Average Typing Speed, and What the Number Is For

The largest published study of typing measured an average of 51.56 words per minute across 168,000 volunteers, with 1.17 percent of characters left wrong. That figure describes people who chose to sit a typing test on a computer keyboard, which is a narrower group than it sounds, and reading it correctly is more useful than remembering it.

What is the average typing speed?

Around 52 WPM on a computer keyboard and around 36 WPM on a phone, according to the two largest datasets published on the subject. The computer figure comes from the CHI 2018 study Observations on Typing from 136 Million Keystrokes, run at Aalto University with Cambridge, which recorded 168,000 volunteers and reported a mean of 51.56 WPM with a standard deviation of 20.2. The phone figure comes from its companion study, How do People Type on Mobile Devices, which measured 37,370 volunteers at 36.2 WPM with 2.3 percent uncorrected errors.

Two things in that table deserve a second look. Thumbs on glass reach roughly seventy percent of the speed of ten fingers on a full keyboard, which is a smaller gap than most people would guess. And professional typists of the typewriter era, who used the keyboard for around eleven hours a week as their occupation, were quicker than today’s average computer user, whose keyboard time is spread across writing, searching and messaging.

Does formal training make the difference people expect?

Less than the marketing around typing courses suggests. In the same 2018 dataset, volunteers who reported having taken a typing course averaged 54.35 WPM against 49.00 WPM for those who had not, a difference of about five words per minute with a small effect size of 0.27.

What tracked speed more closely was how many fingers people actually used. Participants reported anything from one or two fingers up to nine or ten, averaging 6.95 fingers, and 47.6 percent said they used nine or ten. The correlation between finger count and speed was r = 0.38, the strongest of the self reported factors. The study also sorted its participants into eight typing styles that differ in hand use, accuracy and overlap between keystrokes, so a single average hides at least eight quite different ways of getting the job done.

The reasonable conclusion is that a course is one route to using more fingers, not a separate magic ingredient. Anyone assembling their own route will find the sequence in learn touch typing and a four week version of it in the 30 day plan.

What speeds do job ads ask for?

Adverts rarely publish WPM. Data entry roles quote keystrokes per hour, and continental European material quotes characters per minute, so the first job is converting the requirement into the unit your test reports. One word means five characters, so CPM is WPM times five and KPH is CPM times sixty.

Read the table in either direction. An advert asking for 8,000 KPH wants about 27 WPM, one asking for 12,000 wants 40, and one asking for 15,000 wants 50, which is where requirements stop describing ordinary office work. The reasoning behind the five character word is in what WPM means, the units themselves in CPM and keystrokes per hour, and what happens in an actual assessment in typing tests for job applications.

Why published averages should be read carefully

Every large typing average comes from people who chose to take a typing test, and the 2018 study is unusually honest about what that means for its own numbers. Its participants were recruited on a commercial typing site: 68 percent were in the United States, the mean age was 24.5, over 70 percent had taken a typing course, and 98.1 percent used a QWERTY layout. The authors list this under limitations rather than burying it.

So the number describes engaged test takers on one continent, not the working population of any country. Nobody has measured a national typing average, and a page that offers one has invented it or borrowed it from a self selected sample like this one.

What can be stated without inventing anything is arithmetic. A 500 word message takes 12.5 minutes of pure typing at 40 words per minute and 8.3 minutes at 60. The difference over one message is four minutes, which tells you plainly how much a change of pace is worth for the work you actually do, and how much it is not.

When the average tells you nothing useful

• Comparing your figure with an average from a different test length. A 15 second run reliably produces a higher number than a 60 second run, because there is no time for attention to drift and no awkward word to meet.
• Comparing a net figure with a gross average, or the reverse. The deduction for uncorrected errors moves a result by several words per minute before your hands are involved.
• Reading an average as a target. Half of any sample sits below its own median by definition, and this distribution leans to the right, so more than half sit below the mean. The standard deviation here is 20.2 WPM, so the ordinary spread is enormous.
• Using a phone figure and a keyboard figure in the same sentence. They are different tasks with different hands.
• Treating one session as your speed. It measures one passage on one day, on one keyboard, at one time of night.

What is your own number actually good for?

Measuring is worth the minute for three groups. People translating a stated job requirement into a unit they can test. People whose work is scheduled or paid by volume, where four minutes per message compounds across a week. And anyone practicing deliberately, who needs a baseline to compare against later.

Measuring is not worth much if the figure is being collected to sit alongside other people’s results, or if your keyboard time goes mostly on short messages and code, where thinking sets the pace rather than transcription. In those cases accuracy and comfort matter and the total does not.

Start with one 60 second run on a passage you have not seen, and write down the accuracy percentage next to the speed. Repeat it a week later, same length, comparable text. If the two figures are within a few words per minute of each other, that is your working pace, and the way it is calculated is set out in how typing speed is measured. Our typing test runs at 15, 30 and 60 seconds and reports speed, accuracy and corrections side by side, so nothing is averaged away behind a single number.

FAQ

What is a good typing speed for office work?

No published standard exists, so the figure that binds is the one printed in the advert you are answering. One public benchmark is available for comparison: California asks for 40 WPM on a proctored five minute test for its clerical classes, which is 12,000 keystrokes per hour in the unit data entry ads prefer.

How fast do professional typists type?

Studies of the typewriter era, cited in the 2018 paper, observed trained professionals at 60 to 75 WPM while using the machine around eleven hours a week. The fastest participants in the modern dataset reach 120 WPM and above, which is a rare tail rather than a benchmark.

Is typing on a phone comparable to typing on a keyboard?

Only loosely. The mobile study measured 36.2 WPM with one or two thumbs, roughly seventy percent of full keyboard speed, and it found that people using autocorrect typed faster while people choosing suggested words typed slower. The tasks differ enough that comparing the two figures says little.

Why do typing sites publish such different averages?

Because they measure different populations under different rules: gross against net speed, 15 seconds against 60, easy word lists against real prose. An average is only comparable with another average collected the same way.

Does typing speed keep rising with practice?

It rises quickly at first and then flattens, which is when accuracy and the number of fingers in use become the things worth changing. The correlations in the 2018 data point the same way: repair work, not raw hand speed, separates the groups.

For a comfortable measurement, Aesop reads better than a test sentence. The Hare and the Tortoise, in the English wording George Fyler Townsend printed in London in 1867, runs to 488 characters and turns on a steady pace rather than a burst, which is the same quality a 60 second run brings out. Alfred Tennyson published The Brook in 1855, and its 482 characters repeat short verbs in pairs, so a tempo arrives without being chased. If you would rather begin with something shorter, Jane Austen’s walk down to the sea at Lyme, from Persuasion in 1818, is 260 characters of plain sentences. All three are waiting in the typing test, each with its author, year and source on its own page.