You do not need an assessment platform to run a fair typing test for employers. You need three things: every candidate taking the same test at the same length, a result you did not receive as a screenshot from the candidate, and a record you can check later. A subscription platform bundles those three with reporting, custom passages and admin seats, and charges from roughly $49 a month upward for the privilege. If you are hiring one receptionist or two data entry clerks this quarter, buy the outcome instead of the software: send every candidate the same locked test link, collect server-scored net WPM against a certificate ID, and look each ID up when a result matters. TypeTimer's recruiter pack does exactly that for a one-time $29 covering 10 verified certificates, with no account for the candidate and no onboarding for you.

Most writing about candidate typing assessment is written for the buyer of a platform. That buyer exists, and for a company screening 400 applicants a month a platform is the right call. This is for the other case, which is far more common and almost never addressed: a small business, an office manager, or an independent recruiter who needs a defensible typing number from a handful of people, once.

How do employers usually test typing speed?

There are four patterns in the wild, and they sit on a spectrum from casual to formal.

The first is no test at all. The posting says "50 WPM preferred", nobody checks, and the number on the resume is taken on trust. This is the most common approach and the least useful, because self-reported speed carries almost no information. Typing speed is not a single stable trait. The Aalto University team behind Observations on Typing from 136 Million Keystrokes, a CHI 2018 study of 168,000 volunteers, found typists divide into eight distinguishable groups that differ in performance, accuracy, hand and finger usage. Where a candidate lands depends on the keyboard, the text and the day. A number with no context attached tells you very little about any of that.

The second is "send me a screenshot". The candidate takes a free test somewhere, screenshots the result, and emails it. This gives you a number, and it gives you no idea what produced it. Different length, different site, different scoring method, and an image file that anyone with a browser inspector can edit in about forty seconds. Two candidates who send screenshots are usually not comparable to each other, let alone to a bar you set.

The third is a live test at interview. Sit the candidate at a machine, run a timed passage, watch. This is genuinely good evidence and the only one on this list that observes the person typing. It is also expensive in your time, impossible for remote-first hiring, and awkward to schedule before you have decided who is worth an interview.

The fourth is a formal, administered exam. Public sector employers run these well and publish the mechanics, which makes them a useful template. The City of Henderson, Nevada documents its skills-based keyboarding proficiency exam in unusual detail: the exam is delivered through eSkill, takes an average of 5 to 15 minutes, requires a computer with a real keyboard, the Chrome browser and a working webcam, and can be taken remotely or in person. Scoring uses standard net WPM. Practice attempts are explicitly not counted, and retakes are not granted for technical trouble unless the platform is at fault. Note what makes that credible. Not the software. The fact that every candidate got the same instrument under the same rules, and that the rules were written down before anyone took the test.

That last point is the whole game, and you can reproduce it without buying anything monthly.

What does an assessment platform cost for one hire?

More than the hire is worth measuring, usually. Published entry tiers for dedicated typing assessment tools start at around $49 a month for a single administrator and a capped number of test takers, and climb into the hundreds for custom passages, extra seats and API access. Broader hiring-assessment suites that include typing as one module tend not to publish a number at all, which in practice means a demo call and an annual contract.

Run the arithmetic for a real small hire. You are filling one admin role. Twelve applicants clear the resume screen and you want a typing number from all of them. On a monthly plan you pay for a full month, learn the admin console, configure a test, screen your twelve, then remember to cancel. The per-candidate cost is fine. The per-hire cost, including the hour you spend setting it up and the month you forget to cancel, is not.

Here is the comparison laid out. Costs are the published or listed prices as of September 2026 and the friction column is about candidate experience, which matters more than buyers expect: a screening step that annoys good applicants filters for patience, not typing.

	Assessment platform	DIY link to a free test	TypeTimer recruiter pack
Cost	Subscription, from about $49/month at entry tier, often annual for suites	$0	One-time $29 for 10 verified certificates, credits do not expire
Setup	Account, admin console, test configuration, sometimes a sales call	Paste a URL into an email	Pay, enter your team name, receive a candidate link and a private results link
Standardization	Strong, you configure the passage and duration	None, candidates pick length and site unless you police it	Every candidate takes the same locked 3 minute test
How results reach you	Reports in the platform dashboard	Candidate emails you a screenshot	Names, net WPM, accuracy and certificate IDs land on your private results page as candidates finish
Verification	Trust the vendor's record, visible only inside your account	None	Public lookup at /verify by certificate ID, checkable by anyone, including a later employer
Candidate friction	Account creation, system requirements, sometimes a webcam	Low, but they have to figure out what you wanted	No account, no payment, one link, three minutes
Best for	Continuous or high volume screening	Informal roles where the number is nice to have	One to a few hires, where the number needs to be real

None of these is wrong. They are priced for different problems. The platform column is what you want when typing screening is a permanent process. The right-hand column is what you want when it is an event.

How do you standardize results without proctoring?

By locking the variables you can lock, and being straight about the one you cannot.

The variables that actually move a WPM score are test length, passage difficulty, scoring method and whether errors are counted against the total. Fix all four and two candidates become comparable. A 3 minute test is the practical sweet spot for screening: long enough that a burst of adrenaline cannot carry someone through, short enough that nobody abandons it. Our own writeup of 1 minute versus 5 minute typing tests covers why the same typist scores meaningfully higher on the shorter one, which is exactly why you should not let candidates choose.

