How to choose a typing test for hiring

TL;DR

• Start from the role, not the test. A support agent typing into a CRM and a claims clerk keying policy numbers need different tests, different scoring units, and different pass marks.
• Accuracy beats raw speed almost everywhere. A 70 WPM typist at 90 percent accuracy creates more rework than a 55 WPM typist at 98 percent.
• There is one published anchor worth borrowing: the US Office of Personnel Management treats 40 words per minute as the typing proficiency bar for its Office Automation roles.
• A data entry test and a typing test are not the same instrument. One scores keystrokes per hour on structured records, the other scores words per minute on prose.
• Whatever bar you set, write down why it maps to the job. That single note is what makes the screen defensible if anyone ever questions it.

Pick the typing test that matches what the person will actually type all day. That is the whole decision in one line. A generic prose test tells you a candidate can copy a paragraph; it tells you very little about whether they can key 400 insurance claim records in a shift without transposing a policy number. Most hiring teams already run a typing test somewhere in the funnel. Fewer have checked whether the test format, the scoring unit, and the pass mark line up with the job.

Why does typing speed still matter in hiring?

Typing speed matters because it is one of the few screening signals that is cheap to collect, hard to fake under proctoring, and directly observable in the work. For any role where the keyboard is the primary tool, throughput and error rate are not proxies for performance. They are performance.

The roles are not disappearing quietly, either. The Bureau of Labor Statistics counts 154,230 data entry keyers in the United States as of May 2023, earning a median of 37,790 dollars a year, or 18.17 dollars an hour. Overall employment in office and administrative support occupations is projected to decline across the 2024 to 2034 decade, and yet about 2 million openings a year are still projected in that group, almost entirely from replacing people who leave. The World Economic Forum's Future of Jobs Report 2025 puts data entry clerks among the fastest declining roles, alongside postal service clerks and bank tellers.

Read those two facts together and the hiring implication flips from what most people expect. Fewer of these jobs are being created, but a great many still have to be filled every year. Volume screening for keyboard roles is not going away this decade. It is just getting more churn-driven, which makes a fast, objective first filter worth more, not less.

Which type of typing test fits the role?

Five formats cover almost every hiring case. Speed tests measure raw words per minute on prose. Accuracy tests weight errors. Data entry tests score structured records. Transcription tests score typing from audio. Simulations put the candidate in something that looks like the real screen. Match the format to the work and the score starts predicting something.

Speed tests

The candidate types a passage for a fixed window, usually one to five minutes, and gets a gross words-per-minute number. Useful as a floor check for roles where typing is incidental, like a recruiter writing outreach notes. Weak as a hiring signal on its own, because gross WPM ignores what the errors cost.

Accuracy and adjusted-WPM tests

Same passage, different math. Every mistake is subtracted before the score is reported, so a candidate who types 68 WPM with a dozen errors lands nearer 55. This is the default worth using for most office roles. If a vendor only reports gross WPM, that is a reason to look elsewhere.

Data entry and 10-key tests

Here the content is names, addresses, dates, order numbers, dollar amounts, and codes, and the scoring unit is usually keystrokes per hour rather than words per minute. Numeric-heavy roles deserve a dedicated 10-key data entry test, because number-pad speed and prose speed are close to unrelated skills. Plenty of fast prose typists are slow on the pad.

Transcription and audio typing tests

The candidate types from an audio clip. This tests listening, retention, and typing at once, and it is the right instrument for medical scribes, legal transcriptionists, and support roles that log calls live. Expect scores well below a candidate's prose speed. That drop is the point.

Work simulations

A simulation drops the candidate into a form, a spreadsheet, or a ticket queue that mirrors the real one, with the real field constraints and the real time pressure. It is the most predictive format and the most expensive to build, so reserve it for high-volume or high-error-cost roles where a bad hire is felt within a week.

Data entry test vs typing test: what's the difference?

A typing test measures how fast and how cleanly someone types continuous prose, scored in words per minute. A data entry test measures how fast and how cleanly someone keys structured records such as names, numbers, dates, and codes, usually scored in keystrokes per hour. Different content, different scoring unit, different pass mark.

The distinction matters because the two skills come apart in real candidates. Prose typing rewards rhythm and muscle memory over familiar word shapes. Structured data entry breaks that rhythm on purpose: every field is a fresh context switch, the hand moves between the number pad and the keyboard, and there is no sentence to predict ahead. Accuracy tolerance is tighter too, because one wrong digit in an account number is not a typo, it is a downstream ticket.

If a role does both, run both. Do not average the two scores into one number, because that hides the weakness you actually needed to see.

What typing speed should you require by role?

There is one published threshold worth anchoring to. The US Office of Personnel Management sets 40 words per minute as the typing proficiency requirement for federal Office Automation positions, and lets applicants prove it by performance test, by certificate, or by self-certification, with test results and certificates good for three years. Federal hiring is not a perfect model for a private employer, but it is a real, documented, defensible bar rather than a number someone made up.

