Card sorting is a technique used to understand how users naturally categorize information into logical groups. Participants are usually given labeled cards and asked to sort them into groups that make sense to them.

This data can come in handy when crafting the information architecture of your product. For example, if you are a furniture-selling website, carrying out a card sorting study among your target audience can tell you what kinds of furniture they would expect to be grouped together under a single page, i.e., according to the room or utility, etc. A card sorting study should ideally have between 30-60 cards and 15-20 participants.^[1]

Card sorting studies can be:

• Closed: Participants sort information into the categories you provide them. Use this method if you want to understand your users’ mental models to build a new product or improve an existing one.
• Open: Participants name their own categories and sort information into them. Use this method when you need to evaluate your existing information architecture, fix glaring problems, and clear up any confusing information.
• Hybrid: Participants sort information into both categories provided by you and those created by themselves. Use this method when you have a general understanding of your user’s mental models but need more clarity in specific aspects.

You can choose to carry out card sorting online or offline, with or without a moderator. In any method, the important thing is to brief your participants adequately about the purpose of your study and what you’re looking for from them.