Card sorting is a user research method that helps you understand how people categorize and group information. It provides valuable insights into users' mental models, helping you create intuitive website structures, navigation systems, and information architectures that match how users naturally think about your content.
Understanding how users mentally organize information is critical for creating intuitive digital experiences. When your website or app's structure aligns with users' expectations, they can:
Card sorting reveals these mental models directly from your users, rather than relying on assumptions or internal organizational thinking. This leads to user-centered information architectures that feel natural and intuitive to your target audience.
There are three primary approaches to card sorting:
In an open card sort, participants organize content items into categories that make sense to them and create their own category labels.
✅ Best for: Discovering how users naturally think about and group your content ✅ When to use: Early in the design process when you're exploring possible structures
With a closed card sort, you provide predefined categories, and participants sort items into these fixed groups.
✅ Best for: Validating an existing or proposed information architecture ✅ When to use: Later in the design process when refining an established structure
This approach combines elements of both open and closed sorting, typically allowing users to sort into predefined categories while also creating their own if needed.
✅ Best for: Testing a proposed structure while remaining open to new insights ✅ When to use: When you have a structure in mind but want to remain flexible
The basic process involves:
Preparation: Select content items (typically 30-60) that represent your site's content. Create cards with one item per card.
Setup: Explain the task to participants and provide clear instructions.
Sorting: Participants group the cards according to what makes sense to them.
Labeling: In open sorts, participants create and name their own categories.
Analysis: Review the results to identify patterns, examining which items were consistently grouped together and how categories were named.
✅ Select participants who represent your target users ✅ Include 15-20 participants for each distinct user group ✅ Recruit users with various levels of domain knowledge
✅ Keep card labels clear and concise ✅ Use plain language that participants will understand ✅ Include a representative sample of your content ✅ Aim for 30-60 cards (too few won't show patterns; too many causes fatigue)
✅ Provide clear instructions without influencing how participants sort ✅ Encourage participants to think aloud during the process ✅ Ask follow-up questions about their sorting rationale ✅ Document observations during the session
❌ Using internal jargon on cards that users won't understand ❌ Including too many cards, causing participant fatigue ❌ Over-influencing participants with leading questions or examples ❌ Ignoring outliers without understanding why they differ ❌ Drawing conclusions from too small a sample size ❌ Forcing a single information architecture when multiple valid approaches exist
Card sorting analysis involves looking for patterns in how participants grouped items:
Look for items that were consistently grouped together across participants, as well as items that caused confusion or were placed in multiple categories.
In-person card sorting allows for rich qualitative insights through observation and follow-up questions. It's ideal for complex content or when deeper understanding is needed.
Online card sorting enables you to reach more participants quickly and efficiently. Tools like Free Card Sort make it easy to run remote studies at scale and automatically analyze results.
Card sorting is a powerful, accessible research method that yields valuable insights for creating intuitive information architectures. Whether you're designing a new website, improving navigation, or reorganizing content, card sorting helps ensure your structure matches users' mental models.
Start with a clear goal, choose the appropriate card sorting technique, and let your users show you how they think about your content. Your information architecture will be stronger for it.
Explore related concepts, comparisons, and guides