Jakob's Law

Jakob's Law, attributed to usability researcher Jakob Nielsen, states that users spend most of their time on other sites, and therefore prefer a site to work the same way as the ones they already know. It is a claim about transferred expectation: people arrive with mental models built elsewhere, and an interface that matches those models requires no learning, while one that deviates forces relearning at exactly the moment the visitor is least invested.

The practical implications are conservative and frequently unpopular with design teams. The logo in the top left should return to the homepage. The cart icon belongs in the top right. Underlined blue text should be a link. A magnifying glass means search. Checkout should proceed in the sequence people have completed hundreds of times elsewhere. Navigation should look like navigation. These conventions carry no inherent merit, but their familiarity is a genuine asset, and discarding it costs real conversion in exchange for distinctiveness that most visitors did not ask for.

The law explains a persistent tension in commercial design work. Novel interaction patterns are professionally rewarding, win awards, and demonstrate craft. They also require every visitor to learn something, and most visitors will not invest the effort, because they came to accomplish a task rather than to appreciate an interface. Sites that pursue distinctiveness in structural elements, such as navigation, scrolling behavior, and purchase flow, generally pay for it, while sites that pursue distinctiveness in visual identity, tone, photography, and content usually gain from it without the cost.

There is a legitimate boundary to this conservatism. Conventions change, and if nobody deviated, interfaces would never improve. The reasonable position is to innovate where the potential gain is large and the deviation is testable, and to conform where convention is well established and deviation buys nothing. It is also worth distinguishing between a convention that is genuinely universal and one that merely dominates within a particular design community, which is a distinction that internal teams frequently get wrong because their reference set is other professionals rather than their actual customers.

The law also has direct consequences for redesigns. Because expectation is built from repeated exposure, existing customers have mental models of the current site specifically, and a major redesign violates those models even when the new design is objectively better. The resulting temporary dip in performance, sometimes called change aversion, is not evidence that the redesign failed, and misreading it as such has caused many organizations to revert genuinely superior work. Measurement over a longer horizon, with new and returning visitors analyzed separately, is the correct response.

The law also implies that conventions should be verified against the audience rather than assumed from the design team's own habits. Expectations differ substantially between markets, age groups, and levels of digital experience, and the patterns that feel universal to a professional working in the field are often specific to a narrow population. Payment methods, address formats, name structures, navigation habits, and the meaning attached to particular icons all vary by country, and interfaces built on assumptions from one market frequently perform poorly in another for reasons that look inexplicable from the inside. Testing with participants from the actual target markets, rather than generalizing from a single one, is what turns this from a plausible principle into an accurate one.

In consultancy practice, applying Jakob's Law usually means resolving arguments about where distinctiveness should live. A product design engagement will typically hold structural patterns close to convention while investing differentiation in content, identity, and the quality of the specific interactions that matter most to the proposition. Where an existing site has deviated from convention in costly ways, those deviations are among the most common high-impact findings in a UX audit, because they are usually invisible to the internal team that has adapted to them.