Home › Guides › How to transcribe an interview
There are three ways to turn a recorded interview into text: use a transcription app (fastest), do it by hand (most control), or hire a human service (most expensive). This guide covers all three, plus how to get a clean transcript, how to handle consent, what to do if the interview runs long, and how long it actually takes.
For most people, a transcription app is the right answer — a draft in minutes instead of hours. Here is the flow with Attesta on iPhone:
Because every line is tappable, you can jump straight to the moment a quote was said instead of scrubbing through audio. And nothing you record is ever used to train an AI model.
A single Attesta recording runs up to 60 minutes. If your interview goes long, it does not get cut off mid-sentence or lose anything — it stops the same way tapping Stop yourself would, and you still get the full transcript and summary for everything captured so far. For a two-hour, oral-history-style interview, plan on stopping and starting a second recording once you are past the first hour.
If you need a break partway through — a phone rings, you refill coffee — pause instead of stopping. Pausing keeps everything in one recording and one transcript, and the paused time does not count against the 60-minute limit, so an interview with a couple of breaks still fits comfortably in a single recording.
Not every interview is one-on-one. Panel interviews, roundtables, and group discussions work the same way — Attesta's speaker labels are not limited to two voices, so a transcript with several participants still comes back split by who said what. See how speaker labels work for more on how that identification happens.
The transcript itself — the full, speaker-labeled text — is read right in the app. It's there whenever you need to check exact wording, and tapping any line jumps playback to that moment — but there is no copy or export action on the transcript text today; treat it as your reference copy, not a file to hand off. What you can export and share is the AI summary: once it's ready, export it as a PDF straight from the recording — the summary, key topics, and quotes in one file — and send it from your iPhone's share sheet: email, Messages, Slack, AirDrop, wherever you need it. For a full transcript in an article or a research write-up, quote from what you read in-app.
If you interview people regularly, the real problem eventually isn't recording — it's finding the one thing someone said three months ago. Attesta's library search checks the full text of every transcript in your account, not just the titles, so a quote you half-remember ("something about the Q3 budget") turns up even when you cannot recall which interview it was in. It's a Pro feature, and it searches your whole history, not just the recordings currently loaded on screen.
If you want full control over every word, transcribe manually:
Manual transcription is accurate but slow — plan on four to six minutes of work per minute of audio. Many people record with an app, let it produce the draft, then hand-correct, which is far faster than starting from a blank page.
Transcribing an interview is fine — as long as you have consent to record it. Laws vary: some places only need one party to agree, others require everyone. The simplest rule is to always tell your interviewee you are recording. Attesta makes that automatic with an audible tone at the start of every recording, so consent is never ambiguous.
By hand, a clean 1-hour interview takes about 4–6 hours; a 2-hour interview, 8–12 hours. With a transcription app, the first draft of either is ready in a few minutes, and your only real work is light correction. That time saving is the whole reason apps exist.
Record an interview, get a speaker-labeled transcript plus a summary and quotes — with an audible consent tone, and never trained on your audio.
Download on theApp StoreRecord it clearly, then run the audio through a transcription app for a fast draft, or type it out manually in short segments. An app like Attesta gives you a speaker-labeled transcript plus a summary automatically.
Not on its own — it needs a speech-to-text step first, and it does not handle speaker labels or timestamps. A dedicated transcription app does all of that in one pass.
Yes, with consent to record. Recording laws vary by region; in two-party-consent areas everyone must agree. Always tell your interviewee — Attesta’s audible tone makes it clear.
By hand, roughly 8–12 hours. With an app, a few minutes for the first draft, then light correction.
Yes — a single Attesta recording runs up to 60 minutes, then stops cleanly (same as tapping Stop yourself), keeping the transcript for everything captured. Pause during breaks instead of stopping; paused time does not count against the limit. For interviews that run well past an hour, start a second recording once the first stops.
Yes — speaker labels are not limited to two voices, so a roundtable or panel with several participants still comes back split by who said what.
Not the transcript text itself — it's read-only in the app today, with no copy or share action on it. What you can export is the AI summary: tap export for a PDF with the summary, key topics, and quotes, then send it from your iPhone's share sheet.
Yes, on Pro — Attesta's library search checks the full text of every transcript you've recorded, not just the title, so a quote or topic turns up even when you don't remember which interview it came from.