Mechanical vs Membrane: Does Your Keyboard Add WPM?

Almost everything written about this question is published by someone who sells keyboards. That is not a conspiracy, it is just who has a reason to write 1,200 words about switch feel. It does mean the conclusion tends to arrive before the evidence does. We sell typing tests, not hardware, so we have no stake in which board you end up on.

Does a mechanical keyboard actually raise WPM?

Start with what is actually known about typing speed in general, because it reframes the question.

The largest dataset on the subject comes from a 136 million keystroke study run by Aalto University and the University of Cambridge, covering 168,000 volunteers. The average participant typed 52 words per minute. The fastest typed around 120. The interesting part is what separated them. The fastest typists performed between 40 and 70 percent of their keystrokes using rollover, pressing the next key before releasing the previous one. And people who had taken a formal typing course typed at broadly similar speeds to people who never had, even though the self taught group used fewer fingers.

Read that again, because it is the whole argument. The single biggest published study on typing speed found the difference between a 52 WPM typist and a 120 WPM typist sitting in overlap, rhythm and finger economy. Not in hardware. Hardware is not even a variable it reports.

So what does the hardware literature say? Less than the marketing implies, and it is mostly about a different question.

The closest published work is a Journal of Applied Biomechanics study on short travel key switches from 2019. Twenty participants typed on four keyboards with different key travel: two very short travel boards at 0.55 mm (a dome switch and a butterfly switch) and two scissor switch boards at 1.3 mm and 1.6 mm. The shortest travel keyboard produced typing speeds about 8 percent slower, 67 WPM against 75 WPM on the longest travel board. Adjusted for errors the gap was 64 against 72, which held up statistically. The short travel board also drew 12 percent more typing force and 6 to 8 percent more forearm muscle activity.

Two things about that result matter, and the keyboard blogs tend to quote the first and skip the second.

The first: travel distance is real. A board with almost no key travel measurably cost people speed and cost them more force to use. If you are typing on something with near zero travel, that is a genuine finding and it applies to you.

The second: that study did not test a mechanical keyboard. It compared dome, butterfly and scissor switches, which is to say laptop and tablet class keyboards. The authors' own conclusion was that key travel alone does not predict typing force or muscle activity, and that the switch mechanism and activation force matter beyond travel. You cannot take an 8 percent figure measured between a butterfly switch and a scissor switch and reprint it as "mechanical keyboards are 8 percent faster". People do it anyway.

There is also a neat finding on how badly we judge our own hardware. A wearable EEG study on keyboard switch pre-travel put 25 participants on five switches that differed only in how far the key moves before it registers. Reaction times on the shortest pre-travel switch were significantly quicker than on the medium and longest ones. But when the same participants were surveyed, their subjective rankings showed no significant differences between the switches at all. The measurable effect existed. The ability to feel it did not.

That is the real trap in this whole category. A mechanical keyboard feels dramatically different the moment you touch it, and your brain files that sensation as "this is faster". Feel and speed are separate measurements, and the only way to separate them is to measure.

Which switch types suit fast typing?

Switch choice matters more for comfort, noise and error rate than for raw speed. Published actuation specs vary by manufacturer and by individual switch model, so treat any single number you see quoted as belonging to one specific switch rather than to a whole category. Here is what each type actually changes about the act of typing.

Type	What it feels like	What it changes for typing
Linear (mechanical)	Smooth, consistent pressure all the way down	No feedback at the actuation point, so some typists bottom out harder and tire sooner. Quiet in the tactile sense, still loud on impact.
Tactile (mechanical)	A distinct bump partway down	The bump tells your finger the keypress registered, which can reduce both bottoming out and missed keys. Common choice among people who type prose all day.
Clicky (mechanical)	A bump plus an audible click	Same feedback as tactile plus sound confirmation. Genuinely disruptive in shared rooms and on calls. Speed benefit over tactile is not established.
Rubber dome (membrane)	Mushy, with a collapse near the bottom	Feedback arrives at the bottom of the stroke rather than at the actuation point, which encourages full presses. Cheap, quiet, and perfectly capable of high WPM.
Scissor (laptop, low profile)	Short, crisp, shallow	Short travel means less finger movement per key. The study above found the 1.3 mm and 1.6 mm scissor boards among the faster devices tested.

