Making it easier for members to use ESI's service / website to find a nearby pharmacy

Role: Researcher

Note: As a UX researcher at Express Scripts, I performed all parts of the research process: research planning, writing protocols, conducting research, data collection, data synthesis, presenting, and making recommendations to stakeholders. In my work, I apply myriad methodologies including, but not limited to: moderated in-depth interviews, contextual inquiry, diary studies, usability testing, surveys, RITE testing, tree tests, and card sorts.

Context & The Problem

A designer asked me to do research for a feature enabling members to find a nearby pharmacy. In order for the project to stay on track as scheduled on the project roadmap, I had about 3 weeks to complete research.

This feature request initially came from a specific client, for which there were specific requirements. However, we envisioned an opportunity to scale, providing the tool for all clients and users. Essentially, the feature was intended to assist a member when trying to find a nearby in-network pharmacy using criteria that's most important to them and a view that's easiest for them.

My Approach

My design partner had already detailed the requirements when I joined the project. With the requirements in mind, our research objectives were:

1. to understand users' behaviors around searching for a pharmacy

2. to evaluate how two flows perform in terms of usability and comprehension

Requirements per our client's asks

Survey

I wanted to conduct some discovery research to better understand users' behaviors, motivations, and priorities around choosing a pharmacy, so I conducted a survey with 90 health-insured people. I created the survey in Userzoom, and included a mix of multiple choice, likert scale, and open text questions. Below are a few questions I included in my survey, as well as a couple graphs depicting results from some selected questions.

Results from the survey validated much of what we already presumed in terms of users' behaviors, mental models, experiences, and needs when searching for a pharmacy, however they still helped inform my designer as she created designs for testing.

Usability Testing

I then conducted usability testing on two flows we wanted to test. I recruited 10 people who were a mix of ESI members and other health insurance members. I tested both flows on all 10 people, alternating the order with each participant so as not to create bias toward one flow. We wanted to know if one flow performed better than the other, or if there were aspects of each flow that users preferred. As usual, I managed my notes in figma to allow collaboration in note-taking with my partners.

Notes from testing 4 of my participants

Upon completing testing, I organized feedbacks into "testing topics" (aka: affinity mapping), in order to try to find themes throughout the feedbacks.

Once I fully synthesized the findings from usability testing, I placed my presentation in Figjam and presented directly from the Figjam file. Placing the presentation in Figjam allows stakeholders to easily revisit the takeaways and recommendations, and serves as a nice change from powerpoint.

A slide from my presentation detailing:
1. The insight summary
2. A break-down of the findings
3. Recommendations

Significant Actionable Insights

Among my most note-worthy actionable insights were:

1. While we had to meet a select filter requirement, there were clear consistent favorite filters used by testers.

My recommendation: So long as we include the required search criteria, we might consider reordering the filters to better match users' priorities.

2. Some of the more critical thinkers among the testers reported they would want a search radius much like facebook marketplace provides.

My recommendation: while this may not be in scope for MVP1, consider adding this feature to the roadmap for MVP2.

3. As we had seen before, test participants had difficulty explaining what "plan-preferred" and "plan-recommended" mean. Related, test participants also reported a lower level of trust that the pharmacies rendered in search results provided a benefit them vs to their PBM.

My recommendation: these findings validated our hypothesis that members don't clearly understand the terminology we're using; we might consider adding an explanation provided via a tool tip or other help feature. Further, we might consider explaining how a "plan-preferred" pharmacy actually benefits them.