Usability heuristics are Jakob Nielsen's 10 rules of thumb for evaluating interface design, published in 1994. They're not detailed guidelines — they're broad principles that trained evaluators use to identify usability problems without running formal user tests. Three decades later, they're still the most widely used inspection method in the field.
All 10 heuristics are useful, but two are directly relevant to information architecture and card sorting.
#2: Match between system and the real world. Your interface should speak the user's language, not your organization's internal terminology. This is the heuristic that gets violated most often in navigation design. Teams label sections based on how they think about the product, not how users think about it.
Here's what this looks like in practice: a heuristic evaluation flags that "CRM Integrations" (internal jargon from the engineering team) should probably be "Connect Your Tools" (language users actually use). That's a hypothesis. Card sorting data confirms it — when you ask 25 users to sort integration-related cards, they group them under labels like "Connect" and "Tools," not "CRM" or "Integrations."
#6: Recognition rather than recall. Users should recognize where content belongs by looking at your navigation, not have to remember from a previous visit. This means category names need to be self-explanatory. If users have to recall that "Resources" contains your pricing page (because where else would you put it?), your IA has a recognition problem.
Card sorting tests this directly. When participants in a closed card sort consistently place cards in the wrong categories, those categories have a recognition problem — the labels don't signal what's inside.
For reference, here are all 10:
Each heuristic addresses a different dimension of usability. During a heuristic evaluation, 3-5 evaluators independently review an interface against these principles, rating each issue by severity. The process takes 1-2 hours per evaluator and typically surfaces 75% of usability problems — a strong return for the time invested.
A common mistake is treating heuristic evaluation and card sorting as either/or. They answer different questions and work best in sequence.
Run a heuristic evaluation first. It's fast and cheap — you can do it with your own team in an afternoon. It will flag navigation labels that use jargon, categories that seem arbitrary, and structures that violate user expectations. These are hypotheses about problems.
Then run a card sort to validate. The heuristic evaluation suspects that "Platform Solutions" is bad labeling. The card sort proves it — zero out of 20 participants use anything resembling that phrase when creating their own categories.
The limitation of heuristic evaluation is subjectivity. Different evaluators flag different issues, and severity ratings vary based on expertise. Nielsen's own research found that individual evaluators only catch 35% of problems, which is why you need multiple evaluators and why heuristic evaluation should never be your only research method.
Card sorting provides the empirical data that heuristic evaluation lacks. Heuristic evaluation provides the speed and breadth that card sorting can't match. Use both.
What are Nielsen's 10 usability heuristics? Nielsen's 10 heuristics are: (1) Visibility of system status, (2) Match between system and real world, (3) User control and freedom, (4) Consistency and standards, (5) Error prevention, (6) Recognition rather than recall, (7) Flexibility and efficiency of use, (8) Aesthetic and minimalist design, (9) Help users recognize, diagnose, and recover from errors, and (10) Help and documentation. Published in 1994, they remain the most widely used framework for evaluating interface usability.
How do usability heuristics relate to card sorting? Two heuristics directly connect to card sorting. Heuristic #2 (Match between system and real world) says your categories should use user language, not internal jargon — card sorting reveals what that language is. Heuristic #6 (Recognition rather than recall) says users should recognize where content belongs without memorizing your structure — card sorting validates whether your categories are recognizable.
Can heuristic evaluation replace user testing? No. Heuristic evaluation identifies probable usability issues based on expert judgment, but it misses problems that only surface when real users interact with the interface. Use heuristic evaluation to catch obvious issues early and cheaply, then validate with user testing. The two methods complement each other — heuristics find about 75% of usability issues, but the remaining 25% often includes the most impactful problems.
Explore related concepts, comparisons, and guides