Design is typically understood as the discipline of making things easier, but some of the most thoughtful design in existence does exactly the opposite: it introduces deliberate friction, raises the effort required to complete an action, and makes getting from A to B harder than it needs to be.

The reason is usually the same; The action in question is one that users should not be able to take by accident, and the cost of an accidental execution is too high to risk a comfortable interface.

Tricky basement doors

Research into building evacuations has revealed a counterintuitive pattern: in an emergency, people moving quickly down a staircase sometimes continue past the ground floor and into the basement, because the staircase gives no indication that the exit level has been reached. For most emergencies, this is merely disorienting. For a fire, it can be fatal. The design response to this finding is to make basement access noticeably harder and visually different from the rest of the staircase. A different door type, a different handle mechanism, a change in the visual environment at the basement threshold: all of these introduce enough friction to interrupt automatic behaviour and prompt a moment of attention. The difficulty is the point. An exit that requires no thought to use is fine for routine use; an exit that needs to be deliberately located is better for emergencies, because it prevents the casualty that results from moving without thinking.

Nuclear launch protocols

Fictional portrayals of nuclear arsenals tend to imagine the launch sequence as a single dramatic button - satisfying for narrative purposes, catastrophic for any serious safety analysis.

Real launch procedures are, by contrast, deliberately multi-stage, multi-authorised, and slow. A typical sequence involves physical safeguards on the weapon, multiple independent authentication steps, confirmation through separate command structures, and waiting periods built into the process.

This is friction by design. The goal is not to make launching nuclear weapons difficult for operational reasons; the goal is to make launching them by accident, under duress, or through a single point of failure as close to impossible as the system can make it. Every additional step is an opportunity for a mistake to be caught, for a cooler moment to intervene, or for a chain of authorisations to break.

The lesson for digital design is not that products should be this hard to use - it is that when the cost of an error is high enough, friction becomes a feature rather than a failure.

Just out of reach

A school whose grounds adjoin a busy road has a straightforward but serious design problem: children, given a door at their height with an accessible handle, will open it. The solution is equally straightforward: latches, bolts, and door handles placed well above a child's reach. The door becomes harder to use, and that difficulty is precisely the requirement.

This pattern - restricting access by placing controls outside the reach of the users for whom access would be dangerous - appears in many physical and digital contexts. Child-resistant packaging on medication is the pharmaceutical equivalent; age verification flows are the digital one. The design inconveniences the adult to protect the child, and that trade-off is, in context, the right one.

Good design is always relative to who it is for; what functions as friction for one user can function as safety for another.

Multi-factor authentication

A single password securing an account is one point of failure, as anyone who knows or obtains it has complete access. The solution that has become standard across most authentication systems is to add a second factor - typically something the user has (a phone, a hardware key) rather than something they merely know.

Two-factor authentication is harder to use than single-factor; it requires an additional device, an additional step, and occasionally an additional moment of frustration.

The friction it introduces is proportional to what it protects. For an account securing critical data, financial records, or communications that should remain private, the minor inconvenience of a second authentication step is an appropriate cost. The design principle at work here is that ease and security are in genuine tension, and where the stakes of a breach are high, the right calibration leans toward security even at the cost of fluency. The mistake is applying this logic indiscriminately - a low-stakes account gated behind multiple authentication layers imposes a cost that the stakes don't justify.

Deletion confirmation passphrases

Some of the most damaging actions in digital products are not errors of misunderstanding but errors of habit. A user who has clicked "confirm" on dozens of dialog boxes may click it one more time without reading the prompt - and delete something they cannot recover. The deletion confirmation passphrase is a design response to this problem: instead of asking users to click a button, it asks them to type the name of the thing they are about to delete.

The act of typing breaks the habit loop. It requires active engagement with the action rather than passive confirmation of it, and it makes the specificity of the action legible - you are not confirming an abstract "delete" but the deletion of a thing with a particular name. Webflow uses this pattern for site deletion; GitHub uses it for repository deletion; several database management tools use it for schema-level actions. The pattern is, in the ordinary sense, worse UX than a simple confirmation button - it is slower, it requires more effort, and it occasionally frustrates users who know exactly what they are doing. But for the small set of actions where the cost of an unintended execution is unrecoverable, that friction is worth considerably more than the fluency it sacrifices.

In Summary

Deliberate friction is a tool, not a failure. The question is not whether difficulty belongs in design — sometimes it clearly does — but whether the difficulty being introduced is the right kind, at the right level, for the right action. Friction that prevents catastrophe is good design. Friction that simply fails to consider the user is not. The difference between the two is the intention behind it.