Synthesis
How a pile of evidence becomes a ranked list of the problems that matter.
Synthesis is the step that turns “we have a lot of feedback” into “here is what to work on, in order.” It reads your evidence, groups related signals into problem clusters, scores each cluster, and ranks them.
How problems are scored
Each problem gets a composite score from real, data-derived signals — not vibes:
- Frequency — how often the problem shows up across evidence.
- Severity — how badly it hurts when it happens.
- Revenue impact — its connection to money (churn, lost sales, expansion).
- Strategic fit — how well solving it aligns with your goals.
The noise gate
A high vote count alone doesn’t win. Synthesis applies a floor: if a problem scores very low on severity or strategic fit, its overall score is capped — so a popular-but-trivial request can’t outrank a genuinely important problem.
“Add dark mode” had 892 votes and still ranked last
Priya, PM on a developer-tools product
The situation. Priya imports a year of feedback. By raw volume, “add dark mode to the CLI” is the runaway winner — 892 upvotes. But two enterprise accounts worth $400k ARR have flagged API reliability as a reason they might churn, mentioned in only a handful of tickets and one QBR note.
What they do
- 1She runs synthesis over the full evidence set.
- 2Dark mode scores high on frequency — but near-zero on severity and revenue impact, because no ticket ties it to churn or activation.
- 3The noise-gate floor caps its composite score.
- 4API reliability scores lower on frequency but high on severity, revenue impact, and strategic fit — so it rises to the top.
The result. API reliability ranks #1, dark mode ranks last. Priya spends her sprint on the thing that actually protects $400k of revenue — and has the evidence to explain the call in her roadmap review.
Lesson: frequency ≠ importance
- What it is
- The composite score blends four signals — frequency, severity, revenue impact, and strategic fit — and a floor prevents any one of them (usually frequency) from dominating.
- Why it matters
- The single most common prioritization mistake is building the most-requested thing. Volume is easy to measure, so it quietly takes over. The floor is there to stop that failure mode by design.
When something ranks lower than you expected, open its scores — a low severity or strategic-fit number is usually why, and it’s usually right.
Emerging signals
Synthesis also surfaces emerging signals — thin, single-source clues that haven’t been corroborated yet. They’re your early-warning system: worth watching before they become big problems.
Every ranked problem links back to the exact evidence it came from. Click through any time to see why it’s ranked where it is.