Open card sorting is a generative research technique that allows you to gather insight into how your users understand your information by asking them to group given examples (or cards) into categories that make sense to them.

In this article, we'll walk you through the steps to create an open card sort. Looking for help making a closed card sorting activity? You'll find those steps here: Creating a closed card sort

There are a few ways to create your cards:

Manually type the names for each card in your task. Press enter or return to quickly move to the next card:

Paste a list of card names from another document into the test builder—one card name per line—to quickly create cards in bulk:

You can also bulk-create cards from image files. Turn on the Images toggle, then drag in a group of image files. You can now add names to the new cards.

Cards can support optional descriptions to add a bit more context for better sorting. To enable this option, switch on the descriptions toggle, then add a description for one or all of your cards:

Here's how the description looks to your participants:

And here it is with an image:

By default, participants will be required to move all cards into one of the available categories. Deselect the checkbox shown below if you want participants to have the option to leave some cards uncategorized.

By default, participants will be required to type a name for all the category groups they create. Deselect the checkbox shown below if you would like to give participants the option to leave some card groups without a name.

Unnamed card groups will appear as "Untitled #*" in your reports. You can customize names, merge categories, and ungroup cards as you would with other categories when you standardize and organize your category results—more on this below.

By default, participants will see all of the cards you have created. You can also limit participants to work with a smaller selection from your list of cards (E.g. participants may be asked to categorize only 10 of your total 50 cards).

When this option is selected, our system will track the number of times each card was shown to a participant, and we'll automatically optimize this to keep views as balanced as possible for all cards. As more participants take your test, we'll present the cards with the lowest frequency first to ensure we don't miss any cards.

By default, the order of your cards will be randomized automatically for each participant. This randomization can be disabled by unchecking the option shown below the list of cards. When unchecked, cards will be ordered as they were arranged within the test builder.

When panelists begin your card sort task, they'll first be shown detailed instructions on how to complete a card sort activity. If you've added instructions to your task, they will replace the bolded text "You're about to do a card sorting exercise":

Our testing interface is responsive and supports card sorting tasks on both desktop and mobile devices. Participants will be able to successfully complete a card sorting activity no matter which device they may be using.

Here's how your task will look on a desktop:

...and here's the same test on a mobile:

After you have collected some results in an open card sort, the next step is to standardize the categories that were created by your participants in order to get useful data and reports. Standardization allows you to group very similar user-generated categories together for more accurate results and reporting.

As an example, a few of your participants might have created a category named “Fruits”, and these should all be grouped together. Another participant created the category “fruits” with a lowercase F, while another typed “FRUITS” in all caps. Our system will automatically group categories with the exact same spelling, regardless of capital or lowercase, so all three variations will be sorted into a single category for you. We'll use the case that was submitted first, but you can customize this or ungroup cards by clicking the category:

You will also have the opportunity to decide which user-created categories mean the same thing. From our example, you might also have a participant that typed the category name "frIUts", misspelling the word fruit. This won't be grouped automatically due to the misspelling, but you can manually group cards together if you think they’re all referring to the same idea.

When you merge two groups, you will need to choose which name to use for the merged group.

All of our results (Eg. card view, category view, agreement matrix) will take the name of the category group, rather than the exact category the participant used. For example, if a participant sorted apples into a group they named “Fruits”, that sorting of apples might end up counting under a group of categories that you chose to name “Healthy foods”.

Now that you've organized your categories, you can check out this helpful article for more details on understanding the results you've received on your open card sort.

Yes, we support closed card sort tasks. You can find more information on closed card sorting in this article.

No, hybrid card sort tasks aren't available just yet. Let us know if you'd like to request this feature!

The file must be a JPEG, PNG, or GIF image. The maximum file size for images is 5MB.

We do not resize your uploaded images, so images with higher resolutions can affect your participants' load times. Also, participants will not be able to zoom into images for a closer look, so we recommend using images that are clear at a small scale.

If your image is cropped, it's because cards work best with a specific shape or aspect ratio. This ratio ensures images fit neatly in the frame—we use a standard aspect ratio of 4:3 (width: 4 units, height: 3 units) on cards. If your image isn't this exact shape, parts may not fit within the card.

Tip: You can adjust your image's shape to avoid cropping. Imagine it's like resizing to fit a frame – add some extra space to the sides so the whole image fits the 4:3 ratio.

No. Categories can only contain a name and an optional description.

Yes. The test builder can support a maximum of 100 cards and 30 categories.

Cards are randomized automatically for each participant, but you can turn this feature off when needed.