The dilemma
Illustrative scenario
Owen runs online sales for a specialty coffee roaster with a subscription service. The shop page lists 38 single-origin coffees, each with five grind settings, four bag sizes and three delivery frequencies. First-time visitors linger on that page and many leave without buying.
The designer proposes a fix: a three-question taste quiz that recommends one coffee, with the full catalogue tucked behind a "See all" link. In a quick prototype test, newcomers reach the basket much faster.
But Owen's subscriber data points the other way. Returning subscribers, who bring in most of the revenue, rotate through origins and email support whenever a favorite is hard to find. The head roaster adds a brand worry: show only a handful of coffees and the shop starts to look like any supermarket shelf.
Owen has to choose a design for the holiday season, when both new gift buyers and loyal subscribers arrive at once.
The short answer
Hick's law, often called the Hick–Hyman law, says that the time needed to make a choice rises with the number of alternatives, roughly in proportion to the logarithm of that number: doubling the equally likely options adds about the same extra time each time. William Edmund Hick reported the relationship in 1952 from choice reaction-time experiments, describing decision time as a function of the amount of information in the choice.
How strong is the evidence?
UX model or heuristic
A model or design heuristic from human-computer interaction. Useful for predictions, not a guarantee of conversion.
In design, Hick's law works as a UX model: a simplified rule of thumb drawn from laboratory findings. The original experiments used simple, well-practiced tasks: a light comes on, press the matching key. The logarithmic pattern holds best when alternatives are equally likely, the person already knows the mapping from stimulus to response, and no deliberation is involved. Practice shortens the effect, and making some options more likely than others (a clear default, a popular pick) changes it. Scanning an unordered list is different: people read items one by one, so time grows closer to linearly.
Hick's law is about speed, not sales. The separate claim that more options reduce buying is the choice-overload hypothesis. Its best-known evidence is the Iyengar and Lepper jam study (see the example). A 2010 meta-analysis by Scheibehenne, Greifeneder and Todd pooled dozens of choice-overload experiments and found an average effect close to zero, with large differences between studies. Overload may occur under some conditions, such as choosers without clear preferences facing hard-to-compare options, but it cannot be assumed.
So: expect more options to slow some decisions; test whether that changes revenue.
Investigating an example
Documented study
The jam study, and the meta-analysis that followed
In a field experiment at an upscale grocery store, Iyengar and Lepper alternated a tasting display between 24 jams and 6 jams. The larger display drew more passers-by to stop, but about 30% of shoppers who stopped at the small display went on to buy a jam, compared with about 3% at the large display. The study became the most quoted evidence that more choice can demotivate buyers.
A decade later, the meta-analysis by Scheibehenne and colleagues found that choice-overload effects average out close to zero across many experiments, with some studies showing the opposite pattern.
Limits: the jam study is one store, one product category and a tasting setting; it measured purchases, not decision time, so it is not a direct test of Hick's law. The meta-analysis shows its result does not generalize reliably. Use it as a question to test, not a rule to apply.
A published study or record, linked below. Its limits are stated.
Choose your move
You are in the dilemma above. Which move would you make? Open each one to see its trade-offs.
A Lead with the three-question quiz and hide the full range behind "See all".
Newcomers get a faster path and fewer may leave. But the quiz adds steps of its own, recommendations can feel generic, and subscribers lose the browsing they rely on. The brand may start to look narrow.
B Keep all 38 coffees, add taste-profile filters and a short "start here" set, and move grind and size to sensible defaults on a later step.
Range stays visible for regulars while the first decision gets simpler. It needs good product tagging, honest defaults that some customers will change, and more design work.
C Cut the range to 12 coffees.
Simpler page and simpler operations, with real inventory savings. But it removes origins loyal buyers rotate through, and the overload evidence does not promise a sales lift to offset that.
No move wins in every business. The law tells you what to check, not what to choose.
How to apply it
- Separate speed from outcome. Measure time to first add-to-cart and conversion separately; a faster decision is only useful if it is also a good one.
- Segment by familiarity. Compare new and returning visitors. Unfamiliar choosers feel the extra options most; experienced ones often use the range.
