I’ve mentioned usability card sorting in the past, briefly as one of the more popular prototyping and visualizing approaches for designing UX. But, I’ve never really talked that deeply about it, and given its popularity, it’s time I remedied that.
However, let me start by saying that I’m not that big of a fan of usability card sorting myself, though I see why others are. It’s a clustering methodology, grouping items into contextual topic groups in order to itemize the various facets and variables which must be addressed for usability. I see where it’d be helpful to certain mindsets, both to test an idea dynamically, or in charting the strategy and model for your user-end design.
Well, in circumstances where I worked on teams, the team opted to use it, and so I have participated in it. I didn’t mind. But, having seen it in action, I think I can give you a simple guide to effectively using this. For the right ways of thinking, it’s a very handy approach.
So, the basic concept, like I said, is to create a series of items on cards, and to pick out a series of topics. Then, you go through the cards, and group them together by shared topical relation. This creates a series of clustered units, which list all the variables to approach.
It is a creative solution, but the problem is that it can get convoluted when you try to be too refined with it. Sometimes, topics fall under more than one category strongly, which causes one of two things to happen. It either means that only one of several important categories apply, or it causes added complexity with multiple cards of the same name, etc.
So, how do you overcome that efficiently? Well, just plan for this multiple category thing ahead of time. Once you’ve picked the categories, when you make your cards, multiple copies of each item – one for each topic. The be ready to have multiples in groups. The ones that don’t get used, just be rid of.
Now, the trick is, how do you pick your categories? This is usually a group activity, so there are two ways to pick them. In open, everyone can contribute categories, and make new ones as it goes. This gets hairy, when you account for those duplicate overlaps. The other, closed, has you picking them in advance, and enforcing them.
A good solution to this might be to have a democratic process for selecting the categories, where everyone contributes and rules on a final set, and then you get to making items to group.
Now, how do you read this data? Well, it doesn’t take a metric. This isn’t really a direct testing tool as much as a modeling tool for designing test parameters, as well as a clustered map of contextual things to address in sets when doing all phases of design.
It works. The only reason I’m not a fan of it is because in a traditional use, it requires location dependency. Everyone has to be present in a physical space, in order to participate.
If it can go digital – maybe an SaaS solution to do group online card sorting will be or even has been designed – I’ll be a much bigger fan of it. Still, usability card sorting isn’t difficult, if you just go into it with the things above in mind. Being ready for conundrums it brings out. So, if you don’t mind the location dependency, and you have the frame of mind that does well with charts and flow plans, then you’ll find it handy.