Project Details:
Type: 4+ month remote, exploratory research to understand an audience for a new strategic initiative at Virgin Pulse
My Role: Lead UX Researcher
Methods Employed: Semi-Structured Interviews | Directed Storytelling | Role-playing | Affinity Clustering | Personas
Tools used: Microsoft Excel | Microsoft Word | UserInterviews | Zoom | Adobe Premiere | Dovetail | Figma
Project Scope:
Virgin Pulse (VP) is a platform centered around employee health and wellbeing. It is a suite of tools, products, and services to serve as a "homebase for health" - and to better meet the needs of users with health conditions, they enlisted me to do large-scale exploratory research with VP users with Type II Diabetes and/or Hypertension.
I was the lead and solo individual contributor for this research which included two rounds of interviews with 15 and 8 participants respectively and over 1500 datapoints (tags in Dovetail). I used those to create atomic insights for Health Condition Management research at Virgin Pulse.
This study helped inform a 0-1 roadmap for a new Digital Therapeutics tool within Virgin Pulse. While development of that tool began prior to the study's completion, the core team was moved by the insight gathered, and impact extended beyond VP's Product Organization.
This screenshot showcases the video footage of an interview (using a moment showing my face to preserve participant privacy) and an example of my DoveTail data-coding
Semi-Structured Interview Methodology
The strength of semi-structured interviews - going in with a script, but being flexible with it - is that I can lean into participants' responses where the attention matters most. By ignoring irrelevant questions and following up on sensitive or important details, I can articulate "moments of truth" in their health journeys - or the "make it or break it" moments in their health care experiences. The method also helps keep things conversational, improving participant comfort and willingness to share.
I wrote a script and made a note-taking sheet to prepare for interviews, and the first two sections were fairly standard: a health background section and a "health journey" session, walking through important moments from pre-diagnosis to today. That information had potential to be valuable, but I added two additional sections for added potential, outlined below:
Directed Storytelling
The Directed Storytelling method allowed me to scaffold my participants' retelling of a relevant health story. I used this to encourage participants to share two recent events in great detail: their most recent "typical health" day, from start to finish, and their most recent "bad health" day, from start to finish. This gave me vision into routines to know tactically where products or services could fit.
Role-Play: Superhero Movie casting
Inspired by Jane McGonigal's behavior modification game "SuperBetter" and the design thinking concept of "reframing", I ended interviews by asking them to cast the "superhero movie about their health" - who is the villain, who is the sidekick, what is the 'kryptonite', and what is your superpower - ending the interview on a silly yet uplifting note.
I view UX Research as a form of advocacy and a touchpoint of my company's brand. Ending interviews on an optimistic, critical thinking, silly note makes me feel good about how I represent myself and my company - and most interviews ended in smiles.
From Data Collection to Data Analysis
The scope of data was massive, and I negotiated to give the data analysis the time it deserved. The first set of interviews, conducted over Zoom, included fifteen participants, and it was my first time using DoveTail. I coded each interview transcript with help from the notes taken in each session, resulting in over 1000 distinct highlights over the dataset. Then I standardized the codes, reducing the number of codes (or "tags") to cluster.
Our data spanned a wide range of topics, including pain points, positive moments, beliefs, desires, tools, sources of education, and more.
Clustering Round 1: Managing huge amounts of data
The first round of tag-clustering was based more upon high-level "type" than their insightful substance. To manage the sheer number of notes, the macro-clusters were based upon whether they were positive or negative, beliefs or desires, or miscellaneous qualities. Each macro-cluster had sub-categories when appropriate - for example, "Pain Points" included "barriers to success", "negative feelings", "negative thoughts", and "miserable 'moments'".
I coded, color-coded, and organized huge amounts of data in DoveTail for easier analysis
Barriers to Health Success fit into several substantive categories including food, work, and family.
Clustering Round 2: Articulating Substance
Despite standardizing codes into a more manageable number of tags, the tags could be clustered into more meaningful categories. I was able to look at the individual code names and group each into distinct themes - with a select few appearing in more than one theme. In this way, the 48 different tags that were all under the sub-category of "barriers to success" could be clustered into ten groups, and the number of individual highlights from the fifteen interview transcripts per group was easily determined. Comparing these across macro-cluster helped me ask questions that led to deeper analysis - for example:
• Why is food SO present as a barrier to success?"
• Why does family appear approximately equally as a pain point and as a positive point?"
• "What makes interactions with doctors lean toward being positive experiences or negative experiences?"
Impact of Research
The initial round of the study was Virgin Pulse' deepest audience exploration from a UX Research perspective to date. Over thirty cross-functional stakeholders attended the research read-out, and the project had multiple immediate impacts:
• I documented three major insights about people's behaviors when managing their health conditions. These had implications within multiple delivery teams' roadmaps.
• Designers on the delivery team most-impacted by the research collaborated with me to create and update our personas to guide their team's research.
• A non-Product team focused on Virgin Pulse's partnerships looked into opportunities for food-support focused benefits integration on platform
This proto-persona came from research prior to this study. These personas were used in tandem with the atomic insight from this study to inform product design.
Follow-up Research
Another delivery team immediately requested follow-up research of the same protocol to explore the needs of people with pre-diabetes and managing their weight to prevent health conditions (to compare to the needs of those managing conditions).
That follow-up round of research - another 500 highlights clustered and analyzed from eight additional sessions - allowed me to articulate an additional two insights and then define an opportunity-space for the updated digital therapeutics product. All insights and opportunity-space diagrams are not included in this portfolio page for confidentiality's sake, though I am happy to present abstractions via case study.
Future workshopping methods and follow-up concept-exploration studies will determine the viability of pursuing digital therapeutics strategies in any of those spaces or intersections.
This portfolio page may continue to update as we continue to do follow-up research in this domain, though no confidential materials will be shared.