Jakob's law: why users expect what they already know
Jakob's law says users expect your site to work like the ones they already know. What it means, Nielsen's examples and when to break the pattern.
What does it say, exactly, and where does it come from?
The statement Laws of UX gives is simple: users prefer your site to work the same way as all the other sites they already know. Nielsen puts it this way in his September 2023 article, where he recalls that he formulated the law in 2000: users spend most of their time on other websites, so they expect yours to work like them. Nielsen Norman Group also spread it in a two-minute video from 2017. Visitors arrive with expectations formed on every site they use, and those expectations are the starting point of any new design.
Nielsen presents it as valid in any setting where people meet many different designs, mobile apps included. Laws of UX sums up the practical consequence: if you build on mental models that already exist, users focus on their task instead of learning something new.
It is not a style rule. It is an observation about behaviour. A person does not arrive at your product with a blank page. They arrive with habits learned from dozens of other products, and those habits are what they use to decide where to click.
Why does it matter to a designer?
Because every familiar pattern is a shortcut. The user does not have to read to know where the search is or what the bag icon means. The effort they save can go to their task. When a design breaks a pattern without a strong reason, the user has to learn something new before doing what they came to do.
This connects with the user’s mental model and with cognitive load. A known pattern lowers the load; an unknown one raises it, even when it is prettier or more efficient in theory.
What are the classic examples?
Nielsen mentions several patterns that are now standard: an open search box at the top, preferably on the right; the logo at the top left; the main navigation at the top of the page; the shopping cart icon; visible scroll bars and standard GUI widgets; pinch-to-zoom in mobile apps; and vivid, prominent call-to-action buttons, because users expect them.
Laws of UX adds a form example: toggles, radio buttons and buttons come from their physical equivalents. The user understands a toggle because they have used a light switch, and that understanding carries over to the screen.
Why not just copy the competition?
The law describes what is already common. It does not say that copying is the best strategy. Nielsen recommends adopting new conventions as they settle, and warns that a small performance gain does not justify a drastic departure. A small change that adds nothing is just friction.
Nielsen also cites a data point from an NN/g study on logos. With a left-aligned logo, 4% of users failed to reach the home page in one click; with a centred logo, 24% did. It is a difference that shows why a familiar pattern, even one that looks arbitrary, can affect the outcome.
Nielsen also applies the law to new interfaces, such as voice, augmented and virtual reality or artificial intelligence: their designers should adopt emerging conventions as they settle. He warns against assuming that new territory lets you discard all the “baggage” of the old world; old conventions should be dropped only reluctantly, and only when the limits of the old technology justified them.
When is it worth breaking the pattern?
There are cases where breaking the pattern is justified:
- When the improvement is large and measurable. A flow that reduces three steps to one can be worth the learning curve.
- When the existing pattern is an accessibility problem. A control that cannot be used with the keyboard deserves changing.
- When the product is new and has no direct competition. Then the familiar pattern is the one from the closest category your audience comes from, and it should be tested with real people.
In any of these cases, give the user a way out: a way to keep using the familiar version, even if only for a limited time. That is what Laws of UX recommends, citing YouTube’s 2017 redesign: desktop users could try the new interface, send feedback and switch back to the old one.
Example
Illustrative example: a banking app wants to replace the “transfer” icon with a circular arrow designed by its brand team. In testing, many people hesitate because the icon does not match the one in other financial apps they use. The team keeps the familiar send-arrow pattern and adds the label “Transfer” underneath. The custom icon is reserved for the brand screen, where it does not compete with a task. The change did not help the task; the label did.
For your next design: list five controls in your product and ask whether someone who uses other apps would recognize them without reading. The ones they would not recognize are candidates for review, but not necessarily for change.
Common mistakes
- Changing a familiar pattern for aesthetics, without evidence that it improves the task.
- Confusing the law with copying the competition, when what it describes is the user’s habit.
- Breaking basic navigation patterns, such as hiding the search.
- Not giving a way out for users who prefer the previous version.
- Forgetting the law applies to mobile too, where gesture patterns already carry their own expectations.
- Using the law to justify never innovating. It describes a tendency, not a prohibition.
Related in the wiki
To see how expectations connect with the order of options, read Hick’s law. The design principles help you decide when a familiar pattern serves the task and when it only serves habit.