Turn a Stack of User Interviews Into a Thematic Research Report
Stop letting interview recordings pile up in a folder nobody re-opens. Transcribe them consistently, code them for themes, and get a synthesis report with real quotes attached instead of a vague gut feeling about what customers said.
Download workflow config (JSON)How the data moves
This is a handoff plan, not an automatic integration. Review outputs before the next step. Unlabeled arrows show order, not a verified data format. Use the capability links to explore tool substitutions.
Step 1
Record and Transcribe Every Interview the Same Way
Run every interview through Otter.ai, whether it's a Zoom call or an in-person conversation you play back through your laptop mic, so every transcript comes out in the same format. Mixed formats are what make cross-interview analysis painful later.
Step 2
Pull Every Transcript Into One Working Document
Copy all the transcripts into a single document with each one clearly labeled by interviewee, so the next step has full context on who said what instead of one anonymous wall of text.
Tools
No catalog tool specified
Step 3
Have Claude Do a First-Pass Thematic Coding
Prompt Claude with a strict schema: theme name, a one-line definition of that theme, the supporting quotes with which interviewee said each one, and how many interviews mentioned it. Forcing the interviewee attribution into the schema is what makes the next review step possible.
Step 4
Review the Coding Against What You Actually Heard
Read the themed output against your own memory of the interviews, not just the transcripts. A model coding out of context can plausibly misfile a quote under the wrong theme, and the person who actually ran the interviews is the fastest check on that.
Tools
No catalog tool specified
Step 5
Know What This Doesn't Replace
A frequency count across 5-8 interviews is not a statistically meaningful signal, it's a prioritization hint at best. This also won't catch tone, hesitation, or sarcasm the way an experienced qualitative researcher listening live would, and it's not a substitute for one on anything high-stakes like a pricing or positioning decision.
Tools
No catalog tool specified
Time Required
2-3 hours for a batch of 5-8 interviews
Expected Result
A synthesized report that groups raw interview transcripts into named themes, each backed by attributed quotes and a frequency count, instead of a stack of individual transcripts nobody has time to re-read.
Related Workflows
Automate Customer Feedback Analysis
Turn hundreds of survey responses, support tickets, and reviews into actionable product insights, without reading every response manually.
View workflowResearch a Customer Persona from Scratch
Build a research-backed customer persona in 90 minutes, using real community data, not guesswork, to sharpen your product positioning and copy.
View workflowBuild a Research Report with Checked Data and Citations
Assemble a repeatable report from an agreed dataset, web context and relevant research papers, with approval before distribution. Plan around three hours to configure a small working flow and 30 minutes for each draft afterward; data cleanup, source verification and editorial approval need additional time.
View workflow