Card Sorting
UX research technique where participants organize cards representing content to reveal their mental model and structure information architecture.
Updated on January 31, 2026
Card sorting is a collaborative UX research method that involves asking users to organize cards (physical or digital) containing concepts, features, or content according to their personal logic. This technique helps understand how users categorize information and align a product's information architecture with their natural mental models. It serves as a fundamental tool for designing intuitive navigation and effective taxonomies.
Fundamentals of Card Sorting
- Three main types: open (participants create their own categories), closed (predefined categories), or hybrid (combination of both approaches)
- Based on cognitive psychology and mental model studies to understand how users structure information
- Can be conducted in-person with physical cards or remotely via specialized digital tools
- Ideal number of 30-60 cards and 15-30 participants to obtain statistically significant results
Benefits of Card Sorting
- Reduces designer bias by basing information architecture on actual user perceptions rather than assumptions
- Identifies areas of confusion or disagreement in content organization before costly development
- Significantly improves content findability and reduces abandonment rates related to complex navigation
- Facilitates team consensus through objective data from real users
- Relatively low cost compared to other UX research methods while generating actionable insights quickly
Practical Example: E-learning Platform Redesign
An e-learning platform wants to reorganize its catalog of 150 courses. The UX team conducts an open card sorting with 25 representative participants. Each participant receives 45 cards representing different courses and must organize them into logical groups, then name each group. Analysis reveals that 78% of participants group courses by professional objective ("Become a Developer", "Advance in Management") rather than by technology ("JavaScript", "Python") as the old architecture did. This discovery leads to a complete navigation restructure, resulting in a 34% increase in conversion rate and a 42% reduction in catalog page abandonment.
Implementation of a Card Sorting Session
- Define objectives: complete site architecture, section reorganization, validation of existing taxonomy
- Select appropriate type: open to explore, closed to validate, hybrid to optimize existing structure
- Prepare cards: identify 30-60 representative items, write clear and unambiguous labels, avoid technical jargon
- Recruit participants: 15-30 users representative of target audience, diversify profiles to capture different mental models
- Conduct sessions: clearly explain the task, observe without influencing, encourage thinking aloud, time sessions (typically 20-60 minutes)
- Analyze results: use similarity matrices and dendrograms to identify patterns, calculate agreement rates, detect outliers
- Synthesize into architecture: create a sitemap based on majority groupings, name categories using user terminology, test with tree testing if possible
Pro Tip
Always combine card sorting with other complementary methods: start with user interviews to understand context, then conduct card sorting for structure, and finally validate with reverse tree testing. This methodological triangulation ensures an architecture that is both intuitive and performant. Also document participants' hesitations and comments during sessions: these qualitative insights are often as valuable as quantitative data.
Related Tools
- Optimal Workshop (OptimalSort): comprehensive platform for online card sorting with advanced statistical analysis and visualizations
- Miro / FigJam: collaborative whiteboards enabling remote real-time card sorting
- UserZoom: UX research suite including card sorting with integrated participant recruitment
- UXmetrics: French specialized tool with card sorting and tree testing modules
- Physical cards + Post-it notes: effective low-tech solution for in-person sessions with direct observation
Card sorting represents a strategic investment for any organization seeking to improve the user experience of its digital products. By revealing how users actually think about information organization, this method enables the creation of architectures that reduce cognitive friction, increase engagement, and directly improve business metrics like conversions and retention. Its high ROI and relative simplicity of implementation make it an essential part of the modern UX toolkit.
