Humana faced a defining challenge in Medicare Advantage: improve Star Ratings while differentiating their value proposition in an increasingly competitive market. Across the industry, traditional approaches to closing care gaps—reminder calls, generic mailers, financial incentives—struggle to drive sustained behavior change. The question became: how do you activate members at scale while respecting that each person's health journey is deeply personal?
Within Humana's organization-wide Bold Goal initiative, we were engaged to pilot a new approach to member activation. The pilot targeted critical Star Measures: medication reconciliation post-discharge, diabetes management, and blood pressure control.
Rather than defaulting to reminder calls and generic outreach, we hypothesized that sustainable behavior change required understanding what actually motivated each member—their intrinsic needs for autonomy, competence, and relatedness.
We developed a gamified behavioral assessment that revealed these motivations, creating a foundation for personalized engagement across 100,000 Medicare Advantage members.
Health insurers excel at identifying care gaps through claims and clinical data—knowing precisely which members need mammograms, diabetes screenings, or medication reconciliation. But this data reveals what needs to happen, not why gaps persist. Traditional engagement tools—reminder calls, generic mailers, financial incentives—treat populations uniformly because they lack the lifestyle and motivational insights needed for true personalization.
The challenge: capture motivational & lifestyle data at scale, integrate it with clinical and social determinants data, and use it to personalize engagement for 100,000 members.
Healthcare outcomes are shaped by far more than medical care. Social determinants of health (SDOH)—where people live, access to transportation and healthy food, environmental hazards, education, economic stability—often matter more than clinical interventions.
For example, neighborhoods with poor air quality and limited grocery access correlate with higher rates of respiratory illness and diabetes, while areas with limited public transportation see lower preventive care participation.
We sourced and integrated SDOH data from government databases, census records, and environmental reports into our data warehouse. Combined with Humana's clinical records and our motivational assessments, this created a multi-layered view of each member—enabling personalization that went far beyond demographics to address the real-world context of their health decisions.
Source: Institute for Clinical Systems Improvement, Going BeyondClinical Walls: Solving Complex Problems (October 2014)
Missingness map showing spots in our database with missing values (black = missing)
Integrating Humana's medical records into our HIPAA-compliant data warehouse was the technical foundation. However, we immediately hit a practical barrier: incomplete contact information. Many member records lacked current email addresses and phone numbers, requiring a parallel effort to locate and verify this data before personalized outreach could begin.
We established a controlled experiment with two populations: a test group receiving personalized communications based on motivational data alongside Humana's standard outreach, and a control group receiving standard communications only. Chi-square goodness of fit tests confirmed the groups were statistically comparable across demographics and health status. This experimental design enabled us to measure the incremental impact of personalization using two-sample proportion tests for each Star Measure.
We designed a five-stage measurement framework tracking member progression: Engagement (program participation), Insights (motivational and lifestyle data quality), Activation (members taking ownership of their health decisions), Gap Closure (Star Measures improvements), and Behavior (sustained preventive care habits). Each stage built on the previous—members who completed the motivational assessment and engaged with personalized communications showed measurably higher rates of care gap closure and maintained those behaviors over time.
We designed a gamified behavioral assessment grounded in Self-Determination Theory (Deci & Ryan, 1985), which identifies autonomy, competence, and relatedness as fundamental human needs. Using card sorting—an established UX research methodology—we created a digital experience that invited members to reflect on their gifts, passions, values, and desired impact. This approach proved far more powerful than traditional external motivators, helping members connect health goals to their deeper sense of purpose.
Caption: card-sorting game used to gather member’s motivations & lifestyle data. Our approach has been informed by various practitioners in the purpose/meaning space, including Richard Leider's work on calling.
Demographic and clinical variables—age, diagnosis, income, location—produce predictable member segments. Adding motivational data from our behavioral assessment revealed non-obvious patterns.
For instance, members expressing strong family-oriented values showed higher rates of missed appointments, suggesting they prioritized caring for others over self-care; an insight invisible in traditional data.
We employed K-means clustering with silhouette analysis to identify natural groupings of members sharing similar motivations and circumstances. This enabled segment-specific communication strategies. For example, family-focused members received messaging about modeling healthy behaviors for loved ones, while autonomy-focused members received content emphasizing personal agency and self-determination.
We designed a content framework with 108 communication tracks and tone variations targeting specific care gaps and motivational profiles. NYU Langone Health experts populated this system with evidence-based clinical content.
Members were routed to content tracks matching their unique combination of clinical needs and motivational drivers, with messaging adapted to their values, priorities, and decision-making preferences.
We grounded our approach in the UK Behavioral Insights Team's BASIC framework (Attractive, Brief, Clear, and Doable). To avoid high-frequency messaging or manipulative tactics, we optimized timing, framing, and presentation to help members act on their own health goals. This approach respected member autonomy—communications aligned with stated values, supported intrinsic motivations, and emphasized informed choice rather than increasing message volume.
Caption: Beyond tracking gap closures (avg 9.4% improvement vs control), we quantified program impact by mapping closed gaps to potential cost avoidance using published research. For example, early detection through mammography represents ~$11,500 in lifetime savings per case (Health Affairs, 2015)—just one of several preventive measures we tracked. Note: Estimates based on peer-reviewed studies; actual results may vary.
The pilot delivered $1,598 in estimated annual savings per engaged member through personalized, bilateral engagement—the first time Humana integrated lifestyle data and two-way SMS at this scale.
Targeted Star Measures showed statistically significant improvements: medication reconciliation post-discharge increased 9.5%, blood pressure control improved 6.8%, and diabetes care increased 5.4% compared to the control group. Overall weighted gap closure across all measures reached 9.4%.
More significantly, non-targeted preventive services showed even larger improvements: breast cancer screenings increased 26.9% and colorectal cancer screenings increased 22.1%. This demonstrated that personalized engagement created broader health activation—members who felt understood and supported in managing chronic conditions became more proactive across all aspects of preventive care, not just the measures we explicitly targeted.
Real-time analytics enabled continuous optimization, translating closed care gaps into quantified cost avoidance using published research on preventive care value.
Health insurers have long recognized the importance of engaging members to improve the value of care
Engaged members received personalized health guidance reflecting their actual priorities and circumstances. The gamified assessment gave them agency to share what mattered, and communications responded accordingly.
Two-way SMS enabled members to identify specific barriers to care—cost concerns, fear, confusion, forgetting—allowing targeted support.
Members whose motivations, lifestyle, and barriers informed their care coordination showed higher participation in preventive services—proving that understanding and addressing obstacles drives better health behaviors.
The pilot validated the economic model: at $1,598 in annual savings per engaged member, optimized engagement across the 100,000-member population could unlock over $100M in annual cost savings—establishing a foundation for member engagement that transforms both patient experience and healthcare economics at scale.
