Think Aloud Testing: Results and Revisions

We’ve been working on redesigning mobile banking apps in Rapid Ideation and Creative Problem Solving. Here’s the process so far:

  1. Create a concept map representing the current version of the app.
  2. Create a concept map representing a proposed revision.
  3. Write scenarios based on the top goals that users have when using the app.
  4. Sketch storyboards based on the scenarios, including preliminary product shots.
  5. Create a first draft of hero flows based on everything done to this point.

That’s where we were after the first couple of weeks. Here are the links to previous posts for more detail:

This last week, we took the hero flows and put them in front of users to see where they were falling short in terms of usability. To do this, we used a simple method called “think aloud testing”. You just have people narrate their actions while they interact with something. It works because people have the ability to describe what they’re doing verbally without distracting themselves from what they’re doing with their hands. In this case, they’re interacting with a prototype I created using a web app called InVision. Here’s how the testing was set up:

Screen Shot 2015-11-17 at 3.28.18 PM


I had a total of five participants attempt to accomplish five tasks using the prototype. It was immediately apparent that I had a big flaw in the app. Not one participant immediately found the menu I was using to house links to primary activities (making payments, transferring funds, and depositing checks). In fact, one person was totally stumped and never found it.

The following images show comparisons of the original screens in my prototype that contained issues I observed during the test along with my proposed revisions.

Here’s a full pdf of the flows, incorporating the revisions below: USBank_Wireframes_Kade_r3.pdf

USBank_ProposedFixes_r1-01 USBank_ProposedFixes_r1-02 USBank_ProposedFixes_r1-03 USBank_ProposedFixes_r1-04 USBank_ProposedFixes_r1-05

Next Steps

Now that I’m relatively confident that the primary flows are pretty solid, I’ll be moving on to making the existing flows more robust and adding additional, secondary functionality. For example, I haven’t yet determined what it will look like to set up a recurring payment or an account notification.