Fitts's law says the time it takes to hit a target grows with its distance and shrinks with its size: in the common form, movement time = a + b × log2(distance ÷ width + 1). For conversion work that means primary buttons and menu links should be big and close to where the thumb or cursor already is.
It predicts how fast people can click, not whether they want to, so A/B tests of "bigger, closer" buttons win often but not always.
Fitts's law is one of the classic laws of UX, along with Hick's law, Miller's law and Jakob's law.
Below is what each one actually says, where it came from, and what real A/B tests in our database show when stores and SaaS teams applied it.
Laws of UX: the four that matter most for conversion
"Laws of UX" is a loose label for findings from experimental psychology and usability research that designers use as rules of thumb.
The name was popularized by Jon Yablonski's Laws of UX site, which now lists 30 principles, and the companion O'Reilly book (2nd edition, 2024).
They are not all the same kind of thing.
Fitts's and Hick's laws are lab-measured models with formulas.
Miller's law is a memory finding that is often misquoted.
Jakob's law is an expert observation, not an experiment. Treat them with matching levels of trust.
| Law | Origin | What it says | Typical CRO use |
|---|---|---|---|
| Fitts's law | Paul Fitts, 1954 | Time to reach a target depends on distance ÷ size | Button size, sticky CTAs, tap targets, menu layout |
| Hick's law (Hick-Hyman) | W. E. Hick, 1952; Ray Hyman, 1953 | Decision time rises with the log of the number of equally likely choices | Fewer nav items, fewer plans, grouped options |
| Miller's law | George Miller, 1956 | Short-term memory span is limited (about 7 items in Miller's data; about 4 chunks in later work) | Chunking forms into steps, grouping specs |
| Jakob's law | Jakob Nielsen, 2000 | Users expect your site to work like the other sites they use | Keeping conventional headers, carts, form controls |
Fitts's law: the formula and a worked example
Paul Fitts published the original result in 1954 in the Journal of Experimental Psychology, using information theory to measure how fast people could tap between two targets.
The version most used in human-computer interaction is the Shannon formulation from Scott MacKenzie's 1992 review:
MT = a + b × log2(D / W + 1)
- MT: movement time.
- D: distance from the starting point to the center of the target.
- W: width of the target along the direction of movement.
- a, b: constants measured for a given device and person. The log term is the "index of difficulty" in bits.
Worked example. A mobile add-to-cart button is 40 px tall and sits 400 px below where the thumb rests.
Index of difficulty = log2(400 ÷ 40 + 1) = log2(11) = 3.46 bits.
Double the height to 80 px and it becomes log2(6) = 2.58 bits.
Make it sticky so it sits 100 px from the thumb, at 80 px, and it is log2(2.25) = 1.17 bits.
With illustrative constants of a = 0.2 s and b = 0.1 s per bit, those three designs take about 0.55 s, 0.46 s and 0.32 s to reach.
The gains are fractions of a second. That matters for frequent actions and for people who struggle with small targets, but it rarely explains a double-digit conversion lift on its own.
Platform guidelines turn the law into minimums. Apple's Human Interface Guidelines give 44x44 pt as the default control size on iOS and 28x28 pt as the minimum.
Google's Android accessibility guidance recommends 48x48 dp. WCAG 2.2 success criterion 2.5.8 sets 24 by 24 CSS pixels as the AA floor, with spacing exceptions.
What A/B tests show for Fitts's law
Making targets bigger and easier to reach has a good record when the old targets were genuinely hard to hit:
- Oliver Kenyon's team found user frustration with small, hard-to-tap mobile categories and rebuilt the menu to open full screen with large touch zones. Result: +$50,568 per month, with the category hierarchy also reorganized.
- Locations Hawaii renamed its "Inquire" button to "Contact Agent", set it in larger type and moved it higher: +22.97% inquiry conversion rate over 26 days and 37,000 visitors. Copy changed too, so size alone gets only part of the credit.
- Safety Gear Pro enlarged its mobile search bar and saw +8.33% purchases (VWO case study, no significance reported).
- Conversion Fanatics replaced tap-to-expand option groups with expanded, easier-to-tap key options: +36.6% mobile conversions. An earlier version that expanded every group lost about 4%. Easier to tap only helped when it did not also add clutter.
Sticky buttons, the most common Fitts's-law tactic, are mixed.
Blend Commerce's desktop sticky add-to-cart bar for PerTronix lifted conversion rate 25.95%.
A Clean Commit mobile bottom sticky add-to-cart raised add-to-carts 6.8% but conversion rate fell 7.7% and revenue per visitor fell 22.1% (not significant).
Another Clean Commit test that fixed the mobile filter button to the bottom of the screen lost 16.0% conversion rate.
Faster to reach is not the same as the right thing to reach.
See sticky add-to-cart tests for the full set.
