Understanding Card Sorting

What is card sorting?

Card sorting is a research technique that looks to inform the information architecture of a digital product. Once the information architecture has been established, the navigation design is based on this information.

Card sorting involves participants organising content into groups that are intuitive to them. It also allows us to understand how people think about and make sense of content. The groups they create and the language they use provide insight into their mental models. Mental models help us to refine the organisation of content on the website and labels within navigation to make them more intuitive for users.

What is the method for card sorting?

During a card sort, participants are presented with a cross-section of content from a website or product. Each card represents a piece of content and contains a descriptive label. 

• In a closed card sort participants organise this content under group headings that have been predetermined. 

• In an open card sort participants are able to give each group a heading that makes sense to them.

• In a hybrid card sort there is a mixture of open and closed groups. 

Ideally, card sorts are completed by a group of participants who are representative of a selected product audience.

Once the card sort has been completed, participants are often asked a series of open-ended questions to gather qualitative data. 

These questions might ask if there were certain cards that they found challenging to group, or if certain cards were simpler to group. 

Originally card sorting was performed in person with physical cards, as seen in the video below.

In more recent times, however, card sorting is commonly carried out using an online tool such as Optimal Workshop. This allows participants to complete the activity in their own time and space. Online card sorting also provides efficient data collection, potentially saving time on the project.

Card sorting delivers:

• An understanding of which pieces of content belong together.

• The type of language that people naturally use when referring to these groups.

For content-heavy products or websites, more than one round of card sorting may be needed to provide sufficient information and refine results.

What is the outcome of card sorting? 

The final output of a card sort is usually a recommended navigation structure (including navigation labels) or site map. Key insights that have informed the recommendation are captured in an accompanying report.

At Nomat, we understand that the analysis of the card sort data is one of the most critical parts of the process. This data synthesis is performed by unearthing the factors that participants may have considered while sorting the cards. The qualitative data collected helps to give texture to the research and often provides useful quotes to personalise the information. 

Observing the factors participants consider when sorting the cards is a key takeaway of the activity. This information is key to producing a practical and useful website or product structure.

Why should tree testing be performed after card sorting?

Card sorting is rarely performed in isolation. A final validation of the structure determined by the card sort is usually recommended. This can be done via tree testing with participants, which acts as a checking mechanism of the proposed information architecture. 

What have other organisations learned from card sorting?

At Nomat, we have experience running card sorting for a broad range of corporate and government clients. All of these projects have delivered navigation structures that support end-users to locate content easily.

Card sorting was included in a project for a major Australian retailer, which led to a 22% eCommerce sales increase in the electronics category after the identified research changes were implemented. Another organisation that benefitted from a card sorting activity as part of a broader website redesign was cohealth, a not-for-profit community health service. The resulting changes to the website information architecture helped cohealth better support its clients by making it intuitive and easy for the community to complete their top tasks such as booking appointments or locating information about health services.