Tutorial 6: How To Conduct A Card Sort

“Hello again! 💁I’m back to walk you through the process of card sorting, which is one of the most common user research methods. If you’re enjoying these tutorials, make sure to check out our full UX Design Program and follow us on Instagram and YouTube for regular tips and tutorial videos!”

What are we going to do today?

Welcome to tutorial six of your UX research short course! In this lesson, we’ll focus on another very useful research technique: card sorting. First, we’ll consider what exactly card sorting is and how it works. Then we’ll introduce some different types of card sorting before showing you, step by step, how to conduct your own card sort. We’ll round off with some tips and best practices and, as always, a practical exercise. By the end of this tutorial, you’ll be ready to conduct your own card sorting sessions as part of your user research.

We’ve divided this lesson into the following sections:

1. What is card sorting?
2. Different types of card sorting
3. A guide to card sorting
4. Card sorting tips and best practices
5. Practical exercise

Ready to become a card sorting pro? Let’s begin.

1. What is card sorting?

Card sorting is a UX research technique used to either evaluate the existing information architecture of a product, or to generate ideas for how the information architecture of a new product should be organized. But what exactly is information architecture? Let’s recap briefly.

Information architecture is concerned with how content is organized, structured, and labelled. Think of how you navigate around a website using menus and submenus, or the content you see on the homepage—that’s all mapped out as part of the information architecture. A carefully-planned information architecture makes it easy for the user to navigate your product and find what they’re looking for. That’s absolutely crucial when it comes to creating a good user experience! You can find a comprehensive guide to information architecture here.

So where does card sorting come into it? Well, as a designer, you want to determine the most logical, user-friendly way of organizing your content. While you may have your own ideas about how the information architecture should be structured, what’s logical for you isn’t necessarily logical for your users. That’s why card sorting is so useful: it shows you how your users categorize and classify information, allowing you to remain user-focused when designing your information architecture.

How does card sorting work?

As the name suggests, card sorting is done using cards. First, you’ll pick a set of topics based on the content you want to include on your website or app. You’ll write a different topic on each card, shuffle the cards into a random order, and then hand them over to the user to sort into piles. The aim of card sorting is to get the users to group information in a way that makes sense to them. This helps you understand your users’ mental models and thus determine the best way to organize your site content.

What is a mental model?

A mental model is based on what the user believes about a given system; in other words, how they expect a certain system to work. When it comes to websites and apps, the user’s mental model will influence how they navigate and interact with the interface. So, as a designer, it’s important to design products that match the users’ expectations. Card sorting is an excellent way to learn about how your users expect information to be categorized and structured.

2. Different types of card sorting

Now we know what cord sorting is and why it’s so useful, let’s take a look at some of the different types of card sorting.

Remote vs. in-person

Card sorts can be conducted either remotely or in person. In an in-person card sort, you’ll write your topics on index cards and set the user up at a large desk or table. In a remote card sort, the user will drag and drop digital cards. You can conduct remote card sorts using specific tools, such as OptimalSort. We’ve included a link to further card sorting tools at the end of this tutorial.

There are pros and cons of each. The good thing about in-person card sorts is that the user has plenty of space to work with and can easily rearrange the cards if they change their mind. The downside is that you’ll have to document the results manually for analysis later on. With a digital card sort, the software will collect and analyze the results for you. Online card sorts are also generally easier (and less time-consuming) to organize.

Open vs. closed

Another distinction to make is between open and closed card sorts. In an open card sort, the users are free to create their own labels or category names for the groups they’ve created. In a closed card sort, the users are provided with a set of predetermined categories. Open card sorts are the more common method, and will offer you much greater insight into the user’s mental model. A closed card sort, on the other hand, doesn’t tell you anything about how users conceptualize a set of topics; in fact, it can often feel like you’re merely testing the user on their ability to place the content in the “right” category. When conducting a card sort, you want to see how users group the information of their own accord, and an open card sort gives them the freedom to do so. However, a closed card sort can come in handy if you want to test how logical your existing information architecture is.

Moderated vs. unmoderated

Card sorts can also be moderated or unmoderated. Moderated card sorting usually takes place in person. You’ll ask your users to think out loud while sorting the cards, or to explain the rationale behind their choices at the end of the session. Some useful questions might be “Were there any cards that were especially difficult to categorize?” or “Were there any topics that you felt could have been placed into more than one group?” If the user has left any cards unsorted, it’s also a good idea to ask them why. Moderated card sorting is a great way to gain further qualitative insights into your users’ behaviour.