Blend Commerce
Conversion Fanatics
Clean CommitClean CommitConversion FanaticsHick's law: fewer choices, faster decisions
W. E. Hick's 1952 paper had people press keys in response to one of up to ten lights.
Reaction time rose with the logarithm of the number of choices, and Ray Hyman's 1953 follow-up showed it tracks the information in the choice, in bits. A common form is:
RT = a + b × log2(n + 1)
Going from 10 equally likely options to 5 drops the log term from 3.46 to 2.58 bits, about a quarter less.
Two limits matter. The law covers choices among options people already understand, so it says little about scanning an unfamiliar list of menu labels, where reading time adds up item by item.
And it assumes equally likely options, which nav menus almost never are.
The tests that cut choices tend to cut irrelevant ones:
- Harvard Business Services trimmed a 10-tab homepage nav, removing Blog, About Us and Contact Us, adding "How to Incorporate" and renaming two tabs: +15.68% completed orders across about 32,000 visitors.
- Peter Christensen removed the 'Track Order' link from the main nav and reported +30% revenue per visitor and +24% conversion rate at 98% significance, after only 6 days.
- Blend Commerce replaced a long mobile menu of overlapping categories with broad ones: +11.11% revenue per visitor.
For more on how choice count affects buying, see the paradox of choice and navigation menu tests.
Miller's law: chunking, not "seven menu items"
George Miller's 1956 paper observed that people can hold roughly seven items, plus or minus two, in immediate memory, and that grouping items into chunks stretches that limit.
Later work by Nelson Cowan put the core capacity at about four chunks.
The common UX reading, "never show more than seven menu items", is a misuse.
A visible menu does not need to be memorized.
The useful idea is chunking: break long tasks into groups so people hold less at once.
That is the logic behind multi-step forms and reducing cognitive load.
HomEquity Bank split a one-page inquiry form into seven steps for +154% conversion rate, then cut seven steps to four for another +53%.
More chunks was not always better. Performics India cut a paint cost calculator from five steps to three, grouping paints into Economy, Premium and Luxury: +40% leads in a one-week split URL test.
Jakob's law: why familiar beats clever
Jakob Nielsen stated it in the July 2000 article End of Web Design: "Users spend most of their time on other sites.
This means that users prefer your site to work the same way as all the other sites they already know."
It is a design principle drawn from usability studies, not a controlled experiment with an effect size, so it predicts direction rather than size.
In practice it means keeping the conventions people learned elsewhere: a cart icon top right, a logo that goes home, a search box that looks like one, radio buttons for single choices.
The tests where teams broke conventions are the clearest evidence:
- Clean Commit replaced a multi-row header with a single-row header and an image mega-menu. The redesign lost 22.5% conversion rate and 12.3% revenue per visitor, with a 99.5% probability the original was better.
- Uncommon Knowledge tested a simpler, trend-style opt-in page against its older magazine-style page. The original beat the redesign by 19.55% at 99.99% significance, with an audience skewed 45 and over.
- Titan Casket showed purchase options as green tabs that could look like buttons. Switching to standard radio buttons with benefits visible gave +24.6% conversion rate, per Blend Commerce.
Jakob's law does not mean "never change anything."
A Clean Commit test that reverted a mobile menu to an older layout did worse, not better (-5.7% conversion rate, not significant).
The convention that matters is the one users bring from other sites, not your own old design.
Using UX laws in A/B tests without overclaiming
These laws are good at telling you where to look and weak at telling you how much you will gain.
Here is how to use each one, and when it is the wrong tool.
| Law | Good hypothesis source when | Wrong tool when |
|---|---|---|
| Fitts's law | Session recordings show missed taps, rage taps, or long scrolls back to the buy button | People reach the button fine but do not want to click it; the problem is value, price or trust |
| Hick's law | Menus or pricing pages list many similar, low-use options | Options are unfamiliar and need reading; or removing them hides what people came for |
| Miller's law | Long forms or comparisons ask people to hold many facts at once | Used as a hard "7 items" cap on visible lists |
| Jakob's law | A redesign breaks patterns users know from big retailers or apps | Used to block every new idea; your audience's expectations may differ from the web at large |
Three habits keep the tests honest:
- Measure the final outcome. Easier clicks often lift the click and not the sale. The Clean Commit sticky button raised add-to-carts while revenue per visitor fell, and JellyTelly's nav removal gave +105% visits to the sign-up page but only +5% sign-ups.
- Isolate the law where you can. Many winners above changed size, copy and position at once, so they support the idea without proving the mechanism.
- Watch for early noise. The Conversion Fanatics expand-all test led by 40% early, then finished about 4% down. Locations Hawaii's winners lost heavily for the first 10 days.
The laws themselves hold up well in the lab.
Fitts's and Hick's results have been reproduced for decades, and the magic-number-seven claim has been revised down rather than dropped.
What does not replicate cleanly is the jump from "faster to click" to "more revenue."
That has to be tested on your own traffic.
Browse more psychology principles or search every test in the database.