Why Typing Speed Still Matters (and Where It Doesn't)

Typing speed still matters because typing never became optional. It became invisible. Almost everything you do at a computer — searching, messaging, taking notes, filling forms, writing documents, describing what you want to a tool — passes through a keyboard, and the amount of text an ordinary desk job expects has gone up rather than down. Note-taking is the one item on that list where more speed is not obviously better, for reasons set out in whether typed or handwritten notes serve you better.

What changed is the shape of it. Long, uninterrupted transcription is mostly gone. In its place is a constant stream of short, exact, consequential text: a message, a search, a command, a correction. That pattern rewards different things than the old one did — accuracy over stamina, and low effort per keystroke over top speed.

This piece makes the case honestly, which means it also makes the case against. There is a point where faster typing stops buying you anything, and it arrives earlier than most people expect.

The prediction that typing would stop mattering

The case against typing has been made confidently for well over a decade, and it has arrived in waves. Voice recognition would make speaking faster than typing. Touchscreens would move everyone off keyboards. Most recently, generative tools would write the text so nobody would need to produce it. Meanwhile the keyboard quietly became the primary interface for competitive gaming too, where typing in chat under time pressure is its own small skill.

Each wave contained something true. Dictation genuinely is faster for prose you have already composed, though not in a support queue where you cannot speak your replies aloud. Phones did absorb an enormous share of casual messaging — though texting speed measured properly still lands well short of a keyboard. Tools genuinely do produce first drafts now. And yet the keyboard did not recede, because none of those changes addressed what the keyboard is actually for.

It is not primarily an instrument for producing volume. It is the interface for expressing something precise to a machine — a query, an instruction, an edit, a correction — and that need has grown with every new tool rather than shrinking. The parts of typing that disappeared were the mechanical parts. The parts that remained are the parts where exactly what you write determines what happens next.

The real cost of slow typing is attention, not minutes

The time arithmetic is the most quoted argument for typing faster and the least interesting one. The larger cost is cognitive. While typing still takes conscious effort — locating a key, choosing a finger, checking what appeared on screen — it is drawing on the same working memory you need for whatever you were trying to say.

That is why people who learn to touch type describe the result as typing becoming effortless rather than fast. The speed increase is real and secondary. What changes their experience is that the keyboard stops competing for attention, so the sentence they were composing survives the act of writing it down.

It shows up most clearly in tasks where you are thinking and writing at the same time: working through a problem in notes, drafting a message you are not sure how to phrase, describing a bug you only half understand. Any attention spent on the mechanics is taken directly from the part that matters. Someone who has to look at the keyboard is, in a small but constant way, interrupting themselves.

What the time difference adds up to

For completeness, here is the arithmetic, using 2,000 words as a moderate day of combined output. The shape of it is more instructive than any single figure.

The important feature is the diminishing return. Going from 30 to 40 WPM saves seventeen minutes a day; going from 60 to 70 saves under five. That is simultaneously the honest case for learning to touch type and the honest case against chasing speed indefinitely — almost the entire practical benefit is captured before you become fast. Where your current number sits in that range is set out in average typing speed.

Where typing is still the interface

It is worth being concrete about how much of an ordinary day runs through the keyboard, because the total is easy to underestimate when no single task feels like typing.

No line in that table looks like a typing task. Added together they are most of a working day, which is the reason the skill compounds so quietly. Nobody notices the keyboard being slow; they notice the day feeling like more work than it should have been.

Why voice never replaced the keyboard

Voice recognition has been essentially solved for years and still has not displaced typing for desk work. That is not inertia or habit. Dictation is genuinely faster than typing for one task — producing continuous prose you have already worked out in your head — and worse for nearly everything else a keyboard is used for. The controlled evidence, including a study that measured speech at three times keyboard speed a decade ago, is worked through in voice typing versus typing.

The last row is the underrated one. Typing lets you think mid-sentence, back up and change direction without anyone hearing the false start. Speaking forces you to commit to a phrasing before you have finished deciding on it, which is why people who dictate well usually compose first and dictate second.

Coding, prompting and AI tools

Two areas deserve a specific mention, because both are often given as reasons typing no longer matters and both turn out to cut the other way.

Code is harder to type than prose, even though there is less of it. Very little of a developer's day is spent typing long stretches, so raw speed is genuinely not the constraint — but the difficulty is concentrated exactly where self-taught typists are weakest. Brackets, braces, semicolons, pipes and angle brackets sit on the weakest fingers or need Shift. camelCase puts Shift mid-word rather than at the start of a sentence. Ports, indices and versions drag in the number row, which almost nobody drills. Underscores and hyphens sit far from home position. And terminal commands are short, exact and unforgiving, typed dozens of times an hour. What the character counts actually show puts numbers on every clause of that paragraph, measured against real published prose.

