Meeting types · Focus group
Focus group minutes, without the note-taker.
Focus group analysis lives or dies on attribution — an unlabelled transcript of eight people is data you cannot code.
A focus group produces the messiest audio in qualitative research and the most attribution-dependent analysis. It matters enormously whether an opinion was the majority view or one insistent participant, whether agreement was independent or followed the loudest voice, and whether a minority position was voiced once and abandoned. All of that is lost in a transcript that reads as one continuous voice.
The traditional answer is expensive: send the recording to a transcription service, wait days, pay per audio minute, and repeat for every group in the study. On a project with eight groups the transcription bill and the delay both shape what the research can afford to be. Turnaround matters as much as cost: if transcripts arrive after the last group has run, nothing learned in group two can shape group three.
GetNotes transcribes the session with automatic speaker labels, so you get separated, attributable text you can take into your coding software the same day rather than the following week. For multilingual fieldwork it handles 90+ languages and code-switching, which matters when participants slip between languages depending on whether they are describing a feeling or a price. Coding can begin while the session is still recent enough for the moderator's own margin notes to mean something.
The consent burden here is heavier than in most meetings, and rightly so. Participants are usually recruited, often paid, and frequently discussing personal experience in front of strangers. They need to know they are being recorded, what happens to the audio, how it will be reported and that they may withdraw — before the session, in writing, not as an afterthought.
Where it earns its keep
One participant dominating
A confident participant sets the tone in the first ten minutes and the rest of the group starts agreeing. Whether the later agreement is independent or conformity is a real analytical question, and you can only answer it from a transcript that shows who spoke, when and after whom.
Eight groups on a fixed budget
A study runs eight focus groups across three cities. Paid per-minute transcription for all of them consumes a serious share of the budget and delays analysis by weeks. Same-day transcripts change how quickly the fieldwork can be adapted between groups.
Participants switching languages
Participants describe practical details in one language and emotional responses in another, sometimes within a sentence. A transcript that garbles the switch destroys precisely the phrasing the study is trying to analyse.
How GetNotes works
- Record or upload. Hit record on your phone or laptop in the room — or drop in an audio file you already have. No bot joins anything.
- It writes it down. Every word transcribed, every speaker labelled, in 90+ languages including ones that switch mid-sentence.
- Share the minutes.Summary, decisions, and action items with owners — ready before everyone's back at their desk.
Questions, answered
How should consent be handled for a focus group?
In writing, before the session, covering that it is recorded, who will hear the audio, how findings will be reported and anonymised, how long the data is kept and the right to withdraw. If your work goes through an ethics committee, their template governs. GetNotes processes the audio to produce your transcript and notes and stores it privately to your account, but the consent process is entirely yours to run.
How accurate is speaker separation with eight people?
GetNotes applies automatic speaker diarization to every recording, which gives you attributed turns rather than an undifferentiated block. Eight strangers with overlapping speech is the hardest condition for any system, so expect to correct some turns during coding. A central microphone position and a moderator who discourages talking over each other both improve the result.
Can I use the transcript in NVivo or similar software?
There is no direct integration with NVivo, Atlas.ti or Dedoose. The speaker-labelled transcript sits on the meeting page in GetNotes as text, so you take it into whatever package you code in yourself. GetNotes is not a qualitative analysis tool — it gives you the attributed transcript and a written summary, which is the input to your coding rather than a substitute for it.
Is the audio used to train anyone's models?
Your recordings are processed for one purpose: producing your transcript and your notes. They sit in private storage with row-level security, so only your account can access them, and deleting a meeting permanently removes the audio and the transcript with it. If your ethics approval commits you to a deletion date, you can actually meet it.
Other meeting types
Your next meeting can write its own minutes.
Record it, and see. Your first meeting is free — thirty minutes, any language, no credit card.
Try your first meeting free