Automated transcription has got good enough to be dangerous. It handles a clean one-to-one recording in a quiet room, and falls apart on exactly the material qualitative researchers collect: overlapping speech in focus groups, strong accents, clinical and technical vocabulary, background noise from a ward or a community hall. What comes back reads plausibly and is quietly wrong in the places that matter.
We transcribe with humans in the loop. You choose full verbatim — every stammer, repetition and pause, which you need for conversation analysis or discourse work — or intelligent verbatim, cleaned of filler while every substantive word is preserved, which is what most thematic projects should use.
Formatting is agreed before we start, not after: speaker labels matching your participant IDs, timestamps at whatever interval you want, and anonymisation applied consistently across the set so your ethics conditions are met in the file itself rather than in a later pass you have to do yourself.
What you receive
- Word and PDF transcripts, formatted to your template
- Speaker labels tied to your own participant IDs
- Timestamps at your chosen interval
- Anonymisation and pseudonymisation applied consistently
- Translated transcripts with the source language retained alongside
- Files named and structured ready for NVivo or ATLAS.ti import
How this service runs
- Send the brief — Email us the data type, the volume (audio hours or cases), your method if you have one, and the deadline. No form-filling marathon.
- Scope call & fixed quote — We come back with clarifying questions, then a written scope and a fixed price. NDA signed first if you'd rather talk under one.
- Analysis with checkpoints — You see the codebook or the analysis plan before we run the whole dataset — so the direction is agreed while it's still cheap to change.
- Delivery & walkthrough — Files, project files, syntax and write-up delivered together, plus a call to walk you through the reasoning. Revisions included.
Interviews, focus groups, field recordings, oral histories, and any recording in a language your team does not read.
Services that usually run alongside this one
Coding & codebook development
Inductive, deductive or hybrid coding built into a documented codebook — every code defined, with inclusion and exclusion criteria and an anchor quote you can point to.
Thematic & content analysis
Themes derived from your data with the trail intact — Braun & Clarke's six phases, framework analysis, or quantitative content analysis with frequency counts where you need numbers.
Quantitative & statistical analysis
From cleaning and assumption checks through to the model your design actually calls for — with the syntax file, so a reviewer or supervisor can reproduce every number.
Send us the brief. Get a price.
One email with your data type, volume and deadline is enough for a fixed written quote — back with you within 1 business day, NDA first if you prefer.
Quote for transcription & translation
Email us a paragraph, or fill the brief on the right. Either way a real analyst reads it and replies within 1 business day — no sales sequence.