability gap: missing integration, data loss, broken workflow.
- High frequency + high severity — widespread pain. Drop everything.
- Low frequency + low severity — log it and move on.
Name opportunities, not solutions
The output of synthesis should be opportunity statements, not feature specs. "Users need to share results with their boss quickly" opens solution space. "Users want a PDF export button" closes it. The moment you name the solution in your synthesis, you've stopped learning and started building — and the solution you pick first is rarely the best one. Keep the opportunity space open as long as you can.
The synthesis pipeline
Template
Do's and Don'ts
Written with ❤️ by a human (still)