Psychologists initially devised the practice of card sorting to investigate how people arrange and classify the information they have. However, in the information technology space, software application designers, developers, and information architects have to constantly navigate around issues related to arranging features, information and functions in a user-friendly way so that users can locate them conveniently. Utilizing card sorting in design can help them determine the most appropriate and relevant schema of information.
The original card sorting technique involved writing labels on the cards, each representing a series of abstract and concrete concepts, and making the participants sort or classify the cards into stacks with shared similarities. After completing the card sorting exercise, the participants name the piles using words or phrases to highlight the similarity shared by all the cards in that pile.
What is card sorting in design?
Card sorting in design is a quantitative research tool. It is mainly used by UX designers and information architects. It enables them to evaluate or establish a pattern for grouping, classifying, labelling, and structuring information.
As in the original method, here, too, users are instructed to segregate a range of content topics into groups or categories which they feel make appropriate sense.
The findings serve as valuable tools for UX designers and information architects to understand their users’ thought processes and expectations of the product. This knowledge is key to helping them design apps or websites that are relevant to customers and meets their needs.
The insights obtained from the card sorting technique also let you seamlessly integrate helpful content with a host of relevant features in an appropriate and easily accessible way. Thus, card sorting has a vital role if you want to remain successful in a competitive market space.
How is card sorting done for UX?
Card sorting in UX is, in principle, similar to the card sorting experiments done by psychologists. In UX, too, the designers use cards to organise the activity based on a predefined set of criteria. The process is designed based on whether the purpose of employing this technique is evaluative or exploratory.
The designers then write the different topics, one on each card, randomly shuffle the pack and then give them to the user participants to arrange into stacks. For the cards, some use actual physical cards, while some prefer to use one of the many software tools designed for online card sorting.
The goal of card sorting exercises in UX is to develop an awareness of how different individuals dissect and look at the pieces of a problem, task or concept and reorder them based on their priorities and the comparative value of these analysed items.
Thus, card sorting can elucidate how the mental model of the different users shapes their interaction with the apps and websites and how they navigate and engage with the product or service. Card sorting throws light on how your users expect the system to work.
As a designer, it is your responsibility to design products that match users’ expectations. Card sorting is a valuable tool that tells you exactly how users expect information- in what category and structure.
The use and significance of card sorting in design
Card sorting helps designers understand the grouping of items and how to implement the navigation.
Consider the example of Drupal.org. The card sorting test yielded that many labels featured on the Drupal.org website, though well-comprehended by the “Drupal-insiders,” were perceived to be confusing by its new users.
The Drupal case highlights that web creators often have a rather keen understanding of the classification required but fall short when they have to label these pages. In addition, the study highlighted that over-reliance on the organisation’s or industry terminologies, and jargon makes it difficult for new users to come on board.
Using the card sorting method helps designers assess the website’s information architecture. For example, developers gain insights into which content should feature on the homepage and the number and types of sections the website needs.
Finally, developers can successfully identify the most suitable way to present content to users to make it simple for them to locate and fulfil a task. Card sorting is an efficient best tool to help you construct a website that is perceptive and easy to navigate.
Types of card sorting
1. Open card sorting
The users get instruction to arrange the cards into groups, based on what seems appropriate to them. Finally, the users must name each group they have created with a label that best depicts the group. The open card method is generally used for new/existing information architectures when organising products on a site or creating a new information architecture.
Open card sorts are more flexible. Developers can gain insights into multiple facets, including the number of categories users would like on your website and the labels allocated to these categories. This is an excellent opportunity to understand the terminology used by your target audiences.
On the other hand, though the results are slightly more complex to interpret; they offer feedback comprising very interesting patterns in both the categories created and their naming.
2. Closed card sorting
The participants get both the content cards and category cards and asked to place the cards in the respective categories.
This method is more suitable for testing an existing site. It lets you know whether the names given to the categories are comprehensible or if there were unutilised categories.
It’s also ideal for incorporating new items into the existing structure and studying how users sort them into categories. A closed card sort also helps acquire a second perspective after doing an open card sorting.
3. Remote card sorting
During a remote card sorting session, the users work remotely on their computers, sorting cards provided by online software tools. These online software tools, such as UXtweak, Optimal Sort, allow you to set up and allocate cards to as many users as needed for the test. Online software tools also provide several ways to interpret and analyse the data.
However, since the test is held remotely, the absence of contact with users eliminates opportunities to understand why users have sorted the cards in a particular manner.
4. Face to face card sorting
These sorting sessions are conducted in person, in the presence of an observer. The users are handed a set of cards to move around and are asked to share their thoughts and rationale supporting their decision. The observers can clarify any observations they are not sure about to get better insights into the choices made by the user.
