How to pass a typing test for a job

Most advice on how to pass a typing test for a job tells you to practise more. If your assessment is on Thursday, that is not advice, it is a shrug. The useful part is narrower: employers score typing differently from the site you have been practising on, and the gap between those two scoring systems is worth more words per minute than a week of drilling.

This is about the test you have booked. If you are still working out what speed the role wants, its sibling covers that: typing speed requirements by job.

What a pre-employment typing test actually measures#

Two numbers, and the second one decides more outcomes than the first. Speed gets quoted in the job ad. Accuracy gets checked when they read your result.

Which WPM figure they report#

Typing tests generally start from a gross speed: total characters typed, divided by five, divided by minutes. The five is a convention rather than a measurement, and it is why a long word costs you no more than a short one.

Then the scoring splits. The formula most often used on pre-employment tests takes that gross figure and subtracts a penalty for each uncorrected error:

Net WPM = gross WPM minus (errors divided by minutes)

Typepace does something different. It counts only the characters you got right, and divides those by five. No penalty is subtracted afterwards, because the wrong keystrokes were never added.

Both results are called net WPM. On the same run they disagree. Here is one sixty-second run scored three ways, with 400 keystrokes and eight of them wrong:

Six words per minute separate two numbers that share a name. If you have been quoting your practice score to yourself as the one an employer will see, you may be six ahead of where you actually stand. Assume the stricter reading and you will never be surprised in the room.

One consequence matters on the day. A penalty of one word per error, spread across the test length, means errors cost far more in a short test. Five errors in a one-minute test take five words per minute off you. The same five errors in a five-minute test take one. If you get to choose the length, choose the long one.

The accuracy threshold, which is usually the real gate#

Speed is the number on the ad. Accuracy is the number that fails people, and it is often unstated.

The clearest published example is the US federal standard for office automation roles. The Office of Personnel Management's qualification standard for clerical and administrative support positions asks for 40 words per minute, measured on a five-minute sample with three or fewer errors.

How you prove it is more flexible than the number suggests. Applicants may pass a performance test, present a certificate of proficiency, or self-certify, and performance test results and certificates stay acceptable for three years. An agency may still test a self-certified applicant.

Read the error allowance rather than the speed. Three errors across a five-minute sample is a tight allowance. The OPM standard specifies the number of errors but does not explain the error-counting method in this section. The speed is modest. The allowance is not.

That pattern repeats in private assessments with the numbers hidden. When an ad asks for a "high degree of accuracy" and gives no figure, you cannot know where the line sits, so practise close to clean rather than close to the speed.

The platforms, and what you can actually find out#

Many employers use a third-party assessment rather than building their own. The names that recur are Kenexa Prove It, Criteria Corp and eSkill, and staffing agencies run their own variants on top of them.

You will find dozens of pages claiming to know exactly how each one scores. Be careful with them. Vendors change scoring between versions, employers configure their own thresholds, and most of those pages are selling practice packs. A confident table of per-vendor formulas is usually somebody guessing.

What you can rely on is the invitation email. It usually names the vendor, the test length, and sometimes the passing score. If it names none of those, ask the recruiter. Asking how long the test runs and whether errors can be corrected is a reasonable question that makes you look prepared rather than nervous.

Everything else you can work out in the first ten seconds of the test itself, which is the next section.

During the test: whether to correct your errors#

This is the decision that moves your score most, and almost nobody tells candidates how to make it. It depends on which of three behaviours the test has, and you can identify yours by deliberately mistyping one character at the start.

• The cursor refuses to advance. The test blocks wrong keys. You have no choice: fix it and move on. Speed here comes from not stopping, because every wrong press costs you the time to press again.
• The character turns red and the cursor advances. Errors are recorded and you are allowed to leave them. Leave them. Backspacing costs additional keystrokes and time, and under a one-word penalty that trade only pays if the correction is nearly instant.
• The character turns red and backspace clears the mark. The test scores your final text rather than your first attempt. Correct everything, because an uncorrected error here is a permanent error and a corrected one is free.

Spend one keystroke finding out which you are in. It is the cheapest information in the whole test.

One caution on the second case. Leaving errors is right under a penalty of one word each, and wrong if the employer also reads the accuracy percentage, which many do. If the invitation quotes an accuracy threshold, fix your errors even when the test lets you run past them.

Pacing, and why starting fast loses#

A typing assessment often takes the same shape: a fast opening thirty seconds, a spike of errors, then a careful crawl that never recovers the average. Starting above your sustainable speed costs twice, once in the errors themselves and again in the caution they produce.

