On Optimal Workshop this week, we evaluated 3 types of unmoderated user testing to establish which test would be the best to use to test our mobile app and why. Here are some of our insights.

Initial Impressions

Open Card Sort

This test would allow us to design a useful navbar structure for intuitive navigation through pages and resources. I really liked the PCA analysis and the adjustable information architecture slider. With it, I picked out an architecture I was happy with that contained the number of groups that I might put on a navigation bar, and did so in only a matter of seconds. While this information was very useful, I feel as though it would be less productive to use after having already designed an app or interface.

 

Closed Card Sort

The closed card sort felt like the natural follow-up to an open card sort, after implementing the natural next design iteration. I found the results were nothing like what I would have predicted and that the results didn’t make as much intuitive sense to me. This particular test feet more like trying to fit a card where Bananacom wanted you to place it and not where it made the most logical sense to place it. For that reason I would only use this sort if the design team had a very specific reason to try and sort material into a limited range for some reason.

 

Tree Test

While I did find this test quite useful, I feel it would not be as ideal as other tests to assist in early stages of design. Once designed, this test would be an excellent tool to test and measure the success of specific tasks. This feels like a great fine-tuning tool that the team could use to validate the success of an established design in late-stage iterations.

 

The Winner

To test an app that has been developed I would use the Tree Test. I believe this would give us the best measure of success with the designed structure and flow of the app. The results from this testing can be evaluated and directly influence the next design iteration. The ease with with a user gets to the page they’re trying to find directly impacts the user experience. A hard-to-find asset buried deep within an application that no one can find heavily impacts usability and engagement. When assets are obvious and easy to find our users can quickly and easily navigate to/through the app making for a pleasant, intuitive, and natural user experience.

For early development stages I might choose an open card sort to start to understand how the design likely “should” be structured. From there, I can create a structure that leverages that information, but also applies basic good design principals while aligning with certain brand and client considerations. Once a rough structure is settled on, a closed card sort would be a good way to validate the chosen grouping while capturing additional insight from users on what the final structure should look like. After one more rounds of design iteration, taking this data into consideration, the almost-final version should be tested using the tree test for a one last pass-or-fail evaluation of the nearly-finalized structure. Any failings here should be re-evaluated and final design changes implemented, re-testing with the same style of tree test once that last iteration is complete. Circle back as needed until successful.