For card sorting to be a success, the participants have to represent your existing or potential target user group; or the results will be distorting. Try considering demographic variables such as age, gender, nationality, and level of digital familiarity – natives or tourists.
The following example concerning Eurostar highlights the importance of selecting the right target group. Card sorting can help describe the labels for the main pages of the website among participants- who have similar experiences and who do not.
Eurostar being a widely used service, it was imperative to include both groups in the study. The sorting activity results interestingly showed that the inexperienced online groups chose real-life words and phrases such as “Planning Your Travel”, “ “Before You Go”, “At the Station”, etc., as category labels. Conversely, the experienced online users opted for more conventional online labels like “About Us”, “Manage Your Account”, “Help”, and “News”.
In this case, card sorting helped sort out online labels and provided insights into possible user behavioural patterns. In this case, it was amply clear that inexperienced users would benefit from support during interactions on the site.
Advantages of card sorting in design
– It’s an established technique and relatively easy and affordable to conduct card sorting sessions.
– A speedy and viable option to conduct research and obtain the results, primarily during remote user interaction sessions
– Provides developers with an understanding of the users’ psyche and their expectations related to the organisation. This helps in structuring of material, as well as the intuitive grouping of content.
– Creates a sustainable foundation structure of a product or site and as a method when studying label quality.
Disadvantages of card sorting
– Results may vary, and there may be no conclusive pattern in the data.
– Depending on the complexity of the data received, the analyses of the results may take time.
– It does not consider the users’ tasks and objectives. So, chances are one may end up with an unviable structure, not suitable for the user to complete the given tasks. Alternatively, the users may provide a superficial or faulty analysis without trying to understand the problem you are aiming to resolve.
– If the presentation of the labels are not set in the proper context, they may lose meaning and purpose.
Card sorting resources and hacks to ensure your card sort is glitch-free.
Audit your web content:
Before performing a card sort, you should have a clear understanding of the information present on your website. Therefore, the critical first step involves conducting an audit of your web content, prioritising it, and identifying the various topics based on the different types of content that you plan to incorporate into the product.
Select the appropriate format for card sorting
In unmoderated card sorting, participants arrange content into groups on their own. However, in moderated card sorting, there is a moderator. Likewise, the topics are generally written on sticky notes in paper card sorting. In digital card sorting, participants drag and drop the digital representations of cards into the respective categories using tools. Select an option that best matches your needs.
Create well-designed cards and use images
Make the cards meaningful for your participants. Test them with your audience to ensure that participants comprehend the information on a card and that cards are groupable. Inconsistencies and poorly designed cards are barriers to creating coherent groups.
Similarly, images can be beneficial in representing concepts and items. Including pictures to explain or clarify the text on your cards is an incredible hack.
Limit the number and assign a number for each card
Target between 30 and 60 cards to collect enough valuable data without overwhelming the participants too much.
Numbering each card will make the analysis easier at the end of the session. Catalogue all the cards in a spreadsheet or table to make the concepts and topics more organised and easy to understand.
Have simple topic labels; Use words carefully
Keep your topic labels brief, precise, culturally sensitive, and to the point so that your user participants find them easy to understand and do not feel pressured. Undue stress may lead to impaired sorting.
Experts caution against reusing the exact words in the topics. For example, “Fruit growing (Card A)” and “Growing Fruit(Card B)” can be confusing. Moreover, participants tend to group those cards together.
Conduct a mock-run
Before you conduct your card sorting, it’s good to do a trial run with your friends and colleagues. It gives an excellent opportunity to weed out typos, ambiguity, or other errors occurring in the system. It also offers a chance to incorporate steps for a smooth flow of the event. A mock run can go a long way to ensure that the actual card sorting activity goes as planned. An anomalous card sort may compel you to discard actual user data entirely.
Begin the session with the cards in random order
Beginning the session with random, assorted eliminates the possibility of bias and skewed results.
Provide an estimated timeframe
Inform your user participants in advance about the time allocated for this task. It will help them complete the activity in the given time frame and also facilitate a serious approach to the activity.
After gathering the responses, you must analyse the data, looking for common trends. Cluster analysis software is one way to go about it. Or, if your sample size is small, you could enter the data into a spreadsheet and scan for any patterns in the labelling and categorising.
Card sorting is a great way to test websites with many categories and subcategories, such as e-commerce, news services, banking, travel and educational sites. It is a resourceful and proven technique in information architecture and plays a crucial role in understanding user behaviour through user research and testing. In addition, card sorting can help you design your products intuitively and easily to use. Whether you are using an open or closed card sort, run as many tests as you think are needed to interpret the data and perceive how your target groups make sense of the information they are presented with.