The billable-hours arithmetic
No typing tier converts speed to money as directly as transcription. Conversational audio runs 130–160 spoken words per minute; a 70 WPM typist therefore spends roughly two to two-and-a-half hours per hour of audio (typing time plus replays and corrections), while a 90 WPM typist compresses the same job toward the 1.5–2 hour ratio that makes per-audio-minute pay rates work. That’s why postings state floors where they do: below ~60–70 WPM the ratio breaks and the work stops paying; above it, every additional WPM is margin. It’s also why sustained speed is what matters — transcription is a 3–5 minute-mode skill, never a sprint.
What the tiers ask
| Tier | Posted range | What else is tested |
|---|---|---|
| General transcription | 60–75 WPM floor | Audio-handling, formatting, grammar |
| Experienced general | 80–100 WPM | Turnaround-time commitments |
| Medical transcription | Listings at 80–100 | Clinical vocabulary, formatting standards, HIPAA compliance |
| Legal transcription | 65–80+ with 98% accuracy | Citation formats, terminology; precision weighted like speed |
Test at transcription length — the 5-minute standard
Click the words and start typing — the timer starts on your first keystroke. space next word · Esc restart
The skill nobody posts: typing while listening
Transcription’s real test is divided attention — producing accurate text while parsing ongoing audio, which is measurably harder than copying visible words. Practical implication: your copy-test WPM overstates your transcription WPM by 10–20% until the listening skill is trained. If you’re targeting this field at, say, a solid 64 (context here), plan the climb to 70+ and practice against audio, not just test screens. Domain vocabulary (medical, legal) is the third leg, and postings test it separately.
Is the climb worth it?
Uniquely, yes-if: transcription is the tier where the pay premium reliably tracks the speed requirement, so the 60→80 project has a defined return. It’s also the natural stepping stone toward captioning and, at the far end, court reporting — where the numbers change so much they need their own page.
Sources
- Posted ranges: aggregated requirement text from live transcription listings and published role guides (2026); medical/legal tiers per listed requirements.
- Spoken-speech rate: standard conversational English measurement, 130–160 WPM.