What is it?
Card sorting is a quick, cheap and versatile tool for the early stages of a project, specifically valuable in auditing and structuring content.
Users physically sort cards labelled to represent areas of content into groups and hierarchies that feel appropriate to them. This allows us as designers to understand their ‘mental model’, the way they group, sort and label tasks in their own mind.
Why is it good?
When facilitated ably it can provide a solid basis for designing the information architecture of new and existing projects with a small outlay of time and cost. The process is also flexible, with the ability to conduct the task without any pre-defined groups for users to work with, or with some structure already in place.
If facilitated carefully card sorting can not only give clear direction on content, but also begin to shed light on how users think about tasks. To gather detailed information about tasks, a full task analysis within context should also be considered.
How do we use it at PDR?
We use card sorting for projects where the organisation of content is important. This includes screen based interfaces and websites.
The greatest benefits come from leaving the exercise completely under the user’s control, without pre-defined group titles. This format allows for a true understanding of how users naturally see content fall into groups and hierarchies.
An open session also welcomes the user into the decision making process, as they assign group titles themselves and have the opportunity to add labels for content as they see fit. Making it clear that they are free to voice their concern or confusion over any content they are not comfortable with, creates an environment where the user does not feel tested or under scrutiny. When planning any research, making sure users are put at ease is a high priority
Limitations of Card Sorting
As an exercise it is content focussed, not covering navigation and task based issues. When facilitated well with a user who is very comfortable with the content it is possible to gather extra details, however this can’t be planned into the card sorting stage. At PDR we are careful to make sure we have research methods planned at appropriate phases of the project. This is far more productive that attempting to get a whole range of data from one limited method.
When analysing the results of card sorting it is important to bear in mind the results of any task analysis undertaken as the two sets of results should be weighed against each other. As is the case with many methods, there is skill in knowing how to use the collected results to achieve the greatest benefit for the user.
For example, Card sorting will not provide direction on user interface design details, although understanding the information architecture will have big implications for the user interface.
Is it appropriate for your project?
If the design outcome will include an interface, card sorting would be an appropriate method alongside others that PDR regularly use. If you already have a defined product with a need for an interface a card sort may be one of the first exercises we look to undertake with users. If the project is at earlier stages and more research is required into user needs, a card sort may become useful later in the project.
As with all methods, we assess the suitability of them for the project and can adapt them as needs be. We work closely with each client to ensure they are aware of the steps we are taking to move the project forward in a way that is focussed on the user’s needs.