People first. Information first.

Re-Architecture of UPMC Health Plan’s Member App


The What

A yearlong effort to revamp and re-architect our flagship member mobile app. With collaboration and research at every turn, we introduced a new IA, remodeled the home screen, surfaced important content, and laid foundations for ongoing development.


In the five months after launch compared to the five months before:

  • 1.8 times the number of sessions per user
  • 15-point jump in the percentage of people using the app more than once a month 
  • 3.5 times more live chats with Member Services
  • Twice the number of doctor searches
  • 80–100 monthly submissions of in-app feedback received from users, with previous average of 10–15

My Collaborators

Alexis Miller [Product manager], Himen Patel [Business analyst], Srikanth Chinthakindi [Engineer], Larissa Jantonio [Designer], Marya Markovich [Scrum master], Roman Velykodanov [Engineer], Vladimir Shelogurov [QA], Mour Diop [Engineer]

My Contributions

Content strategy, Documentation, Facilitation, Information architecture, Interaction design, Prototyping, Stakeholder critiques, UX strategy, Usability testing, User research, Visual design


The Why

“I wanna be able to chat with someone.” This was one of my favorite requests we had received from users of our flagship member app. Although this person believed they were asking for a new feature, live chat with Member Services was in the app. And it underscored a point: You can't use what you can't find.

By summer 2016, it was clear the member app wasn't living up to expectations of its users or business stakeholders. The app had been conceived and then evolved without enough thought about its underlying architectures. And now people couldn't find useful features, important information was buried and disjointed, and too many users were logging in once and never returning. There was nowhere to add functionality or content without overcrowding the home screen's "junk drawer," as we called it.

To shift the app's user experiences and prime it for innovative growth, I believed we needed to start with architecture. This would mean revamped structures for its organization, navigation, and home screen.


Very early on, this concept poster was created to communicate ideas about what a re-architecture could mean for the app and its users. It was displayed around our floor and in the office of our senior director, often stirring and giving focus to conversations.


The How

As the app’s lead designer, I made the case in conversations, in illustrative visuals, and in IA research. A series of card sorts and tree tests revealed people's mental models of our content.

Insight: We had been dividing our content between categories of “coverage and benefits” and “claims and spending,” but people didn’t grasp that distinction. All they saw was a single bucket of insurance jargon.

Playing with ways of organizing before another card sort.

Using a do-go map to see the flows within proposed structures.


When we kicked off the project, we did it collaboratively with design, product management, engineering, and QA. During workgroups I facilitated with colleague Larissa Jantonio, the team forged a shared understanding of what we were doing and why.


Excerpt from the project brief, synthesized with the materials produced and discussed during our kickoff sessions.

Another project brief excerpt. It was important to me that everyone could see how collaborative this was, and that teammates could see contributions in action.


I then used our kickoff’s design principles to guide concept exploration. We knew that with a project like this one, we needed feedback from stakeholders and SMEs yet also needed to avoid being derailed by any single stakeholder’s personal tastes. The answer was to format our concept reviews as critique sessions. This allowed stakeholders, without being design experts, to focus on what they knew best and help us improve our ideas.

As we converged and began developing, I orchestrated and contributed to the planning and execution of user and usability research that would test key components of our re-architected app. What we found was a few places where language needed revision, where affordances needed to be made more obvious. But all in all, we were on the right track. Our discovery, alignment, and research at the very start had paid off.

In one of our validation studies, participants were asked what they liked about our changes to the app. To which one person answered: “Everything, and don't let who came up with this version be fired!”

Before, on the left. And after, on the right.


Our re-architected app launched December 2017, and from what we can see so far, the changes are having exactly the impact we intended. Nearly 80% of sessions are now people using the app more than once per month, up from 63%. There’s been a 16-point jump in the percentage of multiple sessions in a single day. And those members who would visit once and never return? We’ve reduced the percentage of single-visit users from 14% to 8%.

Of all the highlights to this project, these are the most satisfying to me because they signal (along with additional early returns that merit further analysis) that we truly made the app more useful and meaningful to our members’ lives. Though for our user who just wanted to chat with someone, it is equally nice to see the total number of live chats are up by 350%.