Pepys

Blog

How Long Does It Take to Transcribe a Research Study by Hand?

A full qualitative study runs 120 to 210 hours of hand-transcription. Here's the math, from interview count to working weeks.

By Pepys ·

The short version

A full qualitative study takes 120 to 210 hours to transcribe by hand, about 3 to 5 working weeks. That is a 30-interview study at one audio hour each, multiplied by the 4 to 7 hours of typing per audio hour reported by the [University of Bath](https://library.bath.ac.uk/research-data/working-with-data/transcription).

How long does hand-transcribing one audio hour take?

Plan on 4 to 7 hours of typing for every hour of audio, according to the University of Bath research-data guide. Peer-reviewed work sets a wider band: Bailey (2008) in Family Practice reports at least 3 hours per audio hour, rising to 10 hours for fine verbatim detail.

Detail level drives that spread. A clean transcript, with filler words removed, sits near the bottom of the range. A fine-grained version that marks pauses, overlaps and nonverbal cues climbs toward Bailey's 10-hour figure. Audio quality, accents and the number of speakers all add time on top.

Published ranges per single audio hour: - 3 hours: fast, low-detail floor (Bailey, 2008) - 4 to 7 hours: typical academic range (University of Bath) - Up to 10 hours: fine verbatim with visual detail (Bailey, 2008)

Working from a single recording rather than a whole project? See our breakdown of how long it takes to transcribe an hour of audio for the per-file view.

How many interviews go into one research study?

A typical qualitative interview study runs a median of 15 to 31 interviews, based on Vasileiou et al. (2018), which analyzed 214 published studies. Per-journal medians landed at 31 interviews (BMJ), 30.5 (Sociology of Health & Illness) and 15 (British Journal of Health Psychology). Assume one hour each (interview length varies by type) and that is 15 to 31 audio hours.

Could you interview fewer people? Some researchers stop once themes repeat. Guest, Bunce & Johnson (2006) found thematic saturation within the first twelve interviews of sixty. Hennink, Kaiser & Marconi (2017) drew a sharper line: code saturation at nine interviews, yet 16 to 24 interviews for full meaning saturation.

Here's the catch. Even a lean, saturation-driven study of 16 to 24 interviews still leaves you 16 to 24 audio hours to transcribe. Cutting interviews trims the interviewing, not the typing that follows. The transcription workload scales with every recorded hour you keep.

See our full breakdown of how many hours of audio a qualitative study contains to map your own project.

The whole-study math: hundreds of person-hours

Do the arithmetic and a full study runs into the hundreds of person-hours. Take a 30-interview study at one audio hour each: 30 audio hours times the 4 to 7 hours of typing per audio hour from the University of Bath equals roughly 120 to 210 person-hours. Treat this as an illustrative calculation, not a single published figure.

How it scales by study size (one audio hour per interview, 4 to 7 hours of typing per audio hour): - 15 interviews: 60 to 105 person-hours - 24 interviews: 96 to 168 person-hours - 31 interviews: 124 to 217 person-hours

Use Bailey's wider band and the totals stretch further. At 3 to 10 hours per audio hour, that same 30-interview study spans 90 to 300 person-hours. The exact number depends on detail level, audio quality and speaker count, but the order of magnitude holds: this is weeks of full-time work.

Splitting long recordings into manageable pieces helps. See how to transcribe long audio for a chunking workflow.

From person-hours to working weeks and days

Convert person-hours to a calendar and a full study eats 3 to 5 working weeks. The U.S. standard workweek is 40 hours (29 CFR 778.101). A 30-interview study at 120 to 210 person-hours therefore fills about 3 to 5.3 weeks, or roughly 15 to 26 eight-hour working days.

Treat these as illustrative conversions, not published totals. The 40-hour week is the federal maximum-nonovertime standard; splitting it into an 8-hour day is a common convention. One person rarely transcribes eight focused hours a day, so real calendars stretch longer once you add breaks, review and correction passes.

