In short
Hick's law, from experiments by William Hick and Ray Hyman in the early 1950s, states that the time to make a decision increases logarithmically with the number of equally likely options. For interfaces, it means long menus, crowded toolbars, many pricing tiers and forms with many equal choices slow users down and raise abandonment, and that grouping, ordering and progressive disclosure reduce the effective number of choices. It applies to simple selections among similar options, not to reading or to expert users who already know what they want.
The law
In 1952 William Hick, and in 1953 Ray Hyman, measured how long people took to respond when faced with a varying number of possible stimuli, and found that reaction time grows with the logarithm of the number of choices. The usual formulation:
T = a + b * log2(n + 1)where T is decision time, n is the number of equally likely alternatives, and a and b are constants. The logarithm means the cost per option falls as the count rises; going from two options to four costs more than going from twenty to twenty-two. But it never stops rising, and in practice the effect on interfaces is compounded by a second cost the law does not measure: the confidence lost when a user is not sure which of many options is right.
It is the counterpart to Fitts's law: Fitts governs how long it takes to reach a target, Hick how long it takes to pick one.
What it implies
Navigation. A menu with seven items is scanned and chosen quickly; one with twenty is a search task. The remedy is not to hide the twenty but to group them so the first decision is among a handful of categories, each of which leads to a handful of items.
Toolbars and actions. Every equal-weight button on a screen is a choice the user has to consider. One primary action, a few secondary ones, and the rest behind a menu is faster than fifteen buttons in a row, even though the fifteen are all visible.
Pricing pages. Three tiers are chosen quickly; six invite comparison paralysis. The pricing page convention of three plans with one highlighted is Hick's law applied.
Forms. A select box with fifty options is slower than a text field with autocomplete. Radio groups over about five items become lists to read rather than choices to make.
Onboarding. Asking a new user to choose among many templates, plans or paths before they have done anything is the moment abandonment spikes; a default plus one alternative gets them moving.
Search and filters. Twenty filters shown at once is a wall; the five people actually use, with the rest available on request, is a tool.
The misreadings
"Fewer options is always better." Not what the law says. It says decisions among equally likely, similar options take longer as their number grows. An expert who knows exactly which of forty tools they want is not making a Hick's-law decision; they are recognising and reaching. Power-user interfaces with dense controls are not violations when the users are experts.
"Hide the options." Removing choices from view does not remove them from the task; it moves the decision to a place the user cannot see. The remedy is structure, not concealment.
"It applies to everything." It describes simple selections. Reading a page, comparing products on several attributes, or composing an answer are not modelled by it, and applying it to them produces silly conclusions like "articles should have three headings".
"Three is the magic number." The law is continuous. Three is often good because it is small; five, seven or nine can be right when the options are well ordered or the users know the domain.
When more options are right
When users arrive knowing what they want and need to find it, a longer, well-organised list beats a short one that forces extra steps. When the choices are genuinely distinct and equally important, hiding some misleads. When the audience is expert and uses the interface daily, density is efficiency. And when the options are ordered by something the user already knows, such as alphabetically or by date, the decision is a search with a known target and the law's cost shrinks.
Reducing the effective number of choices
- Group into categories, so the user picks a group then an item: two small decisions instead of one large one.
- Order by likelihood or by a familiar scheme, so the expected option is found early.
- Highlight a default or a recommendation, which turns a choice among many into a confirm-or-change decision.
- Progressive disclosure: show the common options, offer the rest behind "more".
- Remove what nobody chooses. Analytics usually show that most options are rarely used; cutting them costs little.
- Replace selection with input where the set is large: autocomplete over a long dropdown, search over a long list.
- Split the decision across steps when the options depend on each other, as a good checkout does.
Testing it
Hick's law predicts direction, not magnitude, in real interfaces. Time-on-task and abandonment at the decision point, measured before and after reducing or restructuring options, tell you whether the change mattered. Most reductions do; the ones that hurt are usually cases where the removed options were the ones some users needed. The usability test with five people reveals that quickly.
If a product's screens feel crowded and nobody can agree what to remove, book a call; the usage data usually makes the decision for you.
Common questions
What does Hick's law say?
That the time to choose among options grows with the logarithm of their number, so more equally likely choices mean slower decisions, with each extra option costing a little less than the last but never nothing. It comes from reaction-time experiments by William Hick and Ray Hyman in the early 1950s and is widely applied to menus, forms and other interface choices.
How is Hick's law used in UX design?
To limit and structure choices: grouping long menus into categories, keeping one primary action per screen, offering three pricing tiers rather than six, using autocomplete instead of long dropdowns, showing common options and disclosing the rest on request, and highlighting a default so a choice becomes a confirmation.
Does Hick's law mean fewer options are always better?
No. It applies to simple selections among similar options by users who do not already know what they want. Expert users who recognise their target are not slowed by dense interfaces, and hiding options moves the decision rather than removing it. The remedy is structure and ordering, not concealment.
What is the difference between Hick's law and Fitts's law?
Hick's law governs how long it takes to decide which option to choose, rising with the number of options. Fitts's law governs how long it takes to move to and hit the chosen target, rising with distance and falling with size. Together they cover most of the time cost of using an interface control.
How many items should a navigation menu have?
Hick's law argues for keeping the first-level choice small, commonly five to seven items, with further items grouped beneath them. The exact number depends on how distinct and well-ordered the items are and how familiar users are with the domain; the test is whether people find things quickly, which a tree test or usability test shows.
