Card sorting is a fast, participatory research method that reveals how users mentally organise and categorise information. Instead of the design team deciding what goes where, participants make those decisions themselves, producing a map of the user's mental model rather than the designer's assumptions. The method is one of the most direct and reliable ways to inform the information architecture of a website, app, or any system that requires users to find and navigate content.
What It Is
In a card sort, participants are given a set of cards, each labelled with a piece of content, a feature, or a concept from the system being designed. They are asked to group the cards in ways that make sense to them and, in most variants, to name each group they create. Open card sorting gives participants full freedom to create their own categories. Closed card sorting asks participants to sort cards into predefined categories. Hybrid sorting combines both approaches. The results are analysed to identify the most common groupings and the most natural category names.
How to Run It
Prepare twenty to forty cards representing the content or features relevant to your project. Use clear, jargon-free labels that your target users will understand. Recruit eight to fifteen participants who represent your intended audience. Sessions can be conducted in person using physical index cards or remotely using tools like Optimal Workshop or Maze. Ask participants to think aloud as they sort. After grouping, ask them to name each group and to identify any cards they found difficult to place. Analyse results by visualising agreement rates and common grouping patterns, often using a similarity matrix or dendrogram.
When to Use It
Card sorting is most valuable when designing or redesigning the navigation and information architecture of a digital product. Use it before creating a site map or navigation structure to ensure that categories reflect user expectations rather than internal organisational logic. It is also useful for evaluating an existing structure: a closed card sort can reveal where the current categories cause confusion or where users expect to find things in unexpected places.
Tips for Success
Keep the number of cards manageable. More than fifty cards leads to participant fatigue and unreliable results. Pilot your card labels with a colleague first to check for ambiguity. In remote sessions, follow up with a brief interview to understand the reasoning behind unusual groupings. Use card sorting data as an input into information architecture decisions, not as a final answer: combine it with tree testing and user interviews to build a complete picture.
