SharpAssessmentGet started

explainer

What is a good typing speed by job?

Typing speed benchmarks by role, why accuracy beats raw WPM for most jobs, and how to set a threshold for a job posting without excluding good candidates.


Most published typing benchmarks are either vague or borrowed from somewhere nobody names. Here is a straightforward version, with the caveat stated up front: speed matters less than the numbers suggest, and accuracy matters more.

Rough benchmarks

Speed What it means in practice
Under 25 WPM Hunt and peck. Fine for a job where typing is incidental
25 to 39 WPM Below average but workable outside typing-heavy roles
40 to 54 WPM Typical for someone comfortable with a keyboard
55 to 74 WPM Above average. Sufficient for essentially any office role
75 to 94 WPM Fast. Rarely required outside transcription and captioning
95 WPM and up Unusual, and almost never the binding constraint on the work

By role

Typing is incidental (managers, field roles, most trades): no threshold worth setting. Someone at 30 WPM will not be slower at the job in any measurable way.

Typing is regular but not the job (customer support, administrative, coordinator roles): 40 to 50 WPM with high accuracy. Support work in particular rewards accuracy and clear writing far more than speed, because the time cost of a confusing reply dwarfs the seconds saved typing it.

Typing is the job (data entry, transcription, medical records): 50 to 70 WPM, and accuracy is the real requirement. Transcription is the one field where 80 WPM and above is a genuine advantage.

Numeric entry (bookkeeping, payroll, claims): WPM is close to irrelevant. Measure keystrokes per hour on a numeric keypad and error rate instead. These are different skills and a fast prose typist is not automatically a fast 10-key operator.

Why accuracy wins

Take two candidates over an hour of data entry:

  • 65 WPM at 92 percent accuracy
  • 45 WPM at 99.5 percent accuracy

The faster one enters more characters and produces roughly eight times as many errors. If each error takes a minute to find and fix, and some are never found at all, the slower candidate is worth more. In records work, where an unnoticed wrong digit propagates, they are worth much more.

This is why net WPM, which subtracts uncorrected errors, is the number to ask for. Gross speed alone rewards exactly the behaviour you do not want.

Setting a threshold you can defend

If speed genuinely matters for the role, three rules keep the requirement useful:

  1. Set it from the work, not from ambition. Time someone competent doing the actual task and work backwards.
  2. Test the format the job uses. A prose test does not predict form or numeric entry speed. If the job is forms, test forms.
  3. Verify it or drop it. An unverified requirement in a posting filters out honest candidates and lets confident ones through, which is the opposite of what you wanted.

And test for long enough to matter. One minute measures a burst; five minutes starts to measure concentration, which is what the job actually consumes.

Measure yours

The free typing test on this site reports net WPM and character accuracy over one minute, in the browser, with no signup. For the hiring versions of these measurements, the data entry accuracy test and 10-key test describe how each is scored. Both are planned, not live.

Frequently asked questions

Is 40 WPM good?
It is around average for someone comfortable with a keyboard, and it is enough for most jobs where typing is incidental rather than central. For a role that is mostly typing it is on the low side.
Is 60 WPM fast?
It is comfortably above average and sufficient for nearly every office role, including most data entry work, provided accuracy is high. Speeds above 80 WPM are genuinely fast and rarely a job requirement outside transcription.
Should I put a typing speed requirement in a job posting?
Only if the work genuinely depends on it, and then set it at the level the job needs rather than the level you would like. An inflated threshold filters out capable candidates without improving the hire, and a requirement you do not actually verify is worse than none.