Use case: user research
User Interview Analysis: Find Themes Across Research Sessions
Analyze user interview notes and transcripts with AI. Group answers across sessions into themes with counts and quotes, and get to synthesis faster.
By Apercu team · Last updated
Synthesizing user interviews is where research projects slow down. After ten or twenty sessions you have pages of notes, and turning them into a clear set of findings takes days of sticky notes and spreadsheets. Apercu groups interview answers into themes across sessions, with counts and quotes, so you can spend that time interpreting instead of sorting.
Why interview synthesis takes so long
- Notes are scattered across documents, transcripts and note-takers.
- Participants describe the same thing differently, so matching ideas across sessions takes careful reading.
- Counting is tedious. "How many participants mentioned this?" means going back through every session.
How to analyze user interviews with Apercu
- Structure the notes. Create a spreadsheet with one row per participant per question: participant ID, segment, question and answer. For unstructured transcripts, split them into passages per topic.
- Analyze one question or topic at a time. Filter the sheet to one question, such as "Walk me through the last time you did X", and upload it.
- Describe the study in the context box, for example "Answers from 18 interviews with operations managers about month-end reporting".
- Review the themes. Each theme shows how many answers it contains and the quotes behind it. Rename themes, merge similar ones and move answers that landed in the wrong group.
- Synthesize with the team. Use the themes as the starting point for a synthesis session, then write insight statements and recommendations.
Example: themes across 18 interviews
An illustrative example for the question "What is the most frustrating part of month-end reporting?":
| Theme | Participants | Example quote |
|---|---|---|
| Manually reconciling data from several tools | 11 of 18 | "I export from three systems and fix the mismatches by hand." |
| Last-minute changes from stakeholders | 7 of 18 | "The numbers change the day before the deadline, every month." |
| No single source of truth | 6 of 18 | "Everyone has their own version of the spreadsheet." |
| Formatting reports for leadership | 4 of 18 | "Half my time goes into making the slides look right." |
Reporting "11 of 18 participants" instead of a percentage is more honest for small samples.
Tips for better interview analysis
- Keep participant IDs so you can count participants, not just mentions.
- Separate observation from interpretation. Code what people said first, then discuss what it means.
- Look for disagreement. Themes where segments differ are often the most useful findings.
- Keep the quotes traceable to a participant and session for your research readout.
Related pages
Frequently asked questions
How do I prepare interview transcripts for Apercu?
Apercu analyzes one text per row, so split transcripts into answers or passages. A common setup is one row per participant per question, with columns for participant ID, question and answer. Then analyze one question, or one topic, at a time.
How many interviews do I need for thematic analysis?
Patterns often start to repeat after 5 to 8 interviews on a focused topic. Apercu is most useful when you have many sessions, or many short answers per session, because grouping and counting them by hand takes the longest.
Can Apercu replace affinity mapping?
It replaces the slow sorting step. Apercu groups similar answers into named clusters, much like a first pass at an affinity map. The team still reviews, renames and interprets the themes together.
Should I use Apercu or a research repository like Dovetail?
Research repositories store and tag recordings, transcripts and insights across many studies. Apercu is a focused tool for grouping and counting text from one dataset. Teams that only need fast synthesis of one study often find a focused tool simpler.
Is participant data safe?
Remove names and identifying details before uploading. Apercu does not use your content to train AI models, and you can delete projects and your account at any time.