Hick's Law and choice overload: why fewer options often sell more

Picture a small bakery's website. The home page has eleven buttons above the fold: order online, catering, gift cards, our story, the blog, three product categories, a newsletter box, careers and contact. Every link is useful, and the owner is proud of it. Yet few visitors place an order, and the owner can't see why.

Nothing on that page is wrong on its own. Together, the options ask the visitor to make a decision before they have a reason to. That is the problem Hick's Law describes, and it is one of the most useful ideas a non-designer can learn.

What Hick's Law says

In 1952 the British psychologist William Edmund Hick ran a simple experiment. He set up ten lamps, each paired with a key, and measured how long people took to press the right key when a lamp lit up. The more lamps that could light up, the longer people took. The time did not grow in a straight line, though. It grew logarithmically, which means each extra option adds a little less delay than the one before (Hick, 1952).

A year later the American psychologist Ray Hyman ran his own version and tied reaction time to the amount of information in the choice (Hyman, 1953). That is why you will also see it called the Hick-Hyman Law.

The usual way to write it is:

T = b × log₂(n + 1)

T is the average time to decide, n is the number of equally likely options, and b is a constant you measure for the task (Wikipedia summary of the law). You do not need the formula to use the idea, but it shows one thing clearly. Here is the log part of the formula for a few menu sizes:

Going from 2 options to 16 does not make a decision eight times slower, but it does make it roughly two and a half times slower on this model. On a web page, where people are already deciding whether to stay at all, that extra hesitation adds up.

The designer Jon Yablonski puts it in one line on his Laws of UX site: "The time it takes to make a decision increases with the number and complexity of choices" (Laws of UX). Notice the word "complexity". Ten simple, familiar options can be quicker to scan than four confusing ones.

Where Hick's Law stops working

Hick measured people reacting to lights, not shoppers browsing a menu, so be careful about stretching it.

• It is about the time to decide, not whether people buy. Hick's Law predicts hesitation. It says nothing directly about sales.
• It assumes the options are equally likely. If most visitors want one thing, they find it fast however many other options sit beside it.
• It does not hold for searching an unfamiliar list. When people scan a random list item by item, the time rises roughly in a straight line with the number of items, and the logarithmic curve no longer fits. A sorted list of names people already know, like an alphabetical country menu, is different, because people can jump straight to the right part (Wikipedia summary of the law).

So the law gives you a direction (fewer, clearer choices are faster) and a warning (structure matters as much as count).

The jam study, and the debate that followed

The claim that fewer options "sell more" comes mostly from a different line of research, usually called choice overload.

The famous example is the jam study by Sheena Iyengar and Mark Lepper, published in 2000. On two Saturdays at an upmarket grocery store in California, they set up a tasting booth that showed either 24 jams or 6 jams, swapping the display every hour. Of the 242 shoppers who passed the big display, 60% stopped to taste. Of the 260 who passed the small display, only 40% stopped. So the big range was more attractive. But when it came to buying, nearly 30% of the people who stopped at the 6-jam booth went on to buy a jar, against 3% of those who stopped at the 24-jam booth (Iyengar & Lepper, 2000).

That is a striking result, and it has been quoted in thousands of marketing talks. You should also know the rest of the story.

In 2010 Benjamin Scheibehenne, Rainer Greifeneder and Peter Todd pooled 63 conditions from 50 experiments with 5,036 participants. The average effect of choice overload across all of them was "virtually zero", with large differences between studies (Scheibehenne, Greifeneder & Todd, 2010). In some studies more choice hurt, in others it helped, and in many it made no difference.

Then in 2015 Alexander Chernev, Ulf Böckenholt and Joseph Goodman ran a second meta-analysis of 99 observations (N = 7,202). They found that choice overload is real when certain conditions are present, and named four of them (Chernev, Böckenholt & Goodman, 2015):

1. The options are complex: many attributes to compare and no clear winner.
2. The task is hard: little time, unclear information, many steps.
3. People do not know what they want, because they have no prior preference or expertise.
4. People want to spend as little effort as possible, as when they are browsing.

My honest summary: a long list does not automatically lose sales. A long list of similar, hard-to-compare options, shown to people who are not sure what they want, often does. That describes most landing pages, pricing pages and menus I see.

A worked example: one landing page, before and after

Let us go back to the bakery. Here is the top of the page as it is, and how I would change it.

Nothing was deleted. Every page still exists. What changed is the number of things competing for the first decision. Look at it through the four conditions above: the visitor is browsing, does not know the full range, and was facing eleven paths that looked equally important. The "after" version removes most of that work.

I won't promise you a percentage for a change like this, because the only number that counts is the one you measure on your own site (more on that below).

Ways to cut choices without losing anything

You rarely need to remove options. You need to decide when people see them.

• Pick one main action per screen. Make it look different from everything else. Secondary links can stay, quieter.
• Recommend a default. On a pricing page with three plans, mark one as "Most popular" or "Best for most people". A recommendation turns a comparison into a yes or no.
• Show a short list first, the full list on request. Designers call this progressive disclosure: revealing extra options only when someone asks for them. Laws of UX describes Slack doing this, hiding most features from new users and introducing them after they have learned to send a message (Laws of UX).
• Group what remains. Eight items in two labelled groups of four are easier than eight in a row, because people can rule out a whole group at once.
• Make options easier to compare. If you must show many, line them up on the same few attributes, such as price, size and who it is for. This attacks the "complex options" condition directly.
• Sort by something people already know. Alphabetical, price low to high, or size. Remember the exception above: familiar order keeps long lists fast.