Guide · From the book

Why task success doesn’t mean users understood

Task success shows that people reached the right screen, not that they understand what the system did. People often succeed by following visible cues or memorized steps, so they fail again when the interface or the situation changes. To see understanding, ask people to predict the outcome before it appears and explain it afterwards.

Why success and understanding come apart

Most usability tests record whether a participant reached the right outcome and how long it took. If most participants succeed, the interface is judged usable. But success can come from clicking the only obvious button, tracing a highlighted path, or exploring until the right option appears. None of these requires a model of how the system works.

Cognitive psychology has documented this for decades: people can get better at a task without gaining knowledge of the rules that govern it (Berry and Broadbent 1984). An employee may know that a report needs one menu, two options and a confirm button, but not why. When the menu moves, the procedure breaks and has to be rebuilt from scratch.

What each common metric misses

  • Success rate records the outcome, not whether the person knows what produced it.
  • Time on task rewards memorized sequences, which are closer to muscle memory than to understanding.
  • Error rate shows that something went wrong, not why. A misread term, a wrong model of the workflow and a wrong model of the data all produce the same error.

The book calls the combination of low effort and low understanding easy but opaque: every performance metric reports success while the person cannot say what happened. It is the condition that standard testing is least able to see.

What it looks like in practice

In a pilot of the Everyday Services Study, participants used bank, medical booking and license renewal prototypes written in institutional language. They finished at normal speed and reported no unusual difficulty. Their explanations told a different story: “it said the protocol was filed”, “it showed stage 2 of 6”. They quoted the system back to itself. That is what high cognitive distance sounds like. The pilot was small and qualitative, so it illustrates the pattern rather than proving it.

The second question to add to every test

Usability asks: can users complete the task? Cognitive distance adds: do users understand what the system did, and why? You can test the second question in the same session with three additions:

  1. Predict before the result. After the person confirms, and before the result appears, ask what they expect to happen. Correct prediction beforehand is evidence of a causal model; a correct answer afterwards may only repeat the screen.
  2. Explain afterwards. Ask “In your own words, what just happened?” and score the answer against a short rubric.
  3. Ask if it is done. “Is the task finished?” with three options: certain, think so, not sure.

The answers feed the Cognitive Distance Index, which turns them into a 0 to 100 score you can compare across versions.

Questions people ask

Is task success rate a bad metric?

No. It is necessary but not sufficient. Keep it, and pair it with a measure of understanding, such as prediction and explanation questions, so you can see when people succeed without knowing why.

What is procedural learning in UX?

It is learning a sequence of actions that reliably produces a result, without understanding why the system works that way. It looks like competence until the interface changes, and then it breaks.

How do you know if users understood a flow?

Ask them to predict the outcome before it is shown, to explain it afterwards in their own words, and to say whether the task is done. Agreement between the three is good evidence of understanding.

Read more in the book

This guide is drawn from From Cognitive Load to Cognitive Distance by Izaias Cavalcanti. The whole book is free to read online and as a PDF.

Cite this guide

Cavalcanti, I. (2026). From Cognitive Load to Cognitive Distance: Reframing Usability and Accessibility in Information Systems. https://book.izaias.xyz/