By name: discovery-evidence-synthesis description: Use when a product manager wants to synthesize customer interviews, user research, support tickets, sales notes, analytics, surveys, feedback, win-loss notes, or usage evidence into themes, opportunities, confidence levels, product decisions, and next research actions.
Synthesize Customer and Product Evidence Into Decisions
Goal
Turn messy product evidence into decision-grade synthesis. The output should show what the evidence supports, what it does not support, and what the PM should do next.
Inputs to request when missing
Ask for context that changes synthesis:
- Decision the evidence should support.
- Target segment, user, buyer, workflow, and geography.
- Evidence sources, sample size, recency, and collection method.
- Current hypothesis or product bet.
- Known metrics, baselines, or counter-evidence.
If unavailable, synthesize what exists and label evidence limits clearly.
Senior PM standard
Synthesis should determine what decision can be made now, not just summarize themes.
Look for:
- Segment-specific signal instead of blended averages.
- Contradictions between qualitative and quantitative evidence.
- Bias from sales pressure, support noise, stakeholder preference, recency, or selection.
- Whether evidence proves problem severity, solution desirability, willingness to pay, usability, retention risk, or only interest.
- Decision thresholds: enough evidence to build, test, narrow, stop, or research further.
Evidence lenses
Classify evidence by:
- Source: interview, support ticket, analytics, sales note, survey, usability study, call transcript, review, internal stakeholder.
- Segment: customer type, role, market, plan, lifecycle stage, geography, platform.
- Strength: direct observation, self-report, proxy metric, anecdote, stakeholder opinion.
- Recency: current, stale, unknown.
- Decision relevance: validates problem, validates solution, reveals risk, indicates demand, indicates churn, indicates usability issue.
Workflow
- Inventory the evidence sources and note gaps.
- Normalize statements into observations, not conclusions.
- Cluster observations into themes.
- Identify jobs, pains, triggers, current alternatives, and desired outcomes.
- Separate signal from noise:
- Frequency is not automatically importance.
- Loud customers are not automatically representative.
- Sales demand is not automatically product demand.
- Internal opinions are not customer evidence.
- Assign confidence to each theme: high, medium, low.
- Convert synthesis into product implications, decisions, and research actions.
- State what evidence would change the recommendation.
Output format
Summary
Lead with what the evidence most strongly supports.
Evidence Map
Use a table:
| Theme | Evidence | Segment | Confidence | Product implication |
|---|
Opportunities
List opportunities in a clear format:
- Customer/job.
- Pain or unmet need.
- Evidence.
- Potential solution direction.
- Confidence.
Anti-Evidence and Gaps
List evidence that contradicts the main themes and what is missing.
Decisions Supported
State which decisions can be made now and which need more evidence.
Decision Thresholds
State whether the evidence supports build, test, narrow, stop, or research more, and why.
Next Research
List the smallest set of follow-up questions, interviews, analysis, or experiments.
Quality bar
- Do not overgeneralize from weak evidence.
- Preserve nuance across segments.
- Call out contradictions.
- Keep synthesis tied to decisions.
- Avoid treating frequency as importance without severity or business impact.