How to Use Card Sorting for UX
Card sorting is a UX research technique in which users organize topics into groups. It is used to create or improve an Information Architecture (IA) that suits users’ expectations (following their mental model). It helps to understand how users think about content. Card sorting can be supplemented with other IA methods to identify issues in the category structure of a website/app.
Conducting a Simple Card Sorting Session
- Choose a set of topics, write them down on cards (these can be either virtual or paper index cards), and mix them
- Let the user arrange each topic into groups (allow ‘not sure’ items)
- Ask the users to label those groups
- Ask the user to explain the reason behind their decisions (or ask them to think out loud)
- Repeat with 15 to 30 users (the more the better)
- Analyze the data and make design decisions accordingly
- Don’t forget to prepare your script in advance, and conduct a practice session
Variations in Card Sorting
This is the most common type of card sorting exercise. Users are free to assign whatever names they want to the groups they’ve created with the cards in the stack.
Users are given a predetermined set of category names, and they are asked to organize the individual cards into these predetermined categories. It is used to evaluate how well an existing category structure supports the content, from a user’s perspective.
One caveat: this method does not reflect how users naturally browse content. We recommended doing Tree Testing instead (giving users a task to perform with the current navigation).
Ask users the reasoning behind their choices. This helps to gain qualitative insights into users’ rationale for their groupings.
Task users with organizing content into groups on their own, usually via an online tool, with no interaction with a facilitator. It is generally faster and less expensive. Unmoderated card sorting can be useful as a supplement to moderated card sorting sessions.
The traditional form of card sorting. Topics are written on index cards and users are asked to create their group on a large workspace. This method is very easy. It’s a flexible process where all cards can be moved around and/or start over. The downside of paper card sorting is that it requires manually documenting each participant in the analysis.
A software or a web-based tool simulates topic cards, which users then drag and drop into groups. It’s the easiest for researchers because the software can analyze the results from all the participants and reveal patterns. The downside is tool usability might cause frustration or even prevent users from creating the exact groups that they want, and therefore can impact the results.
- For the best results, use different types of card sorting, or combine them and take advantage of the best features of each (quantitative and qualitative results)
- Be flexible, include “I don’t know” cards and the possibility to add new categories
- Limit the number of cards (don’t test everything at once)
- Prepare some follow up questions to get more user insights about the reasons behind user decisions