The practical read: tactile switches are the safest default for long form typing because the feedback arrives where the keypress actually happens. Linear switches are a preference, not a speed upgrade. Clicky switches are a social decision. And a decent scissor or membrane board is not the handicap the category assumes it is.

One more thing the table cannot show. Switching switch type resets your muscle memory for a week or two. Every switch change costs you speed before it gives any back, which is exactly why casual before and after comparisons are so unreliable.

When is hardware genuinely your bottleneck?

Sometimes it really is the keyboard. The honest tells:

• Keys that miss. If you are typing a word correctly and a letter does not appear, that is a chattering or worn switch, and it is costing you real accuracy. This is the one unambiguous case for replacement.
• Keys that repeat. Same problem, opposite symptom.
• Near zero travel. A board where the keys barely move is the one configuration with published evidence of a speed and force penalty behind it.
• Pain. Wrist, forearm or finger pain after a typing session is a hardware and posture problem, and no amount of practice fixes it. Address this before you address speed. It outranks WPM entirely.
• A layout you fight. A cramped arrow cluster, a half sized right shift, a function row you hit by accident. Layout errors cost more WPM than switch type ever will.
• You are already past roughly 100 WPM. At that point you are genuinely operating near the limits of what the hardware reports back to you, and small differences in feedback and consistency start to show up in the numbers.

If none of those apply, your keyboard is not the thing standing between you and a better score. The route from 60 to 80 WPM runs through rhythm and error control, not through a shopping cart.

How do you test a keyboard's effect on your own speed?

This is the part no review site can give you, because the answer is specific to your hands. Every result here saves to your browser automatically, so the comparison data builds itself while you run the tests.

Here is the protocol. It takes about twenty minutes of testing spread over a few days.

1. Give the new keyboard an adaptation period first. Use it normally for at least a week before you measure anything. If you test on day one you will measure unfamiliarity, not hardware, and you will get a confidently wrong answer. This step is the one most people skip and it invalidates everything after it.
2. Fix your test conditions and do not change them. One duration and one mode for every single run. The 3 minute typing test is the right length here: long enough that a single lucky sentence cannot carry the score, short enough that you will actually do ten of them. Do not mix prose mode and words mode, because that alone moves scores by a double digit percentage.
3. Warm up and throw the warm ups away. Two untimed or discarded runs on each keyboard at the start of every session. Your first test of the day is always slow and it is not information.
4. Alternate the boards, do not block them. Run A, B, B, A, A, B, B, A, A, B rather than five of A followed by five of B. Blocking lets fatigue and time of day load onto whichever keyboard you tested second.
5. Record five counted runs per keyboard. Ten runs total. Keep them in one sitting if you can, or run the same block at the same time of day across two or three days.
6. Compare medians, not personal bests. Your best run is your luckiest run. The median is the number that describes your actual typing.
7. Compare accuracy and consistency too. A keyboard that leaves your WPM flat but lifts your accuracy is a real improvement, and WPM alone hides it.

A worked example. Say your five runs on the membrane board come out at 68, 71, 74, 69 and 72, and the five on the mechanical at 70, 76, 73, 71 and 74. Medians: 71 against 73. That is a 2 WPM difference, and your own spread within each keyboard is 6 WPM. The honest conclusion is that this experiment did not detect a difference. It did not prove there is none, it just showed that any real effect is smaller than your own noise. That is a genuine finding and it will save you money.

Decide the threshold before you look at the data. A reasonable rule: if the gap between medians is smaller than the spread within either keyboard, call it noise. If the mechanical board's median comes out 8 or 10 WPM clear with tight spread in both groups, you have found something worth paying for.