Type the first line at a pace you could hold for an hour. Your speed rises through the middle of a test on its own, because your hands settle and the text stops being a surprise. Give that the room to happen.

If you have a five-minute test, treat the first thirty seconds as a warm-up you happen to be scored on. The arithmetic supports it. At five minutes, a slow opening costs you a fraction of a word per minute, and the errors it prevents cost a full word each.

What to do after a bad first line#

Keep going at the same speed. The instinct is to slow down and make it back, and the arithmetic says otherwise: a bad line is already spent, and slowing for the rest of the test spreads its damage across every remaining word.

Candidates lose more to the ninety seconds of rattled typing after a mistake than to the mistake. Read the next word, type it, and let the previous line stay where it is.

Preparing with a week#

A week is enough to move accuracy and not much else. Speed responds to months. Accuracy responds to days, because most of it is a habit of not rushing rather than a physical limit.

So spend the week on the number that is easier to move and more likely to be the gate.

If your practice text is all lowercase common words, you are training for an easier test than the one you booked. Employer samples carry capitals, commas, full stops and often figures. Practise on ordinary prose with its punctuation intact.

Preparing with a day#

Do three things and skip everything else.

• Take one timed test at the length your assessment uses, and stop after it. One measurement, not five. Repeated tests in one sitting measure your fatigue.
• Read your accuracy rather than your speed. If it sits below 97%, spend the remaining time typing slowly and correctly rather than fast and hopefully.
• Sleep. Tired hands make more errors, and a late night costs more than the practice it bought.

The hour before#

Use the first few minutes to warm up before the assessment, then leave it alone. On a one-minute test, the opening seconds represent a large share of your result.

Set the desk up the way you practise: chair height, screen distance, both feet down. If you normally type on a laptop and the assessment is on a desktop, the keyboard is the variable to control, not the room.

Then stop practising. The hour before a test is for arriving calm, and no drill in it will change your result.

Unfamiliar keyboards and remote assessments#

An unfamiliar keyboard is a preventable way to lose marks. A different key travel, a shifted Enter, a laptop after months on a mechanical board, and your accuracy drops before you have read the first line.

For an on-site test, ask if you can use your own keyboard. Many employers say yes, and the ones who say no have told you something useful about the job.

For a remote test, control what you can. Close everything else, so a notification cannot land mid-sentence. Use a wired connection if the test streams keystrokes. Check whether the test locks your browser, and if it does, open it once beforehand so the permission prompts happen before the clock starts.

Layout matters more than brand. A UK keyboard and a US keyboard disagree about the apostrophe, the quotation mark and the hash, and business prose is full of apostrophes. If the machine is not yours, type one sentence with a contraction in it before you begin.

If you fail it#

Ask whether you can retake it, and when. Many employers allow a second attempt after a fixed interval, and the ones who do not will say so plainly.

Then find out what you actually failed. A result of 46 wpm against a 45 requirement is not a speed problem, and drilling speed will not fix it. Where the shortfall is accuracy, that is the faster of the two to repair.

If the same score keeps appearing across different sites, the explanation may be the scoring rather than your hands. Why your WPM differs between sites covers what changes between them.

Where an employer accepts a certificate rather than their own test, and federal roles are the clearest case, a dated certificate can settle the question without an assessment at all. Anyone can check a Typepace certificate against the code printed on it.

Questions people ask before an assessment#

What to expect in a typing test: a passage of business prose, a fixed length between one and five minutes, and a result giving speed and accuracy. You will not be asked to type from dictation or from paper unless the role is transcription.

Can I use my own keyboard? On site, ask. Remotely, you already are, and it should be the one you practised on.

Should I correct my mistakes? Run the one-keystroke test in the first ten seconds and let the answer decide. If the test quotes an accuracy threshold, correct them regardless.

How many words per minute do I need? That belongs to the job rather than to typing, and typing speed requirements by job sets out what each role asks for and where those figures come from.

Is 40 wpm slow? Not for most office roles, and the federal standard sits exactly there. The accuracy attached to it is the demanding half.

Do practice sites inflate scores? Some score generously, and none of them use your employer's configuration. Practise for the strictest reading and let the real test be the pleasant surprise.

How long before a test should I stop practising? An hour. Warm up for five minutes and stop.

Run a timed test at the length your assessment uses, on text with its punctuation intact. It is the closest simulation you can get for nothing, and it takes a minute.