folo for UX researchers
Last updated: July 2026
Use cases › UX researchers
A user interview is only as good as what you can pull back out of it — the exact quote, the moment it was said, which participant said it. And it is a recording of a real person who handed you an hour of their time and some honest opinions. With local transcription and summary backends selected globally, folo keeps that processing on your Mac while making the verbatim record easy to synthesize.
Verbatim is the data, not a summary
In research the exact wording is the finding. A verbatim transcript keeps every word as spoken — the hedges, the fillers, the pause before someone admits the feature confused them — and a tidy summary that smooths all that out has thrown the signal away. When you present, a real participant quote carries weight that your paraphrase never will.
With WhisperKit, SpeechAnalyzer or Parakeet selected globally, folo transcribes each session on-device with a full verbatim transcript and who-said-what, so the quotes you lift are the participant’s actual language rather than your memory of it. ElevenLabs Scribe instead uploads meeting audio when selected.
Who said what — participant, moderator, observer
An interview transcript is only useful if it’s attributed. You need the participant separated cleanly from the moderator and from any silent observer or notetaker on the call — and on a longitudinal or diary study you re-interview the same people across weeks. folo’s on-device speaker recognition separates everyone in the session, and Speaker Memory lets you name a voice once so the same participant is recognized across every future session — genuinely useful for a research panel or repeat rounds.
Honestly, diarization is the hard part of transcription everywhere, and heavy crosstalk or a poor mic can still confuse the labels — on-device transcription, explained is candid about where it strains.
From transcript to affinity map
Synthesis is where a study becomes insight, and it runs on quotes. folo lets you ask across every past interview and get answers with clickable, timestamped sources that jump straight to the moment a thing was said — so pulling every mention of a theme across a whole round of sessions doesn’t mean re-listening to ten recordings. Export the verbatim transcript and take the quotes into wherever you actually cluster them: sticky notes and thematic groups in Miro, FigJam or Mural, or a research repository like Dovetail or Condens.
folo isn’t itself an affinity-mapping board or a coding platform — it’s the private, accurate, speaker-labelled transcript that feeds one.
Participant consent and PII
A recording of an identifiable participant is personal data, and under GDPR the participant can withdraw or ask you to erase it. Informed consent should be plain-language, voluntary, and kept separate from any NDA — and you, the researcher, are meant to be the gatekeeper of the raw audio, not hand it to whoever happens to run the transcription.
Selecting local backends so recording and transcript processing stay on your Mac shrinks that problem: participant data lives in fewer places, so honoring a withdrawal or an erasure request means deleting a file you control rather than chasing a vendor’s retention policy. It doesn’t remove the need for consent — you still disclose and get agreement before you record — and none of this is legal advice.
Where folo fits — and where it doesn’t
folo is Mac-only, single-user, and it needs to be running during the session. It is not a shared research repository, a collaborative coding tool, or a participant-recruitment platform — if your team needs a cloud repository where several researchers tag and revisit studies together, something like Dovetail or Condens fits better, and folo vs Granola is honest about where a shared cloud notepad wins. folo’s job is narrower and upstream: with local backends selected, it creates a private, verbatim, speaker-labelled transcript you can take straight into synthesis.
Questions, answered
Can folo give me verbatim participant quotes?
Yes. With a local transcription backend selected, it produces a full verbatim transcript on-device — every word as spoken — with who-said-what, so you can lift the participant’s actual language rather than paraphrasing from memory.
Does it label the participant separately from the moderator?
On-device speaker recognition separates each voice in the session, and Speaker Memory lets you name a voice once so the same participant is recognized across future sessions — useful for panels and repeat rounds. Diarization can still struggle with heavy crosstalk.
Can I get quotes into Miro, FigJam or Dovetail for affinity mapping?
Yes. Export the verbatim transcript into whatever you cluster in — Miro, FigJam, Mural, or a repository like Dovetail or Condens. You can also ask across every past interview and get clickable, timestamped sources that jump to the exact moment a quote was said.
What about participant privacy and consent?
Keeping the recording and transcript on your Mac means participant data lives in fewer places, which makes honoring a withdrawal or erasure request simpler. It doesn’t remove the need for informed, voluntary consent before recording — and this is not legal advice.