The problem with traditional transcription
A 60-minute interview takes an average of 4 to 6 hours to transcribe manually. For researchers and journalists this is a huge investment of time, taken away from the work that actually matters: analysis, follow-up questions, and editing.
With AI transcription you can cut this down to a fraction. But there is a difference between "raw AI output" and a "publish-ready transcript". In this article we share the exact workflow we use ourselves.
The 4-step workflow
1. Prepare your recording properly
Quality beats speed. A few tips:
- Use an external microphone, not your laptop microphone
- Choose a quiet room without background noise
- Pronounce names and difficult terms at the start so the AI can learn to recognise them
- Leave short pauses between speakers — it helps speaker diarisation
2. Upload and let the AI do the work
Once you upload, ForgetLess automatically starts:
- Transcription with speaker recognition (Speaker A, B, C…)
- An AI summary of the main points
- Highlighted action items and quotes
- An optional translation if you want to publish in another language
This typically takes 2 to 5 minutes for an hour of audio.
3. Edit using the transcript editor
The AI output is good, but never perfect. Our advice:
- Play the audio back at 1.5× while reading along
- Correct mistakes directly in the text (every change is saved automatically)
- Use "find and replace" for systematic errors (a misspelled name, for example)
- Only remove fillers and "uhm"s when publishing; for research transcripts you usually want them in
4. Export in the right format
For different use cases we offer different exports:
- DOCX — for editing and publication
- PDF — for archiving and sharing
- TXT — for qualitative analysis tools (NVivo, Atlas.ti)
- SRT — for subtitles on video
A real example
A researcher at a Dutch university ran 12 focus groups of 90 minutes each for a qualitative study. Traditionally: about 80 hours of transcription. With ForgetLess: 6 hours (transcription plus editing) — a time saving of over 90 percent.
"The best part is that I now spend more time on analysis instead of transcribing. The quality of my research has improved as a result."
How to get started
Create a free account at /registreren and upload your first recording. Our free tier gives you 60 minutes of transcription per month — more than enough to test this workflow.