Set up
Try a plugin
The Design plugin ships with /research-synthesis and other research and spec skills as a starting point, already structured to cluster observations and pull supporting quotes. If your admin manages plugins and it's not available yet, skip this step. The prompt below works without it.
/research-synthesisCluster interview transcripts into themed findings with supporting quotes/design-handoffTurn a design into a developer handoff specConnect your tools
Claude is more powerful when it works directly with your systems. You control permissions and access. Learn about tool access(opens in new tab).
Add a connector(opens in new tab) for each tool you want Claude to use.
Set your working folder
Drag the files you'll use (the raw transcripts, your discussion guide, the participant roster) into one folder on your machine, then give Claude access to it(opens in new tab) in the Claude desktop app. Claude reads from it and writes the themed findings doc back to it. If you run studies regularly, create a project(opens in new tab) from that folder so your guide, taxonomy, and memory stay attached.
The prompt
Copy this into Claude
Synthesize the interview transcripts into a research findings doc. Cluster the observations into 5-7 themes, name each plainly, support each with 2-3 verbatim quotes attributed to participant ID, and end with the open questions the team still needs to answer.
Why this works
Give a number range. Five to seven forces real clustering, not a laundry list.
Ask for attributed evidence. Evidence stays in the participant's words, attributed by ID.
End on open questions. The doc tells the team what to study next.
Let the folder supply the context. Every transcript is read; nothing depends on memory.
Get a better draft
Add an example to match. Drop an example you like into the folder and Claude matches your structure and voice.
Ask it to flag uncertainty. Add "flag anything you're not confident about" so you know where to look first when you review the draft.
Make Claude work for you
Tell Claude your own rules for this task, such as your taxonomy and your quote format, then save what you've done as a skill of your own. A few minutes of conversation and it runs with your standards from then on.
Make what we've done in this task so far into a skill called /interview-findings.
Review the skill, and save it if Claude hasn't already.
Make it repeatable
Share it as an artifact
A findings doc goes stale once the next round of interviews comes in. Ask Claude to publish the themes as an artifact and the product team has one link — re-run the skill or schedule it to refresh.
Publish those themes as an artifact for the product team. Keep a "what shifted since last round" note at the top.
Run it after every interview day
Sessions wrap, the synthesis should already be cooking. Ask Claude to set up a scheduled task, or open Scheduled in the sidebar, and set /interview-findings to run each evening, picking up any new transcripts in the study folder.
Set up a scheduled task for weekdays at 6pm: check Research/Onboarding-Study-Q2 for new transcripts and re-run /interview-findings, updating the findings doc and noting which themes moved.
Each weekday at 6pm, checks the study folder for new transcripts, re-runs /interview-findings, and updates the findings doc with a note on which themes moved.
To use a folder on your computer during a scheduled task, keep your computer on and awake with the Claude desktop app open so Claude can access it.
Share with your teammates
Your /interview-findings skill now carries your taxonomy and your quote format. Share it so every researcher on the team clusters the same way, and PMs see a consistent findings doc no matter who ran the study.