Study first
Review the ideas behind the questions
Review the evidence before deciding what the customer insight proves. Strong synthesis keeps observations, themes, metrics, limits, and next actions separate enough for a team to trust.
Match Evidence To The Decision
Mixed-method research is useful because different sources answer different questions. The synthesis gets weaker when one source is forced to answer a question it was not built to answer.
- Qualitative methods fit why and how-to-fix questions, while quantitative methods fit how-many and how-much questions.
- Natural-use research gives more realistic behavior but less control over exactly what the researcher learns.
- Research methods should answer the team's important questions with strong, reliable evidence.
In Practice
Keep The Question Visible
In a synthesis review, write the business decision beside each evidence source so the team can see which source answers which part.
Use Tension Productively
When analytics and interviews disagree, do not pick a favorite. Ask what each method can support and what follow-up would resolve the decision.
Common mistakes
Letting one metric explain the reason behind a customer behavior.
Use the metric to locate or size the pattern, then use a method that can explain why it happens if the decision needs cause.
Q&A
What should happen when research sources point in different directions?
Keep both sources visible, check what question each one answers, and choose the next action that resolves the decision risk.
Why not use interviews to report prevalence?
Interviews can reveal causes and language, but prevalence needs a method designed for how many or how much.
Separate Observations From Findings
Analysis should move from raw notes to patterns and then to findings. The team should be able to trace a finding back to what people actually did or said.
- Observations should capture exactly what was seen or heard, not what the observer thinks it means.
- Group analysis can reduce researcher bias and the influence of individual team members or stakeholders.
- Findings should summarize what the observations show after the team reviews the grouped evidence.
In Practice
Trace The Finding
Before a recommendation goes out, check that the finding can be traced to observations, quotes, behaviors, or grouped evidence.
Do Not Crown A Lone Note
An isolated quote can be useful, but it should be labeled as a signal or follow-up question unless the grouped evidence supports a finding.
Common mistakes
Writing stakeholder interpretation as if it were a participant observation.
Rewrite the note as what the participant did or said, then discuss interpretation during analysis.
Q&A
What makes a finding stronger than a quote?
A finding summarizes a reviewed pattern in the evidence; a quote is support or illustration unless the broader analysis backs it.
Why include observers in analysis?
Observers can add context and challenge one person's interpretation, which reduces bias and stakeholder overreach.
Hand Off Evidence With Limits
Research is useful when the team can act on it. A good handoff explains the finding, why it matters, what is still uncertain, and what action or follow-up fits.
- Research findings should help the team make design decisions, prioritize work, refine needs, and develop the roadmap.
- Findings should be shared soon so the team can use them while decisions are still open.
- Discovery research should examine current experience, existing data, previous research, and what users need before planning or building.
In Practice
Write The Action Boundary
Pair every recommendation with the evidence strength: act now, test a design, check another segment, or run follow-up research.
Share Gaps, Not Just Answers
If a customer segment, channel, or behavior was not covered, say that in the handoff so the next decision does not overreach.
Common mistakes
Waiting until the roadmap is locked before sharing the research.
Share findings while they can still inform prioritization, design decisions, user needs, and the roadmap.
Q&A
What should a readout include when evidence is incomplete?
Include the finding, evidence, consequence, open question, and the smallest next action that can reduce the decision risk.
When should discovery evidence stop a solution plan?
Stop or slow the plan when the team has not yet understood users, current behavior, problems, needs, and existing evidence.