Sep 23, 2015

“Personas, like all powerful tools, can be grasped in an instant but can take months or years to master.”
Alan Cooper

Background

A couple years ago I had a few meetings with a company in Austin about a possible job doing UI design. They suggested I read a book by Alan Cooper called The Inmates Are Running the Asylum, so I happily obliged. What could it hurt, right? In the end I didn’t get the job, but I chalk it up as a good experience, because that book changed the way I think about how I design things. One of the things that really stuck with me was the idea of personas, a user archetype you can refer to when designing user interactions. I had previously tried to make websites that were easy for people to use, but I really had no idea what that meant. I loved the idea of designing for someone other than myself or my client (friend). Unfortunately I sort of shelved the concept away as something cool I had once read, and never really put it into practice. So it’s been cool learning about personas at The Iron Yard, and how to utilize them to make better design choices.

Project Details

Today we had an assignment to create three personas for the Duolingo web app. We were then to create one user scenario, one use case, one user flow, and identify three possible items to test and optimize. I thought I’d introduce one of my personas and share some details of the assignment.

Elroy

The first persona I created is Elroy. He’s a teenager taking a foreign language class in high school. He likes to play games in his free time. He’s not doing well in his class and needs to bring up his grade, but can’t afford a tutor. Additionally, he doesn’t have the attention span for another book.

Simply put, an archetype is a very typical example of a certain person or thing. And a persona is an archetype of a user. So far the hard part of creating a persona, for me, is attempting to strike a balance between too vague and too specific. It’s challenging to make up someone who is typical, but not a stereotype. So I tried to think of Elroy as a real kid. To me, he seems typical, but not too real. Maybe I should have added a personal detail, like that he would rather be playing soccer than reading a book; something to make him less generic.

Elroy’s User Scenario

Elroy told his teacher that he’s using Duolingo to help him learn his language better, and they thought it was a great idea. Now his teacher wants to track his learning progress. Elroy needs to show his teacher what he has learned and where he needs more practice.

Elroy’s Use Case

1. Elroy navigates to duolingo.com.
2. He clicks on the “Settings” link under his username in top navigation.
3. He clicks on “Progress Sharing” from the links in the right sidebar.
4. He inputs information for his teacher and class and clicks “Save Changes.”

Elroy’s 3 Possible Items to Test and Optimize

• Wording, size, color, and placement of the call to action “Set Your Daily Goal”
• Typeface, size, color, and placement of the wording of questions
• Color and styling of focus states for selecting answer

Testing

To test each of these things, I would start with some sort of hypothesis on what I expected the results to be, based on existing research and studies. In the case of the focus state for answer selection, I would hypothesize that a blue focus state would work better than a red indicator, or even a blue one, because red and green carry meaning for a lot of people of being wrong and right, so that could be confusing for a user.

I would then set a goal for testing. In the case of the “Set Your Daily Goal” button, I would be testing multiple variations of the button to see which wording, style, and placement was more effective in getting users to set a daily goal to return to the site. I would use multivariate and A/B testing to find out which option was most effective in providing a better user experience. From that information I would be able to make an informed design decision.

In Summary

Alan Cooper himself notes that while the concept of personas is easy to grasp, it can take a lifetime to master. Personally, I love the idea of designing for someone more specific than just “the user.” After all, it’s a real person that will be using the thing we make, not some nameless, faceless test dummy. One of the things Cooper talks about in Inmates is how too often programmers write software based on how they think. The problem with that is, programmers think very differently than your average person. That means they’re producing a product that is likely to be difficult to use, for the vast majority of people. I think the same idea can be applied to web (app) design. It’s easy for me to design interactions within the realm of my thinking or capabilities. But that’s probably not going to benefit Elroy, is it?