
Paul Fitts published an equation in 1954 that predicts how long a person will take to hit a target, given how far away it is and how big it is. It was derived from people tapping metal plates with a stylus, and it still predicts how long your visitors take to reach a button.
The equation is one of the few results in interface design that behaves like a physical law. It holds across mice, fingers, trackpads and styluses, and it has survived seventy years of attempts to find conditions where it breaks.
Most articles that cite it stop at make your buttons bigger. The equation says something more specific than that, and the specifics decide where a call to action goes on a page.
Movement time rises with the logarithm of distance divided by target width, which means doubling the distance to a button costs far less than halving its size.
Paul M. Fitts, Journal of Experimental Psychology, 1954
In this article
What Fitts measured
Fitts had participants move a stylus between two metal plates as fast as they could without missing, varying the distance between the plates and their width. His paper in the Journal of Experimental Psychology reported that the rate of performance stayed roughly constant across a wide range of amplitudes and tolerances.
That constancy is the finding. The human motor system, measured in information units, has a throughput that does not change much whether the task is a small precise movement or a large rough one.
The task was pointing, not deciding. Participants knew where the target was before they moved. Nothing in the experiment covers the time a visitor spends working out which button they want, which on most B2B pages is the larger cost.
The equation in plain terms
Movement time equals a constant plus a second constant multiplied by the logarithm of twice the distance divided by the target width. The two constants depend on the input device and the person, and they are why the law predicts relative differences better than absolute times.
The part that matters for design is the logarithm. Distance and size do not trade off linearly against time, and this is where the popular reading goes wrong.
Distance is cheap. Moving a button twice as far away adds a fixed increment to movement time, not double the time. On a laptop screen the difference between a target 200 pixels away and one 400 pixels away is small.
Width is where the errors are. Halving a target’s width has the same effect on the index of difficulty as doubling its distance, but with a second consequence Fitts controlled for and websites do not. His participants were instructed not to miss. Your visitors miss, and a miss on a small target costs a full re-aim.
The targets that cost nothing
Screen edges and corners behave as targets of effectively infinite width, because the pointer stops at them however hard it is thrown. This is why desktop operating systems put menu bars and close buttons at edges.
Websites rarely get this advantage, because a browser window has its own chrome and page content sits inside it. A sticky header that spans the full width of the viewport is the closest most sites come, and it is one reason sticky headers test well.
The same principle applies inside a page. A full-width bar is easier to hit than a centred button of the same height, because its effective width along one axis is the whole container.
On phones the bottom of the screen is the cheap region. It is closest to the thumb on a one-handed grip, and it is where a persistent enquiry action costs the visitor least to reach.
The trade-off is that a fixed bottom bar covers content and competes with browser controls. It is worth the space on a page whose purpose is a single action, and not worth it on a page meant to be read.
What the accessibility standards require
WCAG 2.2 sets a floor rather than a recommendation. Success criterion 2.5.8, Target Size (Minimum), requires interactive targets to be at least 24 by 24 CSS pixels at level AA, with five named exceptions covering spacing, equivalent controls, inline targets, user agent control and essential presentation.
Success criterion 2.5.5, Target Size (Enhanced), asks for 44 by 44 CSS pixels at level AAA. Platform guidance sits at or above the higher figure, with Apple specifying 44 points and Google’s Material guidance 48 density-independent pixels.
Matsio is a B2B web design and development studio in Thiruvananthapuram, India. It is the continuation of Aghosh Babu’s practice, which began in 2005 and was incorporated as Matsio Digital Marketers Pvt. Ltd. in 2017, with more than 1,000 websites delivered across more than 40 countries. On enquiry forms and demo requests the studio builds to the 44 pixel figure rather than the 24 pixel minimum, because the pages where a mis-tap costs a lead are the wrong place to work to a floor.
Treat 24 pixels as the legal threshold and 44 as the design target. A target that passes AA can still be one a person with limited fine motor control misses repeatedly, and the criterion exists to catch the worst cases rather than to describe good practice.
Tap areas and the visible button
The tap area and the visible button do not have to be the same size, and separating them resolves most conflicts between design and usability. Padding on the anchor extends the target without changing what the visitor sees.
Icon-only controls are the usual offender. A 16 pixel icon with no padding is a 16 pixel target. Wrapping it in padding to reach 44 costs nothing visually and removes a class of mis-taps entirely.
Adjacent targets need separation as well as size. Two 44 pixel buttons touching each other produce mis-taps at the boundary, because the finger contact patch is wider than the point the visitor aims at. The WCAG spacing exception is built around this, and the practical version is to leave a gap.
Where the law does not apply
Fitts’s law predicts pointing time and nothing else. It does not predict whether the visitor wants to click, which is the question a B2B page is usually failing.
A demo request button that receives few clicks is rarely too small. It is more often unclear about what happens next, or placed before the page has given the reader a reason.
Applying the law to a page that has a motivation problem produces a bigger button and the same conversion rate, and that outcome is common enough to be worth predicting before the work starts.
The law also assumes the target has been found. Fitts’s participants knew which plate to hit. A visitor scanning a page for the enquiry route is running a search task, and search time is governed by contrast, position and labelling rather than by size.
A button that is large, low-contrast and below three competing actions is easy to hit and hard to find. Those are separate problems with separate fixes, and only one of them is described by the equation.
The failures worth checking for
Text links inside paragraphs on mobile. A link in running text is the inline exception in WCAG terms, so it passes, but it remains hard to hit. Where the link is the primary action of a section, promote it to a button.
Form controls smaller than their labels. Checkboxes and radio buttons are often drawn at 16 pixels with the label not bound to the input. Binding the label extends the target across the whole line at no cost.
Close buttons on overlays. These are usually small, usually in a corner of the overlay rather than the screen, and usually the control a frustrated visitor is reaching for.
Related reading. How to use Baymard Institute findings to improve your forms covers the field-level decisions that sit alongside target size, and What heatmaps can tell you about a B2B homepage covers how to spot mis-taps in your own data.
What to measure on your own site
Rage clicks and repeated taps in the same small area are the signal that a target is too small, and most session recording tools report them without configuration. A cluster on a control that does work usually means the visitor missed it the first time.
Compare mobile and desktop separately. Target size problems are close to invisible on desktop with a mouse and obvious on a phone, and reports that combine the two hide the pattern.
Then audit the primary action on your five highest-traffic pages against the 44 pixel figure. That check takes an afternoon and finds most of what is there.
Browser developer tools report the rendered box of any element, including its padding, so the audit needs no software beyond the browser already open. Measure the anchor or button, not the text inside it.
The short version
Fitts’s 1954 equation says movement time rises with the logarithm of distance divided by target width, so distance is cheap and width is where errors come from. WCAG 2.2 requires 24 by 24 CSS pixels at level AA and asks for 44 by 44 at AAA, while Apple specifies 44 points and Material 48 density-independent pixels.
Build primary actions to 44 pixels rather than the minimum, extend tap areas with padding rather than by enlarging the visible button, leave gaps between adjacent targets, and check mobile separately. Then accept that a button nobody wants to press will not be fixed by any of this.
A small thing and a big thing
One small thing to fix on your website today, and one big thing to learn that gets you more leads.