Unmoderated card sorting, on the other hand, doesn’t require any interaction between the user and the person conducting the research. Remote card sorts are usually unmoderated. If you have the time and the budget to spend, it’s best to complement unmoderated sessions with moderated, in-person card sorts.

3. A guide to card sorting

Whether you choose to conduct a remote or in-person card sort, you’ll generally follow the same steps. So what are they? Let’s take a look.

1. Create a list of topics

First up, create a list of topics based on the content of your website or app. Let’s imagine you’re designing a dating app. On each card, you’ll write potential topics (or functions) you want to feature somewhere on the app. For example:

• Privacy settings
• Add photos
• Edit profile
• New matches
• Messages
• Report user

Try to come up with between thirty and forty different topics. And remember: stick to one topic per card.

2. Ask your participant to sort each topic…

Next, shuffle the cards and give them to the user to organize into piles. If the user isn’t sure how to categorize a certain card, they can create an “unknown” pile. It’s also fine for them to change their mind and rearrange the cards throughout. There’s no right or wrong number of piles, and they can all be different sizes!

3. …And create category labels

Once the user has grouped all the topics, ask them to create a name for each group. Remember to bring some spare blank cards for this step!

At the end of the session, you might ask the user to explain how and why they came up with the categories they did. If they struggled to place certain cards, or went back and forth between a few groups, this is a good time to get them to elaborate.

4. Review the results

You’ll want to carry out your card sort with at least five users. Once you’ve gathered enough data, you’ll analyze the results. What common groups did your users come up with? Which cards were frequently paired together? As patterns begin to emerge, you’ll start to get an idea of what the information architecture of your app will look like.

4. Card sorting tips and best practices

You’re almost ready to conduct a card sort! Before you do, here are some tips and best practices to keep in mind.

1. Brief the user. At the beginning of the session, brief the user on what’s about to happen. In fact, this goes for any kind of user research you conduct! If you’re conducting a digital card sort, you can brief the user with a written opening statement. Give them instructions, tell them how long the session will take, and let them know why you’re conducting a card sort in the first place. Your opening statement could go something like this: 

“Thanks for taking part in the session today. Please take a look at the cards one by one and sort them into groups that make sense to you. There are no right or wrong answers, and you can move cards to a different category if you change your mind. The session should take no longer than 15 minutes. Your response will help us organize the information on a dating app we’re designing.”

1. Be mindful of participant fatigue. If you demand too much of your users, they’ll grow tired and the quality of the results will deteriorate. So, use a maximum of forty cards.

2. Randomize! For each card sorting session you conduct, shuffle the cards into a random order. This way, each card—and the information on it—will be sorted at a different point each time. 

Card sorting is a pretty straightforward yet extremely valuable user research technique. The above guide is for an open card sort, but can easily be adapted for a closed card sort; just make sure you prepare your content categories beforehand.

5. Practical exercise

So there you have it: card sorting in a nutshell! Ready to try your hand at a practical exercise?

For today’s task, you’re going to plan and conduct a card sort in order to evaluate the information architecture of the easyjet website. This will get you thinking about the kinds of topics you might put on each card when conducting your own card sort.

First up, open the easyjet website by clicking here. Next, browse the website and think about what topics you might include on your cards. Remember: you’re evaluating an existing information architecture, so you’ll be running a closed card sort. That means you’re working with predetermined categories. Can you identify six overarching or top-level categories and between four and six topics that fall into each category? Write these down on individual cards (you can also use Post-its). Finally, with your main categories in place, ask a friend or family member to sort the sub-topics into each category. Does their categorization match that of the easyjet website? What does your card sort tell you about easyjet’s existing information architecture?

Congratulations—you’ve just conducted your first card sort! You’ve now got some pretty solid theory and some handy research skills under your belt. Great work!

It’s a wrap!

That marks the end of tutorial six, the penultimate lesson of your UX research short course. In the next (and final) tutorial, we’ll show you how to analyze your research findings. Keen to learn more about card sorting? Check out the resources below, and don’t forget to take the quiz.

• 7 of the best card sorting tools
• 8 things I wish I’d known about open card sorting
• A beginner’s guide to card sorting