User interview distiller
Interviews produce stories; decisions need claims. This skill turns transcripts into claims that carry their evidence, so nobody has to trust a summary.
When to use
- After a round of 5–12 interviews, on the transcripts.
- Before a roadmap meeting where "users want X" is about to be said.
Input
Transcripts (with participant ids), the research questions, and the team's assumptions going in (a list — even a rough one).
Process
- Code first, count second. Read each transcript and tag statements with a short code (e.g.
export-manual,trust-numbers). Codes come from what people say, not from the assumptions. - Claims. For each code seen in ≥ 2 participants, write a claim in neutral language and attach: participant count, and the two most representative quotes with ids.
- Single-voice items are listed separately — real, but not a pattern yet.
- Assumptions cross-check. For each assumption: supported / contradicted / not addressed, with the claim that decides it.
- Surprises. Things nobody expected and no question asked about — quoted.
Rules
- Every claim quotes. A claim without a quote is deleted.
- Count participants, not mentions.
- Never turn a request into a claim about need: "P3 asked for CSV export" is evidence for "P3 wants to work with the data elsewhere", and the claim says the latter with the quote.
- Leading questions in the transcript are flagged, and answers to them are down-weighted.
- Preserve the participants' words; do not clean up their grammar.
Output format
## Claims (8 participants)
**C1 · 6/8 — They rebuild our numbers in a spreadsheet before trusting them.**
- P2: "I export it, sum it myself, and only then do I put it in the deck."
- P7: "the dashboard says one thing, finance says another, so I check"
→ contradicts assumption A3 ("users trust the dashboard totals").
**C2 · 4/8 — Onboarding is done by a colleague, not by the docs.** …
## Single voices
- P5 wants a Slack digest (no one else mentioned notifications).
## Assumptions
| A1 weekly usage | supported (C4) | A3 trust totals | contradicted (C1) | A5 mobile use | not addressed |
## Surprises
- Three participants use the product for a purpose we never mention: … (P1, P4, P8)
License
MIT