Scoring should be net WPM, meaning gross speed with uncorrected errors deducted, computed server side rather than in a page the candidate can pause. That is the same standard Henderson names in its exam documentation, and it is the number worth comparing because it prices accuracy into speed automatically. A candidate typing 70 gross with a wreckage of typos is slower in real work than one typing 58 clean.

The variable you cannot lock remotely is identity and supervision. An unproctored test taken in a candidate's own browser is a screening signal, not an exam. Every online typing test for employment works this way, including the expensive ones, and any vendor telling you otherwise is selling you a webcam feed and calling it proof. Treat the remote result as a filter, then confirm your finalist in person at interview with two minutes at your own keyboard. If the role genuinely requires certified proof under supervision, an online certificate from anyone, ours included, will not satisfy that requirement, and the employer or agency setting it will name the testing centre they accept.

How does certificate verification replace platform reports?

A platform report is trustworthy inside the platform. Its weakness is that it is a screenshot the moment it leaves. Export the PDF, email it to the hiring manager, attach it to the offer file, and you are back to trusting an image.

A verifiable certificate inverts that. Each TypeTimer result issues a certificate carrying a unique ID in the form TT-XXXXX-XXXXX, recorded server side against the certified name, net WPM, accuracy, test length and issue date at the moment of the test. Anyone can enter that ID at the public verification page and see whether the details match. An edited document fails the check, because the record was never in the document. This is the same mechanism candidates use when they buy a single verified certificate for $4.99, described in more depth in our guide to typing certificates for employment.

For an employer that changes two things. First, you can accept a certificate a candidate already holds instead of making them take your test, because you can confirm it in ten seconds without emailing anyone. Second, your own results survive the file. Six months later, when someone asks how the keyboarding requirement was assessed, the ID still resolves.

We wrote separately about how employers verify typing speed from the candidate's side. The short version from your side of the desk: you are not verifying a person, you are verifying that a specific measurement happened under known conditions on a known date.

What passing bar should you set for the role?

Start from the work, not from a benchmark table. Henderson's own FAQ puts it plainly: a keyboarding minimum appears when the hiring manager has determined that keyboard proficiency is essential to the job's functions, and the job announcement says so up front. That ordering is worth copying. Decide whether typing speed is genuinely essential, decide the number before you see any candidates, and publish it in the posting.

Then calibrate against the market rather than your instinct. Read ten current postings for the same role in your region and note the minimums they state. Our roundup of typing speed requirements for jobs collects the common bands by role type, and it is a better starting point than a number someone half remembers from a keyboarding class. Two practical notes: set the bar as a floor rather than a ranking device, because the difference between 62 and 68 WPM will not decide who is good at the job, and pair it with an accuracy minimum, since speed without accuracy is negative value in any role involving records.

Because this is a selection procedure, employers running one should know the framework it sits inside. The EEOC publishes technical assistance on employment tests and selection procedures covering how employers use tests to screen applicants and where issues can arise under federal employment law. We build typing tests, we are not employment counsel, and anything about your specific screening process is a question for someone who is.

FAQ

Can I ask candidates to take an online typing test?

Employers routinely do, and the practice is unremarkable when the requirement is stated in the posting, applied to every candidate for that role, and tied to work the person will actually do. Where employers get into difficulty is applying a test inconsistently, or setting a bar with no connection to the job. Decide the requirement before the applicants arrive, write it into the posting, and run the identical test for everyone. For the legal specifics of your situation, ask an employment lawyer rather than a typing site.

How do I know a candidate's result is really theirs?

For a remote, unproctored test, you do not know with certainty, and neither does any vendor charging you monthly. What you can know is that a specific result was scored on our server, at a recorded time, under the same 3 minute conditions as every other candidate, and that its certificate ID resolves at /verify. That is a screening signal, and it is a considerably stronger one than an emailed screenshot. When the result is decisive, retest the finalist in person for two minutes. Discrepancies show up immediately.

What WPM should I require for admin roles?

It depends on how much of the day is spent at a keyboard, so resist a single universal number. General office and reception work sits well below what a dedicated data entry or transcription role demands, and postings for the same title vary by employer and region. Pull the current postings for your role, take the median stated minimum, and pair it with an accuracy floor. Our typing speed requirements guide has the common bands by role category if you want a starting range to sanity check against.

Is a 3-minute or 5-minute screen better?

For screening, 3 minutes. It is long enough to expose the drop-off that shorter tests hide, short enough that candidates finish it, and it fits inside an application flow without feeling like an exam. Five minutes is the better instrument when you need a sustained-performance number for a role that types continuously all day, such as transcription or heavy data entry. If you are choosing between them for a general admin hire, take the shorter test and the higher completion rate.

Screen your next batch this afternoon

If you are hiring one or two people and want a real number instead of a claim, the recruiter pack is 10 verified certificates for a one-time $29. Pay once, enter your team name, and you get two links: a candidate link to send out and a private results page to bookmark. Candidates pay nothing and create no account. They take the same locked 3 minute test, and their names, net WPM, accuracy and certificate IDs appear on your page as they finish. Every ID is checkable by anyone at typetimer.com/verify, including by the next employer who asks. Credits do not expire, so the second hire costs you nothing extra. If you would rather see the candidate side first, take the test yourself and look up your own certificate before you send it to anyone.

Typing Test for Employers: Screen Without a Platform