Get 20% off on all plans of User Evaluation – the AI for Research. Use the code LIMITED20. Sign up today!20% off User Evaluation, the AI for research: code LIMITED20
Interaction & usability

Usability heuristics

Updated UX Pickle editorial team

Definition

Usability heuristics are broad rules of thumb for interface design. The best known are Jakob Nielsen's 10, used to review designs for problems.

Usability heuristics are general principles for interaction design. NN/g calls them broad rules of thumb rather than specific usability guidelines: they describe what good looks like without prescribing a layout. Jakob Nielsen's set of 10 is by far the most widely used, both as a design checklist and as the basis for heuristic evaluation, a quick expert review method.

Where they come from

Nielsen developed the heuristics with Rolf Molich in 1990, then refined them in 1994 based on a factor analysis of 249 usability problems (NN/g). The 10 principles are:

  1. Visibility of system status. Keep users informed about what is going on, through appropriate feedback within a reasonable time.
  2. Match between the system and the real world. Speak the users' language, with familiar words and concepts rather than internal jargon.
  3. User control and freedom. Users often act by mistake and need a clearly marked "emergency exit," such as undo and cancel.
  4. Consistency and standards. Users should not have to wonder whether different words, situations or actions mean the same thing.
  5. Error prevention. Better than a good error message is a design that prevents the problem in the first place.
  6. Recognition rather than recall. Make elements, actions and options visible to reduce memory load. See recognition over recall.
  7. Flexibility and efficiency of use. Shortcuts, hidden from novices, can speed up interaction for experts.
  8. Aesthetic and minimalist design. Interfaces should not contain information that is irrelevant or rarely needed.
  9. Help users recognize, diagnose, and recover from errors. Error messages should use plain language, say precisely what went wrong and suggest a fix.
  10. Help and documentation. It is best if the system needs no extra explanation, but help should be easy to find and focused on the task when needed.

Why they matter

The heuristics give teams a shared vocabulary for usability problems and a cheap way to catch many of them before testing with users.

How to apply them

In a heuristic evaluation, NN/g recommends that three to five people independently review the same interface, since any single evaluator will miss problems. Each notes where the design breaks a heuristic, then the team synthesizes the findings, for example with affinity diagramming, and discusses which issues hurt the experience most.

Do:

  • Review every new flow against the list, not only at the end.
  • Note which heuristic each issue breaks, so fixes can be prioritized and discussed.

Don't:

  • Treat a heuristic review as a substitute for research. NN/g is clear that heuristic evaluations cannot replace user research.
  • Apply the heuristics as rigid rules. Sometimes they pull against each other, such as minimalism versus visible options.

Common mistakes

  • Single-reviewer audits. One person finds a fraction of the issues.
  • Vague findings. Point to the exact screen and element.
  • Skipping prioritization. Without it, cosmetic issues and blockers look equally urgent.

In AI products

The heuristics apply to AI products with some new twists. Visibility of system status covers showing what an agent is doing and which sources it is reading. User control and freedom means undo for AI actions and an easy way to stop generation. Error prevention includes confirming before an agent sends, books or deletes anything. Help users recover from errors includes the case where the AI is the one that got it wrong. For AI-specific guidance, Microsoft's HAX guidelines extend this kind of checklist to human-AI interaction.

  • Recognition over recallRecognition over recall means keeping options, actions and information visible so people can spot what they need instead of remembering it.
  • Cognitive loadCognitive load is the mental effort needed to use an interface. Good design cuts the effort that does not help people reach their goal.
  • Mental modelA mental model is what a person believes about how a system works. People use it to predict what will happen when they act.
  • HAX guidelinesMicrosoft's 18 Guidelines for Human-AI Interaction: design rules for AI features, grouped by when they apply, from first use to long-term use.
  • Deceptive patternsDeceptive patterns, also called dark patterns, are design tricks that push people into choices they would not otherwise make, often to their harm.
  • Postel's lawBe liberal in what you accept and conservative in what you send: tolerate varied user input, but produce output that is consistent and precise.

Sources

  1. Nielsen Norman Group: 10 Usability Heuristics for User Interface Design
  2. Nielsen Norman Group: How to Conduct a Heuristic Evaluation

Browse the full AI UX glossary