How do we know what belongs where on a website or application? How do users expect to find the content they’re looking for—and what logic do they apply when navigating digital environments? Card sorting helps answer these questions, offering a simple yet powerful method to organize information based on how real users think, not just how we assume they do.
What Is Card Sorting?
Card sorting is a user research technique used to inform the information architecture (IA) of a digital experience. Participants are asked to group content or features—represented as “cards”—into categories that make sense to them. This exercise can be done physically with paper or digitally through online tools.
There are two primary types:
Open card sorting: Users create their own groupings and label them.
Closed card sorting: Users organize cards into pre-defined categories.
Each method reveals different insights about how users mentally structure information.
Why It Card Sorting Matters in UX Design
At its core, UX design is about meeting users where they are. We often design from a place of internal structure—business goals, product hierarchies, internal jargon. But card sorting flips that. It reframes the design process from the user’s point of view, asking: How do they naturally look for things? What naming conventions feel intuitive? What structure feels logical to them?
Card sorting reveals patterns. It highlights misalignments between how brands structure content and how users expect to find it. It can expose redundancies, gaps in understanding, and opportunities for simplification.
When to Use Card Sorting
Card sorting is most useful in the early stages of designing or redesigning:
- A website with a complex navigation
- A product with evolving feature sets
- A content-heavy platform like an intranet or knowledge base
It’s especially helpful when there’s uncertainty around categorization or when internal stakeholders disagree on structure. The data from users helps guide decisions with objectivity.
How We Use It at VERSIONS
In our work, we often integrate card sorting as part of our broader discovery and UX research phase. When working on digital ecosystems for enterprise clients or startups building from scratch, we invite real users—internal, external, or both—to participate. The results help us create wireframes, navigation systems, and content hierarchies that align with actual user behavior, not assumptions.
We often follow up card sorting with tree testing—a way to validate the new structure before full design begins. This ensures the IA not only makes sense on paper but functions as expected in the wild.
Inside the Sort: Technical Practices and Examples
A common question we get is: What exactly goes on the cards? The answer depends on the type of project, but the principle stays the same—each card represents a piece of content, feature, or concept users would expect to find in your interface.
For example:
On an eCommerce site, cards might include: “Track My Order,” “Return Policy,” “Gift Cards,” “Store Locator,” “Live Chat.”
For a SaaS dashboard, you might use cards like: “Analytics Overview,” “Manage Users,” “Billing Settings,” “Integrations,” “Support Tickets.”
In a healthcare portal: “Schedule Appointment,” “View Lab Results,” “Find a Specialist,” “Insurance Information,” “Prescription Refills.”
Tip:
Cards should be clearly written in everyday language. Avoid internal terminology—use what users actually see or expect to see.
How to Perform the Sort
You can run a card sort using physical index cards, sticky notes, or digital platforms like OptimalSort or UXtweak. Participants are asked to sort the cards into categories that feel logical to them. In open sorts, they also name each group, giving you direct insight into mental models and labeling expectations.
Important best practices:
Limit the total number of cards to around 30–40 to avoid participant fatigue. Recruit participants who reflect your actual user base (not just internal team members). Encourage participants to think aloud, if the session is moderated—this provides extra context.
Drawing Conclusions: Beyond the Piles
After collecting results, it’s time to analyze. You’ll notice some cards consistently land in the same groupings—that’s your signal for a strong pattern. But conclusions should never be based solely on frequency.
There are two key approaches:
Elimination system: Remove outliers or fringe groupings that don’t appear consistently. These may represent individual bias rather than a pattern.
Collection system: Aggregate overlapping categories across users to find the most common or intuitive labels and groupings.
Be cautious not to cherry-pick data to fit a pre-existing structure. One workshop rarely gives you a definitive answer—run multiple sessions, ideally with different participant segments. This helps account for edge cases and differences between user roles or personas.
How to Avoid Misinterpretation
- Don’t overfit your architecture to one group’s thinking—look for consensus, not perfection.
- Avoid assuming label preferences from one participant apply universally.
- Watch for over-grouping. Sometimes too many items in one pile means the category needs to be broken down further.
Document your findings, but don’t rush to final decisions. Use the card sort as one data point in a larger UX strategy that includes user interviews, tree testing, and real-world usability testing.
Beyond the Sort: The Bigger Picture
Card sorting isn’t just about grouping content. It’s about empathy. It’s about shifting the conversation from what we want to say to what users need to hear—and how they need to find it. When layered into a human-centered design process, it helps deliver digital experiences that feel effortless to navigate and naturally aligned with user intent.
Final Thoughts
Tools and trends come and go, but understanding how people organize information—how they expect things to work—is timeless. Card sorting is a reminder that structure isn’t something we dictate; it’s something we uncover. And when we listen closely, users will show us the way.