UX Design Fundamentals · 18 min · 120 XP
From notes to insight
Turn a pile of interview notes into a few insights you can defend, each with the evidence beside it.
Six interviews produce sixty pages of notes, and the temptation is to skim them and write down what you remember. What you remember is what surprised you, what matched your hunch, and whatever the last person said. Synthesis is the deliberate version: break the notes into small pieces, find the patterns across people, and say what they mean.
Start by pulling out one observation per note: something a participant did, said or described, in their words, tagged with who said it. “P4 keeps a spreadsheet of which client approved which invoice.” Not “P4 needs approvals”: that's already an interpretation. Then cluster bottom-up: put notes that feel related next to each other, then name each group. This is an affinity map. Naming the groups first (“pricing”, “onboarding”) just sorts the notes into the categories you already believed in.
Observations (clustered) People ────────────────────────────────────────────── ────── Keeps a spreadsheet of who approved what P1 P4 P6 Forwards invoices to a manager by email P2 P4 "I find out it was paid when the client calls" P3 P6 Insight Finance teams track approvals outside the product because the product doesn't record who signed off, so they can't answer "who approved this?" without searching email. (5 of 6 participants)
An insight is more than a theme. “Approvals” is a theme. An insight says what people do, why, and what it costs them, and it's surprising enough that someone could disagree. Write each one with its evidence count (“5 of 6 participants”) and a quote or two. A finding that only one person mentioned can still matter, but say that it's one person.
Finally, turn insights into questions the team can work on. A How might we question (“How might we let finance see who approved an invoice without leaving the app?”) is narrow enough to act on and open enough not to dictate the answer. “How might we add an approvals tab?” has already chosen the solution.
Look for the notes that don't fit. The cluster that contradicts your hypothesis is the most valuable thing on the wall. If every note supports what the team already planned to build, check whether you asked leading questions.
Do it this week: take your notes from three conversations, write one observation per sticky note (or card in a whiteboard tool), and cluster them without naming any group until the end. Write one insight with its evidence count.
Loading your workspace…