Attribution
Every voice labelled
Participants are separated as they speak and renameable on the spot, so “P4 disagreed” is in the transcript, not in your memory.
Six voices, one table, a transcript that keeps them apart.
Run the group and watch every contribution land under the right name — so the moderator can moderate, and the analysis starts from who actually said what.
- Speaker-labelled, as you speak
- Holds up in a loud room
- Audio is never kept after a session
Focus groups are the hardest room in transcription
Six or eight participants, a moderator steering, people agreeing over the top of each other — and the whole value of the session is knowing which reaction came from whom. A transcript that merges everyone into one voice has thrown away the finding.
NoiseScribe separates and labels up to twelve voices as they speak, and you can rename them mid-session as you learn who’s who.
What you get
Used by market researchers, UX teams and academic researchers.
Crosstalk
Interruptions survive
The moment two participants react at once is usually the moment that matters — it’s built to hold up there, not fall apart.
Analysis
Timestamped for coding
Every line carries a timestamp, so pulling verbatims into your analysis grid or checking a reaction against the audio takes seconds.
Questions
How many participants can it tell apart?
Up to twelve voices are separated and labelled in a session — comfortably beyond a standard six-to-eight-person group plus a moderator.
Can I name the participants during the session?
Yes. Speakers can be renamed mid-session, so by the first round of introductions the transcript is already showing real names or participant codes.
Can I get the transcript into my analysis tool?
Exports include TXT and DOCX for reading, and JSON with word-level timings if your analysis workflow wants structured data.
Further reading
Also good for