How might we create a mobile app (digital therapeutic) to help people with schizophrenia cope with their symptoms and improve their lives? This sounded like an incredible opportunity so I jumped at the chance to take on this challenge and help an underserved population.
Case Study Overview
1) My willingness to push and be vocal for the things a product team needs
2) I’ll do whatever is necessary to ensure a high quality product is delivered
For this project, my primary role was the design lead where I drove the team towards solving our users’ unique needs with a UX-focused process.
- Brought user research and usability studies to Pear Therapeutics
- Took on the additional role of product owner for a year in addition to being the design lead
- Increased engagement by 50% over the course of 6 months
Blazing a Path for User Research
I’ll start where many projects should - with the user. At Pear Therapeutics, nobody had conducted user research prior to this project. My team pushed the boundaries at Pear of what it meant to create a product from scratch and in doing so, we gained access to interview people with schizophrenia at various mental health clinics. By showing the value of user research, we opened doors for not just our team, but for all future products.
Once we received the green light to conduct user research, I collaborated with the clinical director and product manager to come up with the user research protocol. I then created the note taking format in Airtable (see below), and how to support it technically with audio or video recordings.
The user research was exactly what we needed. From our notes, we were able to come up with the user needs. Having these allowed us to better define the product.
- Define the purpose of the research
- Describe a typical session
- Provide the rationale for the research
- Justify the cost
Defining the Product and Business Goals
Pear Therapeutics had a partnership with Novartis to bring this product to market. However, the only business goal we were provided was to deliver an app for a clinical study. The trio of product management, clinical, and myself in UX wanted to aim higher. So we came up with goals (see goals to the right) that fell in line with our user needs and were measurable to track over time.
I contributed to the thinking around the first two goals of increasing the efficacy of the treatment. My biggest contribution was pushing to raise the floor of engagement. I felt it was more important to raise the engagement of users who had little to no engagement instead of those who were already engaged. If we accomplished that, a broader set of users would derive increased benefit from the therapeutic.
- Personalize the therapeutic experience through increased data collection
- Strengthen and support the moment of skill practice
- 75% of patients will use Thrive at least half of the days in the prescription
Kicking Off Design
As the sole designer on this project, my responsibilities included running the user research, kicking off brainstorming sessions, creating lo and hi-fidelity designs, building prototypes, leading the usability testing, and then collaborating with all the various members of the team to ensure the designs became reality.
I am a big believer in the power of iteration. The input from team members, other designers, users in usability studies, and eventually data from real users, is critical to crafting the best product.
While designing, does this user flow get our users closer to meeting their needs? Does this visual design improve the experience for our users?
Design Example #1: Iterating on Skill Practice Flows
The primary goal of this app was to teach people new skills to manage their symptoms. There are 76 skills broken across 10 different categories. We needed each of these skills to hold a user’s hand step by step as they learned and practiced the skill. The step content was created by our clinical director.
The primary question I was trying to solve:
How might we design a flow where users can either learn the skill or help them practice it at a later time?
After multiple whiteboard brainstorm sessions, I took the designs into Figma. Shown below are a few of the many evolutions of the design and the process by which I got to the final design.
- My ability to design complex flows from scratch for the needs of a specific population (many with cognitive impairment).
- I work in iterative cycles using design reviews and usability studies as a way to get feedback to continually refine the designs
- I created a design system using components that powered 76 different skill practice flows.
Early in the design phase, we tested a version that was similar to filling in an ad-lib. The thinking was that seeing all of the answers together would help people understand the context of why they were answering questions.
For the usability test, I came up with the protocol and prototyped it in Axure to make it feel as real as possible. What we found is the complexity of filling in multiple questions of a broken sentence led to less comprehension. The perceived efficiency of fewer screens actually made the process more inefficient.
In this version we broke out each step into its own screen to reduce complexity. Again, we usability tested it with a protocol and prototype I created.
We found that breaking up questions like this allowed users to focus on one thing at a time, greatly improving comprehension and success.
After multiple additional versions and testing, the final version continued the step-by-step approach with usability tweaks along the way.