Autocomplete and AI assistants reduce keystrokes substantially and, in exchange, increase reading, judging and editing — you spend more time evaluating a suggestion and adjusting it than producing it from scratch. Editing existing text is the one thing voice cannot do, and it is the thing that keeps the keyboard central to the work.

The AI side is similar in structure. These tools removed a real category of typing — boilerplate, scaffolding, first drafts, formulaic email — and replaced it with typing that describes intent: prompts, constraints, examples, corrections, specifications. That replacement text is shorter and more consequential than what it displaced, and the quality of what comes back depends heavily on how many passes you are willing to make. Someone who finds typing faintly unpleasant tends to accept the second attempt; someone for whom it costs nothing takes five without deciding to.

• For coding, there is no speed threshold. Above roughly 50 WPM speed stops mattering, and what matters is producing a bracket, an underscore or a capital mid-word without looking down.
• Drill the material you actually type. Symbols, digits and mixed case transfer poorly from prose practice, which is why competent prose typists often stumble on code — and how to produce symbols without hunting for them covers the characters that are not printed on any key at all.
• For AI work, accuracy outranks speed. A mistyped prompt does not throw an error; it answers a slightly different question, confidently.
• Neither area rewards chasing a high WPM. Both reward the keyboard being automatic, which is a technique question rather than a speed one — the finger assignments are in how to learn touch typing.

Accuracy matters at least as much as speed

If the work has shifted toward short, exact text, then the more valuable metric is accuracy. This is also true arithmetically: only correct characters count toward your score, so every error lowers your speed directly, and each mistake costs time twice — once to make and once to correct.

The practical consequence is that a fast, error-prone typist is usually slower than a moderate, clean one. A run at 70 words per minute with a scattering of mistakes finishes below a clean 50, and the person typing at 70 experiences themselves as the faster typist throughout. The gap between your raw and net speed is the number that settles it, and what WPM, accuracy and consistency actually measure explains how both are calculated.

Errors also cost more than the keystrokes they consume. A typo in a message costs a clarifying reply; in a form, a rejected submission; in a search, a wasted query; in an instruction to a tool, an answer to the wrong question. The habits that produce most of them are covered in common typing mistakes.

Who gains most from typing faster

The benefit is not evenly distributed, and the people who gain most are usually not the ones who think of themselves as typists.

• Anyone whose thinking happens in writing. The keyboard is either transparent or it is editing your thoughts as you have them.
• People in constant written conversation — support, sales, recruiting, anyone running several threads at once with someone waiting on each.
• Students, for real-time note-taking, where looking down means losing the thread of what is being said.
• Developers and analysts, for symbol-dense text rather than volume.
• Anyone who writes specifications, tickets or documentation, where precision is the whole deliverable.
• People who currently type with two to four fingers, who have the most available gain of anyone — the ceiling of that method is around 45 WPM regardless of effort.

Where typing speed genuinely does not matter

Two caveats, because the case for typing is routinely overstated and an argument that ignores its own limits is not worth much. First, typing speed is not a proxy for productivity, intelligence or capability. Plenty of excellent writers and engineers type slowly, and plenty of fast typists produce nothing worth reading. It removes a specific kind of friction; it does not supply competence, judgment or ideas — and past a certain speed, the keystrokes you never needed to type are the larger remaining saving.

Second, for composition — as opposed to transcription — the bottleneck is usually thinking rather than typing. Most people compose original prose far below their transcription speed, which means a 90 WPM typist writing an essay is not going twice as fast as a 45 WPM one. Both are waiting on the same sentence to arrive. The gain concentrates in tasks where you already know what you want to say: messages, commands, notes, forms, code.

Both caveats point the same way. The value is in getting from slow to comfortable, not from comfortable to fast. Past about 70 WPM, further speed is a hobby rather than an investment, and there is nothing wrong with that as long as you are not mistaking it for a productivity gain.

What to do about it

The action is small and the return is unusually durable, which is the whole reason this argument is worth making.

• Find out where you actually are. Take a timed test and record accuracy and consistency alongside the WPM rather than only the headline number.
• If you look at the keyboard, that is the only thing worth fixing. Three or four weeks of short daily practice removes a ceiling that nothing else will move.
• If you already touch type, work on accuracy rather than pace. Get above 97 percent and let speed rise on its own.
• Practice a little and often. Fifteen minutes daily beats two hours weekly, because motor learning consolidates between sessions — the routine is in drills and a practice routine.
• Stop somewhere around 60 to 70 WPM unless you enjoy it. The curve has flattened and the remaining gains cost more than they return.

A few weeks of short daily sessions buys a skill you will use every working day for decades, in a period when the amount of precise text an ordinary job expects is rising rather than falling. That was a reasonable trade when typing meant transcription. It is a better one now that most of what you type is something that has to be exactly right.