How to Do a Card Sorting Exercise and Why Bother

Card sorting is an invaluable tool in deciding site structure and visual hierarchy. In this post, we explore how to do a card sorting exercise and why it is so worthwhile.

There’s something very satisfying about the term Card Sorting. It appears to my inner neat-freak and hints at an outcome that will be both practical and useful, which is exactly what you need if you’re about to start organising (or reorganizing) the content on your website.

What Is Card Sorting and When Is It a Good Idea?

Card Sorting will help you create a robust information architecture for your website. It’s a user research technique that allows you to discover how users expect to find your content presented. At worst, a card sorting exercise will confirm you’re on the right track, and at best, it will take you off the wrong one and send you in a better direction.

Like many user research tactics, the aim of the game is to get inside the head of your user, make sure you speak their language and help them find what they are looking for quickly and easily.

Card Sorting is well worth considering if you need to:

  • Decide on a structure for your website.
  • Find the right words to form your navigation.
  • Group your content or products in a way that will make sense to your user base.

At the end of your card sorting sessions, you’ll have your content organised and labelled in line with the mental model of your users, as opposed to that of your organisation.

What Types of Card Sorting Exist?

Card Sorting comes in three flavours:

  • Open Card Sort
  • Closed Card Sort
  • Bit of both

In all cases, you can choose to moderate the sessions or not and can carry out the sessions using paper or online tools.

Example of open card sorting.
An open card sort allows the participant to group the cards in any way they see fit.

An open card sort allows the participant to group the cards in any way they see fit. They then give each group a name of their choosing. Labelling your website content with the terminology of your users will ensure your content is delivered in an intuitive manner for your audience.

Example of closed card sorting.
A closed card sort can be useful where labels are already set and you want to know how your users would place content within an already existing sitemap.

A closed card sort can be useful where labels are already set and you want to know how your users would place content within an already existing sitemap. Closed card sorting is a great option if you need to add lots of new content to a site or make sense of a site that has been added to over time and needs reorganising.

The last option is a combination of the two. Your participants will sort cards into the groups that you supply, but can also add new groups if they feel it’s necessary. In all cases, you’ll want to find at least 15 participants.

How to Conduct an In-Person Card Sorting Session.

In-person card sorting sessions can be conducted with individuals or in groups.

To prepare for the session, collate a list of 30-60 topics and write them on post-it notes or cards. With large information-based websites, you’ll probably have more than 60 topics, but as we don’t want to overwhelm our participants, just stick to the main phrases. With existing websites, you can always use Google Analytics to find out which pages are the most popular and use these as a starting point.

As the facilitator, you’ll start the session by providing instructions to your participant/s. Explain there are no right or wrong answers and tell them how long you expect the exercise will take. It’s ideal if you can have a second facilitator with you, so they can take notes and help you observe the process.

The advantage of in-person card sorting is you can ask your participant/s to think out loud as they do the activity. Find out why they group things a certain way, see where they get frustrated and let them explain why they do what they do. Be careful not to lead or influence their process. Your role is to keep them talking, so you can gather insights as to how they are rationalising their choices.

If you’re conducting an open card sort, ask your participants to group the content first. Once complete, ask them to give a name/label to each group.

Card sorting with labels assigned to groups.
When running open card sorting, get participants to label groups.

Once complete, you can take a photo of the result, before shuffling the cards ready for the next session.

How to Conduct a Remote Card Sorting Session.

There are a number of tools at our disposal for remote card sorting, so it’s worth considering your options before you dive in. OptimalSort offers monthly plans or one-off fees if you just need to use it for a single project.

OptimalSort is one of many tools available for running remote card sorting

Getting started is easy, just enter your card titles and you’re about ready to go. You can add your organisations’ logo and ask additional questions before and after the exercise if you wish.

The obvious advantages of remote card sorting are that it saves time and allows you to reach a wider audience. The disadvantage is that you can’t question your participant’s motives, or ask questions while they carry out the task. As with many remote testing tools, you can’t control the environment that the task is carried out in, so you can’t be sure that you have your users full attention.

Things to Keep in Mind when Creating Your Cards.

  • Keep your cards uniform – write them the same way, the same size font etc. Don’t use capitals for some and not for others. Avoid any visual indication that could hint towards a certain hierarchy.
  • Avoid the obvious –  You don’t need to include cards such as ‘Privacy Policy’ or ‘Home’.
  • Make sure your cards are ‘sortable’ – A selection that is too random will make the task too difficult for your participants. Make sure the cards you provide can be logically grouped.
  • Don’t be too obscure. – The words or phrases you write on your cards should be explanatory enough to ensure your participant has some idea as to what the content might be.  
  • Quality not quantity –  You don’t need a card for every page on your website, just enough to represent each of the content areas. Make sure that your selection doesn’t bias your participant in any way. Try to include a similar number of cards for each category you are trying to represent.

What to Do with The Sorted Data?

First up, have a good look. Cast your eye over the results from each session and let it all sink in. Look for repeating, common patterns, note similarities and differences between each session. Watch for anomalies and surprises. You’re looking for any insights that you can use to present the data to stakeholders.

Where open card sorting has been carried out, you might see similar categories being used. This is good, but don’t forget to check that the items within the categories are similar too.

Likewise, if you note that some of your cards are being regularly left out or set aside, try to determine whether it’s because the card labels are unclear or the content itself didn’t seem to fit in with other categories.

Card sorting with cards left out of categories.
Pay particular attention to cards that have not be sorted as you need to understand why these proved challenging.

No Reason Not to Use Card Sorting

If you’re looking for ways to organise and label your content or products, card sorting is a great option.

It’s reasonably inexpensive and can be carried out in person or remotely.  It provides really good insights into how your users think and delivers evidence that can help prevent the internal clashes that can arise when information architecture is decided on by internal stakeholders. If your users can find what they need quickly, that’s a good experience!

A good experience for your users almost always results in a great result for your organisation.

Boagworks Boagworld