- Structure before you cut. Group options, add filters, set defaults for low-stakes attributes and ask one decision at a time.
- Offer an honest starting point. Label a recommended or popular option only if it is true, and explain why; a clear starting point makes the choice less evenly balanced.
- Test against the full range. Run an A/B test with guardrails for average order value, returns and repeat purchase, not just conversion.
What to measure
Time to first meaningful action, conversion rate and average order value by segment (new and returning), plus repeat purchase within 60 days. Guardrail: a faster path that lowers order value or repeat purchase among existing customers is not a net improvement.
Worked example
| Equally likely options | Information in bits (log₂ of options) | Extra time compared with 2 options |
|---|---|---|
| 2 | 1 | Baseline |
| 4 | 2 | +1 unit |
| 8 | 3 | +2 units |
| 16 | 4 | +3 units |
| 32 | 5 | +4 units |
Illustrative: one "unit" stands for the extra time each doubling adds, which varies by person and task. The formula is often written with log₂(n + 1); this table simplifies it to show the shape. Going from 16 to 32 options adds the same time as going from 2 to 4. The pattern assumes equally likely, well-practiced choices; reading an unordered list behaves differently.
Where it breaks
- Lab conditions do not match shopping. The model assumes equally likely options and practiced responses; real purchases involve search, comparison and deliberation.
- More choice does not reliably reduce sales. Choice-overload effects average near zero across studies, and range can attract shoppers in the first place.
- Experienced customers are less affected, and hiding options to help newcomers can penalize the people who buy most.
- Defaults can be abused. A default that steers people into the option that pays the seller most, or is hard to change, is a dark pattern, however fast it makes the choice.
- Speed is not always the goal. For insurance, finance or health decisions, taking time can be the right outcome.
What if you apply it incorrectly?
A software vendor collapses four pricing tiers into one "to follow Hick's law." Decisions get faster, but small teams now pay for features they never use and cancel, while large customers have to book a sales call just to ask for an enterprise plan. The team optimized decision speed and made the wrong decision easier to reach.
Myth versus evidence
The social-video version
Every extra option costs you sales, so always show fewer choices.
What the evidence supports
Hick's law predicts slower decisions among equally likely options; choice-overload studies average near zero effect on buying, so structure and test before cutting.
Five-minute field assignment
Open your busiest product-listing or pricing page. Count the decisions a first-time visitor must make before buying (not the number of products), and mark which ones have a sensible default. Then pull time to first add-to-cart for new and returning visitors and note the gap.
Checkpoint
Two questions: one about the idea, one about a decision. Answer both correctly to master this law. Questions are drawn from a larger bank and graded on our server; a retry brings different questions.
Evidence ledger
Every factual claim in this lesson, where it comes from, and how far it has been checked.
-
Hick (1952) found choice reaction time rises roughly with the logarithm of the number of alternatives.
Limit: Simple laboratory reaction tasks with practiced responses.
-
In the jam study, more shoppers stopped at the 24-jam display, but about 30% of those at the 6-jam display bought versus about 3% at the 24-jam display.
Limit: One store and product category; verify the exact percentages in the paper.
-
The 2010 choice-overload meta-analysis found a mean effect close to zero with large variation between studies.
Limit: Pooled experiments up to 2009; later work continues to examine moderators.
-
The coffee roaster scenario and the doubling table.
Limit: Teaching scenario; the table shows the shape of the model, not measured times.
Further reading
- Hick, W. E. (1952). On the rate of gain of information. Quarterly Journal of Experimental Psychology, 4(1), 11–26.
- Iyengar, S. S. & Lepper, M. R. (2000). When choice is demotivating: Can one desire too much of a good thing? Journal of Personality and Social Psychology, 79(6), 995–1006.
- Scheibehenne, B., Greifeneder, R. & Todd, P. M. (2010). Can there ever be too many options? A meta-analytic review of choice overload. Journal of Consumer Research, 37(3), 409–425.
Which other law could explain this decision?
Editorial review: Ali Dezhami, November 8, 2026. Last substantive update: November 8, 2026. Our methodology