{"schemaVersion":"1.0","source":"The Workflow Finder","sourceUrl":"https://www.theworkflowfinder.com","license":"Free to use with attribution and a link to the cited page","generated":"2026-08-28","workflow":{"title":"Turn a Stack of User Interviews Into a Thematic Research Report","slug":"turn-user-interviews-into-a-thematic-research-report","url":"https://www.theworkflowfinder.com/workflows/turn-user-interviews-into-a-thematic-research-report","description":"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.","difficulty":"Intermediate","timeEstimate":"2-3 hours for a batch of 5-8 interviews","schemaDuration":"PT3H","expectedResult":"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.","tags":["Research","Customer Research","Product"]},"steps":[{"position":1,"title":"Record and Transcribe Every Interview the Same Way","instructions":"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.","input":"Workflow inputs: the topic, goal, or raw materials the user is starting from.","output":"Handoff artifact consumed by Step 2 (\"Pull Every Transcript Into One Working Document\").","ioProvenance":{"input":"structural","output":"structural"},"toolSlots":[{"slug":"otter-ai","name":"Otter.ai","url":"https://www.theworkflowfinder.com/tools/otter-ai","category":"Audio","pricing":"freemium","capabilities":["Meeting transcription & summarization"]}]},{"position":2,"title":"Pull Every Transcript Into One Working Document","instructions":"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.","input":"Output of Step 1 (\"Record and Transcribe Every Interview the Same Way\").","output":"Handoff artifact consumed by Step 3 (\"Have Claude Do a First-Pass Thematic Coding\").","ioProvenance":{"input":"structural","output":"structural"},"toolSlots":[]},{"position":3,"title":"Have Claude Do a First-Pass Thematic Coding","instructions":"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.","input":"Output of Step 2 (\"Pull Every Transcript Into One Working Document\").","output":"Handoff artifact consumed by Step 4 (\"Review the Coding Against What You Actually Heard\").","ioProvenance":{"input":"structural","output":"structural"},"toolSlots":[{"slug":"claude","name":"Claude","url":"https://www.theworkflowfinder.com/tools/claude","category":"Writing","pricing":"freemium","capabilities":["Long-form writing","Code generation","Web research"],"hasApi":true}]},{"position":4,"title":"Review the Coding Against What You Actually Heard","instructions":"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.","input":"Output of Step 3 (\"Have Claude Do a First-Pass Thematic Coding\").","output":"Handoff artifact consumed by Step 5 (\"Know What This Doesn't Replace\").","ioProvenance":{"input":"structural","output":"structural"},"toolSlots":[]},{"position":5,"title":"Know What This Doesn't Replace","instructions":"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.","input":"Output of Step 4 (\"Review the Coding Against What You Actually Heard\").","output":"Final workflow output: 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.","ioProvenance":{"input":"structural","output":"structural"},"toolSlots":[]}]}