Sessions and their files
User research produces a mix of recordings. Remote usability tests and interviews on Zoom, Google Meet or Teams leave an MP4, often with a separate audio file. Screen recordings from a Mac are usually MOV, phone screen recordings are MP4 or MOV, and in-person sessions may be filmed on a phone or recorded on a voice recorder.
Upload the video as it is. Audio is extracted on our side and the timestamps match the original video, so 12:40 in the transcript is 12:40 in the screen recording, where you can see what the participant was clicking when they said it.
Running sessions that transcribe well
A few habits make the transcripts more useful:
- Ask remote participants to use headphones so your voice does not echo back through their speakers.
- Remind people to keep thinking aloud; silent stretches leave gaps in the transcript and in your evidence.
- Read task instructions in the same words in every session. It makes tasks easy to find in each transcript.
- Say the participant code, not their full name, at the start of the recording.
- Turn off notification sounds on the test device; they end up in the audio.
From transcript to findings
Use the synced player to review: click a line and the audio jumps there, and the timestamp tells you where to look in the video. Fix product names, feature names and UI labels, which participants often say differently from how you write them.
Speaker identification, on paid plans or with a Minute Pack, separates facilitator and participant; rename them to "Facilitator" and the participant code once. Note-takers who joined the call should stay quiet or be counted as speakers.
Export CSV to get one row per segment with its start time. That is a quick way to build an affinity map or a research repository: paste the rows into a spreadsheet or whiteboard tool, tag them, and keep the timestamp so every sticky note links back to the session. DOCX or TXT suit a session write-up, and SRT or VTT let you caption a highlight clip for a stakeholder presentation.
Consent and participant privacy
Participants should know the session is recorded and how the recording will be handled, including that it is transcribed by an online service. Remove names, email addresses and anything else identifying before transcripts go into a shared repository. Files are never used to train AI models, and you can delete recordings and transcripts once the study is done.
Across a study
Keep one folder per study, upload sessions as they finish, and bulk export at the end. Sessions in other languages are transcribed in that language; use the Google Translate link to read them, and have quotes for reports checked by a fluent colleague. For academic interview studies, see the research interview transcription page, and for groups, the focus group page.