One of the biggest questions in running card sorts is how many cards and users to include. In this video, learn what influences the right number and how to come to the right conclusion for your project.
- You can do as many or as few cards in a sort as you like. The sweet spot depends on what you're trying to learn, whether you're doing a moderated or unmoderated test, and how complex the cards or concepts you're testing. The more complex the cards and the more surrounding discussion you hope to have, the fewer cards you'll want to include. Typically, you want to aim to have moderated sessions be about 20 minutes of pure sorting and at least 20 minutes of conversation. The best way to know if you've hit the goal is to do a pilot test with a colleague. However, if you're working on a very complex system, you may need to have more cards or you may need to allot more time for a discussion in order to get useful results.
If you do an unmoderated study, you'll want to aim for about 15 minutes of sorting. If the session is too long, participants may start to get burnt out and become less attentive. The analysis will also be more difficult with more items. If you're doing a closed card sort with straightforward categories, you can probably get away with more cards because it's easier for participants to sort and easier to analyze against existing categories. With too few cards it can be hard for participants to create enough groupings to spot patterns or to feel confident about your insights.
If you find that you don't have enough cards, you may want to consider broadening the scope of your study. Because card sorting is about understanding patterns and thinking, there is no set number of total participants that will lead you to an objective truth. Just like with the number of cards, the appropriate amount of participants depends on the complexity of the test, the type of test, and the goals you have. Essentially, you'll want to gather enough information to see trends. That may take just five moderated conversations for a closed sort, or it may take 100 participants of an unmoderated open sort.
One thing to note is that you can do one-on-one sessions, or you can moderate a few participants at once. This way, the participants can discuss the grouping and you'll be able to pull more qualitative insights. Moderating a few people at once also allows you to get more people's feedback in a shorter amount of time. Either way, I've found that it's better to perform multiple card sorts with fewer participants each rather than fewer, larger studies. You can also always do a test with a small amount of people to begin and add participants if things look unclear once you've started to analyze.
Remember that the goal of card sorting is to help give you some insight about how people categorize and label things, not to give you definitive answers. With that in mind, I always aim to include at least five participants in moderated sessions and at least 15 participants in unmoderated sessions.
- What's card sorting?
- Open, closed, and hybrid card sorts
- Card planning
- Category planning
- Finding, selecting, and screening participants
- Scheduling and incentivizing participants
- Running sessions