A Card Sort is a loosely moderated, user-directed exercise in classifying pieces of information (represented by index cards) into categories. The method aids the design of your site or app’s navigational structure by letting users tell you how they would classify its features or content.
There are two principal versions of Card Sorting: Open, which asks users to assemble a product’s content into groups of their own devising, and Closed, in which users organize content cards within existing categories. This article will be describing a flexible version of the Closed Card Sorting method.
SCHEDULE TIME & GATHER MATERIALS
Prep Index Cards: 1 Hour
Time per User: 30-60 Minutes
Prepare Report: 4 Hours
A camera to record the completed card sort
CARRY OUT THIS METHOD
Complete a Content Audit first so you have a complete content inventory.
Determine the goal of your card sort. Depending on the site or product, it may be general or more granular. Based on that, choose the content items from the content inventory are relevant to your card sort.
Prepare the index cards by writing a relevant content item on each card.
Write the name of each of your content categories on a post-it note or larger sheet of paper. Place them on a table in your testing area – either a meeting room or usability lab – with enough room between them for index cards.
When your test user arrives for the session, explain the Card Sort task to them.
Ask the user to describe their thoughts as they sort in order to access their attitudes toward the content.
Give the user as much time as they need to complete the sort to their satisfaction. It should only take a few minutes.
Be flexible. If the user wants to add a content card, eliminate a card, or change a label, allow them to do so. This may inadvertently reveal an interesting insight worth pursuing.
Take a photo of the results so they can be documented and compared with other users’ card sorts.
TIPS AND RESOURCES
Try these tips
If you have the time, videotape the card sort so you have the user’s thoughts recorded. Replay may reveal subtle insights that you overlooked during the Card Sort session.
If the user has struggled over a sorting decision or has indicated frustration, follow up after the test for more details.
Card Sorts should reveal patterns — that is, users should consistently put certain groups of cards together. If such patterns do not emerge, it may indicate that your product’s scope is too broad or the terminology used in your product may be confusing to users.
Software can be used for card sorting with remote users. However, many UX professionals prefer the low-tech index card method because it allows user observation.