Every other number here is guidance, and it deserves to be treated that way. The bands here are starting points, not validated cutoffs. The honest way to set your own is to run the test on the people already doing the job well and see where they land.

Pro tip: calibrate against your own top performers before you publish a cutoff. Give the test to five people already strong in the role. If your best claims processor types 52 WPM, a 65 WPM screen was never measuring quality, it was just measuring who practiced.

How do you set a defensible typing test bar?

Write down the link between the score and the job before you turn the screen on. In the United States, the Equal Employment Opportunity Commission expects a selection procedure to be job-related and consistent with business necessity, and it is the employer, not the test vendor, who carries that responsibility under the Uniform Guidelines on Employee Selection Procedures adopted in 1978. A vendor's documentation helps but it does not transfer the obligation.

Rather than starting with a test and hoping it maps to the role, start with the role itself. Identify the competencies that actually drive success, then connect each one to a measurable piece of evidence. For most keyboard-intensive roles, typing speed is only one competency alongside attention to detail, software fluency, and written communication. This approach also makes an important distinction clear: whether typing speed is genuinely essential to job performance or simply the easiest skill to measure.

Where does the typing test belong in your funnel?

Send it after the first screen and before anyone spends live interview time. Early enough to save hours, late enough that candidates who have not been told anything about the role are not being asked to take a timed test cold.

1. Set the benchmark before the job posting goes live, and put the number in the posting. Candidates self-select, and nobody feels ambushed.
2. Trigger the test right after the resume screen or the recruiter call, with a clear window of 48 to 72 hours.
3. Auto-filter against the accuracy floor first, then the speed floor. Reversing that order lets fast, sloppy candidates through.
4. Carry the score into the panel as one input, not a gate. A 3 WPM gap is noise, and treating it as a ranking is how good candidates get cut for nothing.
5. Store the result against the hire and check it against performance at 90 days. That is the only way the bar ever gets more accurate.

One caveat that gets skipped: typing speed is coachable, more so than most skills assessments. Two weeks of daily practice can move someone 10 to 15 WPM. That is an argument for testing accuracy and job-shaped content rather than raw prose speed, and for treating a marginal score on an otherwise strong candidate as a training question rather than a rejection.

What to check before you buy a typing test

Six things separate a hiring-grade typing assessment from a basic free web test:

• Adjusted scoring: Account for errors and report accuracy separately from speed. Gross WPM alone does not provide enough information to evaluate performance.
• Job-relevant content: Use passages, field layouts, or data formats that reflect the actual work. Generic text only tells you how candidates perform on generic typing tasks.
• Integrity controls: Use timed sessions, tab switch detection, and paste blocking to protect the integrity of remote assessments. Without basic controls, a typing score can be difficult to trust.
• ATS integration: Send results directly into the hiring workflow so recruiters do not have to enter scores manually. Manual steps are easy to skip and difficult to maintain at scale.
• Auditable reporting: Provide candidate-level results, benchmark comparisons, and exportable records that hiring teams can review and retain.
• Accessibility options: Support reasonable accommodations such as extended time or alternative input methods without forcing recruiters to create manual workarounds.

A quick way to evaluate any platform is to ask the vendor how their score would change if you swapped the prose passage for your own claim-form content. If the answer is that it would not change, the test is not measuring your job.

Related resources: Explore our guide to assessing candidates with typing tests for practical guidance on evaluating typing skills during hiring.

Assess typing skills with confidence

Set the bar from the job, test on content that looks like the work, and read accuracy before speed. Testlify runs typing, 10-key, and role-based assessments with proctoring and ATS sync built in, so the score reaches the recruiter already scored and already filtered. Book a demo to map your keyboard roles to the right test formats, or start free and run one against your current shortlist this week.

Key takeaways

• The role sets the test, not the other way around. Prose speed, keystrokes per hour, and transcription accuracy are three different skills. Picking the wrong one produces a confident number that predicts nothing, which is worse than no test at all because it feels like evidence.
• Accuracy is the screening variable that carries cost. Errors turn into rework, credit notes, and support tickets. Filter on the accuracy floor first and the speed floor second, so a fast and careless candidate never reaches the panel on throughput alone.
• Borrow the one published anchor. OPM's 40 WPM standard for Office Automation roles is a documented threshold you can point at. Everything above it should be calibrated against your own strong performers, not copied from a blog table, including the one above.
• Write the job-relatedness note before the screen goes live. Three sentences on what gets typed, how much, and what an error costs is what makes the cutoff defensible later. The employer carries that responsibility, not the vendor.
• Typing speed is coachable, so treat marginal scores as a training decision. Two weeks of practice moves most people 10 to 15 WPM. A candidate who is 5 WPM short and strong everywhere else is usually a better hire than a fast typist who is average at the rest of the job.
• Feed results back into the bar. Store the score against the hire, compare it with performance at 90 days, and adjust. A benchmark that never gets revisited is folklore with a number attached.

FAQs