The confounds to name out loud, because they are what ruin casual comparisons: the warm up effect, adaptation to a new board, time of day, caffeine, passage difficulty, testing when you already know which result you want, and the fact that a new keyboard is exciting and excitement looks a lot like speed for about a week.

What matters more than the keyboard?

In rough order of how much WPM each one is actually worth:

Not looking at your hands. Every glance down is a re-acquisition of position and a break in rhythm. Nothing else on this list returns as much.

Accuracy before speed. A backspace is not neutral, it costs the keystroke, the correction, and the rhythm on both sides of it. Typing at 90 percent accuracy and fixing everything is slower than typing at 98 percent, every time.

Rollover and rhythm. This is the finding from the Aalto and Cambridge data that most people skip over. The fastest typists overlap a large share of their keystrokes. Overlap comes from an even cadence, and an even cadence comes from practising at a speed you can sustain instead of sprinting and correcting.

Finger economy, whatever your form. The same study found self taught typists using fewer than ten fingers keeping pace with formally trained typists. You do not need canonical form. You need consistent form, so the same finger makes the same journey every time.

Fixing your worst bigrams. Most people lose a disproportionate amount of time to a handful of letter pairs their hands never learned. Finding those and drilling them is unglamorous and it works. Our guide on how to type faster covers the drilling method in detail.

Then, somewhere well down the list, the keyboard. A comfortable board you enjoy using makes you practise more, and practising more is what actually moves the number. That is a real benefit. It is just an indirect one, and it is not what the switch marketing is claiming.

Buy the mechanical keyboard if you want one. They are pleasant, they last, and enjoying the thing under your fingers for eight hours a day is a perfectly good reason on its own. Just run the protocol first, so you know what you bought.

Frequently asked questions

Will a mechanical keyboard make me type faster?

Not by itself, and not by much. The evidence for hardware effects on typing speed is thin and mostly concerns key travel distance rather than mechanical switches specifically. What is well established is that the gap between an average typist and a fast one comes from rollover, rhythm, accuracy and finger economy. If you are below roughly 100 WPM, technique has far more headroom than hardware. The exception is a keyboard that is actively broken, missing keypresses, or has almost no key travel. Run the ten test protocol above and you will have your own answer in twenty minutes instead of an opinion.

Are linear or tactile switches faster?

There is no good evidence that either is faster in ordinary typing. Tactile switches give you feedback at the moment the key registers, which some typists find reduces bottoming out and missed keys over long sessions. Linear switches are smoother and are often preferred by gamers for repeated presses. Both handle high speed typing fine. The switch pre travel study found measurable reaction time differences between switches that participants could not consciously detect at all, which is a useful reminder that how a switch feels and what it does to your numbers are separate things. Pick on comfort and noise, then verify with a test.

Is a laptop keyboard slowing me down?

Usually less than you think, and it depends on which laptop keyboard. Scissor switch boards with reasonable travel performed well in the biomechanics study, better than the very short travel boards in the same comparison. The configuration with a published speed penalty is near zero travel, around half a millimetre, which showed roughly 8 percent slower typing and 12 percent more typing force. If your laptop keys barely move, that is a real cost. If they have a normal shallow but definite press, a laptop keyboard is not what is capping your score. Posture and screen height on a laptop are usually the bigger problem.

Do fast typists all use mechanical keyboards?

No. Plenty of very fast typists work on laptop and membrane keyboards, and plenty of people with expensive mechanical boards type at exactly the average. The 136 million keystroke dataset did not identify keyboard type as a factor in what separates fast typists from slow ones. What it did identify was rollover, error patterns and consistency. The correlation you see online between fast typists and mechanical keyboards mostly runs the other way round: people who care enough about typing to get fast are also the people who care enough to buy a nice keyboard.

---

Ready to find out? Run five tests on your current keyboard and five on the other one. Every result is saved in your browser, on your device, so you can compare medians rather than trusting your memory of how a session felt. Start the 3 minute test.