One of the biggest variables is you. Outsource to a professional service and the calendar shrinks, but the person-hours do not: someone still types every recorded minute. That labor is why transcription, not interviewing, is often the longest phase of qualitative fieldwork.

How can you cut the transcription time down?

Automatic transcription flips the ratio. Instead of the 4 to 7 hours of manual typing per audio hour reported by the University of Bath, software like Pepys drafts a full transcript in minutes. You then spend your hours checking and correcting, not typing from a blank page. The speech-recognition research is public, including the OpenAI Whisper paper (Radford et al., 2022).

Budget for the human pass anyway. Automated drafts still need correction for accents, crosstalk, jargon and speaker labels, the same things that pushed manual verbatim toward Bailey's 10-hour figure. The win is sequence: you edit a draft instead of building one, which is where the hours disappear.

Two habits protect accuracy. First, follow a consistent method: our guide to qualitative research transcription covers verbatim conventions and speaker labels. Second, plan the next step early, because analyzing interview transcripts is faster when the transcript is clean and searchable from the start.

Questions, answered

How long does it take to transcribe one hour of audio by hand?

Plan on 4 to 7 hours per audio hour, according to the University of Bath. Peer-reviewed Bailey (2008) reports a wider band: at least 3 hours per hour of talk, rising to 10 hours for fine verbatim with visual detail. Detail level and audio quality decide where you land.

How many interviews are in a typical qualitative study?

A median of 15 to 31 interviews, per Vasileiou et al. (2018), which reviewed 214 published studies. Per-journal medians were 31 (BMJ), 30.5 (Sociology of Health & Illness) and 15 (British Journal of Health Psychology). Saturation research suggests 9 to 24 interviews often carry the themes.

How many hours does it take to transcribe a whole 30-interview study?

Roughly 120 to 210 person-hours as an illustrative calculation: 30 one-hour interviews times the University of Bath's 4 to 7 hours of typing per audio hour. Using Bailey's wider 3 to 10 hour band, the same study spans about 90 to 300 person-hours.

How many working weeks is transcribing a research study?

About 3 to 5 working weeks for a 30-interview study, dividing 120 to 210 person-hours by the 40-hour U.S. standard workweek (29 CFR 778.101). That is roughly 15 to 26 eight-hour working days for one person, before breaks and correction passes stretch it out.

Does automatic transcription make it faster?

Yes. Automatic speech recognition drafts a transcript in minutes per audio hour instead of the 4 to 7 hours manual typing takes (University of Bath). You still budget time to correct accents, crosstalk and speaker labels, but you edit a draft rather than typing from scratch.

References

  1. 1.Bailey (2008), 'First steps in qualitative data analysis: transcribing', Family Practice 25(2), 127-131Oxford University Press / Family Practice
  2. 2.University of Bath Library, 'Transcription' research-data guideUniversity of Bath
  3. 3.Vasileiou, Barnett, Thorpe & Young (2018), BMC Medical Research Methodology 18:148BMC Medical Research Methodology
  4. 4.Guest, Bunce & Johnson (2006), 'How Many Interviews Are Enough?', Field Methods 18(1)Field Methods (SAGE)
  5. 5.Hennink, Kaiser & Marconi (2017), 'Code Saturation Versus Meaning Saturation', Qualitative Health ResearchQualitative Health Research (SAGE)
  6. 6.29 CFR 778.101, Maximum Nonovertime Hours StandardU.S. Code of Federal Regulations via Cornell LII
  7. 7.OpenAI Whisper paper (Radford et al., 2022)arXiv (OpenAI)

Keep reading

Don't just take our word for it.

Ask ChatGPT, Claude, or Perplexity what Pepys is and who it's for. One click, and your favorite AI does the homework.

Built for the four complaints above

Any length, pay once, real exports, no training on your audio. Start free with 60 minutes, no card.