I do not expect a decision to emerge automatically when the research folder fills up. Ten interviews, a hundred comments and several competitor reviews can still leave the team unsure what it has learned. Synthesis means establishing useful relationships, not merely reducing the volume of material.
Keep each observation connected to its context
When moving a quotation into a working note, I avoid duplicating unnecessary identifying information but retain the context. Which need was being discussed? Had the person used the product? How was the question asked? A price objection based on experience differs from one formed at first glance.
Source references make it possible to revisit a theme and check an interpretation. Anonymising a quotation should not mean stripping away everything that explains it. Preserve the context needed for the decision without spreading personal details that serve no purpose in the analysis.
Look for contradictions as well as patterns
I group observations by needs, concerns, selection criteria and current alternatives. Then I examine differences within each group. Two people can both ask for something easy while meaning different things: fewer steps for one, less mental effort for another. Shared vocabulary does not always establish a shared need.
An observation that differs from the majority may reveal another use case or audience. Conversely, a striking quotation should not become the central insight just because it makes a compelling presentation slide. The limitations of the sample need to remain visible in the recommendation.
Write an insight as an explanation
“People want simplicity” names a theme. A more useful synthesis describes a situation and tension: first-time buyers may want clearer comparisons more than fewer options. This is an illustrative example, but it leads to a different decision. Improving comparison content may address the issue without reducing the range.
Under each proposed insight, I keep supporting observations, counterexamples and unresolved questions. The team can then debate the interpretation while seeing what it rests on. The value of audience research lies partly in preserving those boundaries as information becomes a recommendation.
Pair the recommendation with a next check
For each proposed action, I ask what should change, why and how we will assess it. A comparison problem might justify a small content revision followed by another comprehension check. Not every finding requires a broad repositioning exercise.
Message testing continues this loop. An interpretation becomes copy; the way people understand that copy produces another observation. I use synthesis as the basis for a more informed decision, not as a final statement that cannot be questioned. That approach keeps the work useful when new evidence appears.