Understanding users’ needs is not an option, it is a necessity. According to PwC, 32% of customers stop interacting with a brand they love after a single negative experience. In this scenario, investing in User Experience (UX) becomes crucial. Every design decision should be based on real data and relevant insights. Among the most effective techniques for designing a user-centered information architecture is Card Sorting.
It is a simple yet powerful methodology that lets you define a clear, consistent and intuitive structure for websites and apps. But how does it work? When should you apply it? And what are the best practices to follow?
In this complete guide we will explore Card Sorting step by step, drawing on some of the leading international sources in the UX field.
What is Card Sorting
Card Sorting is a research technique used to understand how users group and classify information. Users receive a set of “cards”, each representing a piece of content or a feature, and are asked to organize them into logical categories.
Depending on the mode (open or closed), you collect data useful for defining or validating the information architecture of an interface.
Card Sorting is used to:
- build effective sitemaps;
- design intuitive menus;
- improve the findability of content;
- align classifications and terminology with users’ language.
Why Card Sorting is fundamental for UX
A good information architecture is essential to the user experience. According to Nielsen Norman Group, 76% of users leave a site within 5 seconds if they cannot find what they are looking for.
A confusing structure creates frustration, raises the bounce rate and lowers conversions. By contrast, an architecture designed from data:
- makes navigation easier;
- improves understanding of the system;
- strengthens the user’s sense of control.
Card Sorting makes it possible to build information structures that reflect users’ real mental map, going beyond internal assumptions.
Card Sorting and Information Architecture
Information Architecture (IA) is about the organization, structuring and labeling of content in digital environments.
A good IA aims to make content:
- findable;
- accessible;
- understandable;
- discoverable.
Through Card Sorting, it is possible to design sitemaps and menus starting directly from the voice of end users, ensuring greater consistency between what the user expects and what the system offers.
Types of Card Sorting
Open Card Sorting
In this mode, users freely create groups of content and name each category. It is ideal in the early stages to explore users’ mental model.
Advantages:
- Reveals natural language and spontaneous logic;
- Offers rich qualitative insights;
- Useful for generating new structures.
Closed Card Sorting
Users receive predefined categories and only have to assign the cards to one of the groups. It is ideal for validating existing or hypothesized structures.
Advantages:
- Easier to analyze;
- Great for testing design alternatives;
- Allows standardized comparisons.
Hybrid Card Sorting
A mixed mode combining open and closed: users start from suggested groups but can create new ones. Perfect for evolving scenarios.
How to run an effective Card Sorting
1. Define the goals
Decide what you want to achieve: explore new categories, validate an existing IA, gather language insights. This will determine the choice between open, closed or hybrid mode.
2. Create the cards to sort
Prepare a set of 30–60 items, with clear, simple, unambiguous names. Each card represents a piece of content or a feature relevant to the interface.
3. Select the right panel of users
Recruit on-target users, with characteristics consistent with the final audience. Selection criteria can include:
- age;
- role;
- digital familiarity;
- goals regarding the product.
Tools such as screening surveys, CRM or professional UX panels help find the ideal participants. The recommended number is between 15 and 30 users.
4. Run the Card Sorting in person or online
Sessions can be run:
- live, with physical materials (sticky notes, whiteboards);
- remotely, with tools such as OptimalSort, UXtweak or ProvenByUsers.
Make sure to:
- give clear instructions;
- invite participants to think aloud;
- collect observational insights in addition to the data.
5. Analyze the results
With open sorting, look for patterns in the groups created and the terminology used. With closed sorting, analyze the agreement rate and use clustering tools or dendrograms to visualize affinities.
When to use Card Sorting
Card Sorting applies effectively at several stages:
- project start: to define a new IA;
- redesign: to restructure existing sites/apps;
- validation: to test architectures before release;
- participatory design: to involve teams or stakeholders.
It is not recommended for:
- extremely technical or taxonomic content;
- very complex systems with over 100 items;
- environments subject to strict regulation.
Advantages and limits of Card Sorting
Advantages
- Inexpensive and quick;
- User-based;
- Accessible even to non-UX teams;
- Adaptable in agile contexts;
- Easy to combine with other methods.
Limits
- Does not indicate content priority;
- Does not measure understanding of labels;
- Sensitive to bias and fatigue;
- Needs a minimum number of users for statistical validity.
Best practices for Card Sorting
- Limit the number of cards to a maximum of 60;
- Use simple, understandable language;
- Do not steer users’ grouping;
- Integrate the results with other tools such as tree testing;
- Analyze both quantitative data and qualitative insights.
Case study: Card Sorting in a Then project
During a project for a SaaS platform in healthcare, we used Card Sorting to redesign the area reserved for professionals. The analysis revealed a considerable discrepancy between the structure proposed by the internal team and the one that came more naturally to users. The result? A complete reorganization of the menu and a 28% increase in how easily the main content could be found.
This example shows how Card Sorting enables evidence-based decisions, concretely improving the user experience.
Conclusions
Card Sorting is one of the most effective tools for designing user-oriented information architectures. Easy to implement, adaptable to different contexts and able to deliver valuable insights, it is an ideal starting point for any UX project.
In our daily work, Then’s experts regularly integrate it into redesign, product development and navigation optimization projects. An approach that translates into more usable interfaces, more accessible content and more satisfied users.