- We decided to hide the next button until an answer was selected instead of simply disabling it because even when disabled some users thought they could tap it.
- We started to “pull forward” answers from previous steps into the current step to better personalize each step.
- We added a summary screen to remind users of the steps they just completed.
After additional usability studies, we found users were able to comprehend and complete the steps without issues. Success!
Design Example #2: Behavioral Activation and How to Show Progress
Oftentimes, people with schizophrenia have some level of depression. One of our goals was to reduce the severity of depression using a method called Behavioral Activation. It essentially boils down to achieving goals via small, attainable steps. This is one of the most important aspects of the therapeutic since it was a direct response to one of our key user needs.
What I am showing below is the way in which users would track their progress. There’s a larger flow to selecting a goal, learning about it, and committing to the goal. If you’re interested, I can show more of that in detail. The reason I’m highlighting just the tracking progress piece is to give you a glimpse of my process for iterating on a design.
- I will also use an iterative process to design for a single, frequently-used component until it’s right
- I’m not afraid to make mistakes and am ready to pivot quickly when information points me in a different direction.
In the early version of tracking progress, a user had to complete the same activity 4 times in a week. I designed the card on the home screen to give an overview of progress and the detail view to both provide a way to mark the days the task was completed as well as see which other days were checked off.
My designs evolved after design reviews and usability studies where we found the details screen wasn’t clear and that doing the same task 4 times a week was too repetitive.
In order to make the goals feel less repetitive, I worked with the clinical director to break down each goal into 4 distinct and achievable steps. I then updated the home page card to account for these changes. I updated the visual style of the counter, added instructive text, and changed the button label to push people to see what the steps were. The detail view got a complete overhaul. No longer was it important to mark which days a task was completed. What we prioritized was showing the steps and making it easy to check off steps.
In the usability studies, comprehension of the circular progress tracker was low and there was also fear that the checkboxes made it too easy to “cheat”.
In the final version, I took the idea of simplification one step further to address the issues from previous studies. The visual design of step tracking was simplified so that users could see how many steps there were and how far along they were in completing those steps. There’s now just a single button “I did it!” that lets the user tell the app the step was completed. Users can only complete one step at a time. We wanted users to extend their learning rather than trying to do everything in one day.
We found that in the usability studies, this simplification led to much greater task completion and comprehension.
Keeping the team running smoothly
Our team happened to lose team members to other projects and they weren’t replaced. Not great. But it gave us all opportunities to step up and so I stepped into multiple different roles. In addition to my design lead responsibilities, I was the product owner for a year, a developer for 3 months, and wrote regulatory documents. I’m proudest of my ability to become the product owner as it was the biggest need and had a significant impact.
Stepping Into the Product Owner Role
When the team lost our product manager to another project, he was never replaced. That gap in responsibilities gave me the opportunity to lead the tactical aspects of the team. Over the past year, in addition to being the sole designer, I:
- Managed and prioritized the backlog in JIRA
- Wrote tickets for all the new features and tasks
- Completed the backlog grooming
- Organized the sprint planning
Overall, my time spent as the product owner, was a great experience that kept the team running smoothly and allowed me to stretch my skills.
Results and Making Meaningful Products
Over the course of two years, the product continually improved as we learned from clinical trials and user research. We hit each milestone adding new impactful features based on user needs. Through my designs and an iterative process, we ensured that new features always had high comprehension in usability studies so that we felt confident when shipping them.
It’s not easy to get real data when we have to run expensive clinical trials but the data we received was compelling.
One of the important measures was engagement. We found that we were able to significantly boost engagement from one study to the next. In the 12 week study, 75% of patients opened the app each day (a 50% improvement) and 55% practiced a skill (a 40% improvement). These engagement results were incredible!
One of the most meaningful aspects of this project was being able to connect with the people we were building the product for. We all have our favorite stories of a specific person that made an impact on us or a person that really opened our eyes to what life is like with schizophrenia. Ask me about them! For now, here are a couple quotes from people we talked to that showed me the impact of the work we completed.