Gerald Ajam

“Good UX removes friction”

Remove the friction in the interface. Keep the difficulty in the material.

Status Position

Published September 2026

All counterpositionsDesign

Usually said

“Good UX removes friction”

The shift

From: All friction is a defect

To: Sort the friction first

More precisely

Good UX removes the effort the interface causes and protects, or deliberately adds, the effort the task requires.

In short

Most friction in an interface is waste, and removing it is good work. Some effort is the point, though. In a learning product, recalling an answer or telling two cases apart is hard on purpose. The precise version asks which effort you are removing: the interface’s or the task’s.

The argument

From the claim to a better one

  1. The strongest formulation

    Every moment a user spends working out the interface is a moment not spent on their goal. Removing that cost is the core of the discipline, and it is almost always worth doing.

  2. Why it is attractive

    It is right for most products, where the user’s satisfaction and the product’s success are close to the same number. It is also measurable: fewer steps, fewer errors, faster completion.

  3. Where it holds

    It holds for friction caused by the interface: confusing controls, split attention, hidden states, needless steps. That effort is waste in any product, including a learning one.

    Where it breaks

    It breaks when the friction is the learning. Retrieval, generation and telling similar cases apart are effortful by design, and a review process that files every difficulty as a defect will smooth them away. It also breaks where friction is protective: a pause before an irreversible action is deliberate.

  4. What evidence matters

    Bjork and Bjork (2011) on desirable difficulties, and Sweller’s cognitive load theory, which separates load caused by presentation from load intrinsic to the material. The test is not whether users found the task easy but whether they can do it later, unaided.

  5. So, more precisely

    Good UX removes the effort the interface causes and protects, or deliberately adds, the effort the task requires.

Reading it

The precise version, word by word

Good UX removes the effort the interface causes1 and protects, or deliberately adds3, the effort the task requires2.

  1. Effort the interface causes

    Work spent on the tool rather than the goal: finding a control, remembering a hidden state, reading the same thing twice.

  2. Effort the task requires

    Work that produces the result. In learning, that includes retrieving, generating an answer and comparing similar cases.

  3. Deliberately adds

    Some effort is worth putting in on purpose, such as a question before the answer is shown, or a confirmation step before something that cannot be undone.

What it changes in practice

  • In a usability review, label each difficulty as interface or task before deciding to fix it.
  • Judge a learning product on later, unaided performance as well as on ease of completion.
  • Keep deliberate pauses where a mistake is costly, and say why they are there.

A case, in general terms

A practice app shows the answer as soon as a learner hesitates. Completion rates go up and users say it is easier. It is easier, because the app has removed the retrieval that made the practice work. The fix is to keep the question hard and make everything around it easy: a clear layout, one tap to answer, feedback straight after the attempt.

Singapore