Menu

Turning Around a Stalled Diabetes Management Pilot

Turning Around a Stalled Diabetes Management Pilot

KAISER PERMANENTE

Role: Director, Mobile Design & UX

Initiative: Condition Management Mobile Experience - Diabetes Pilot Redesign

KAISER PERMANENTE

Role: Director, Mobile Design & UX

Initiative: Condition Management Mobile Experience - Diabetes Pilot Redesign

At a Glance

Situation

A vendor-built diabetes management pilot had stalled, with low engagement, high abandonment, low CSAT, and weak adherence to recommended treatment.

Leadership move

Leadership move

I took ownership, established a baseline, spoke directly with patients, and reframed the choice from patching the existing app to rebuilding a smaller, higher-value MVP.

Team model

Team model

A close cross-functional team brought together UX design, research, clinicians, patients, Engineering, Product, and a clinical psychologist using an SVPG-inspired product operating model.

Outcome

Outcome

The redesigned pilot launched later that year, increasing engagement and CSAT, reducing abandonment, and improving treatment adherence by 20%+ among app participants.

The Challenge

Kaiser Permanente had commissioned a vendor to build a pilot mobile application for diabetes condition management. The intent was strong, but the experience was not producing the behavior or satisfaction the program needed. Engagement had stalled, abandonment was high, CSAT was low, and adherence to clinically recommended treatment remained weak.


When I took over the initiative, the question was not simply how to improve screens. We needed to determine whether the existing product was worth continuing to invest in, what patients actually needed from the experience, and how to improve both the user experience and the clinical outcome without allowing sunk cost to drive the decision.

My Mandate

My responsibility was to stabilize the initiative, create an evidence-based path forward, align leadership around the tradeoff, and give the team a product model that could deliver measurable patient value within the year.

My responsibility was to stabilize the initiative, create an evidence-based path forward, align leadership around the tradeoff, and give the team a product model that could deliver measurable patient value within the year.

CORE LEADERSHIP DECISION

Rather than spend more time and resources retrofitting a low-performing vendor product, I recommended preserving the useful learning, reducing the scope, and rebuilding a focused MVP around the highest-value patient and clinical needs.

Rather than spend more time and resources retrofitting a low-performing vendor product, I recommended preserving the useful learning, reducing the scope, and rebuilding a focused MVP around the highest-value patient and clinical needs.

How I Approached It

  1. Put patients into the decision process

    I spoke directly with end users to understand where the experience created friction, where people disengaged, and which parts of condition management felt most difficult to sustain. Those conversations helped us separate feature requests from the underlying problems patients needed the product to solve.

  2. Reframe the investment decision

    The evaluation showed that retrofitting the existing application would require more time and resources than rebuilding a smaller product around the core value proposition. I brought that tradeoff to stakeholders and leadership and secured alignment to move forward with a scaled-down MVP rather than continue patching the existing pilot.

  3. Build a genuinely multidisciplinary product team

    We used the Silicon Valley Product Group (SVPG) operating model as a guide and formed a close working team across UX design, research, clinicians, patients, Engineering, Product, and a clinical psychologist. The point was not simply representation. It was to keep product, clinical, technical, behavioral, and patient perspectives in the same decision loop from discovery through delivery.

  4. Design for value first, then scale

    The team worked together to scope, design, and build the MVP around the highest-value interactions. We deliberately kept the first release focused instead of recreating the full feature set. At the same time, we designed the product so the approach could become a template for additional condition-management experiences rather than a one-off diabetes solution.

  5. Launch and measure against the same baseline

    The redesigned pilot was released later that same year. After launch, we returned to the same categories we had baselined at the start so the team and leadership could assess whether the new product was actually changing behavior and satisfaction.

Results

We re-sampled the same categories used in the baseline so the team could compare the new experience against the original state.

Measure

Before

After

Engagement

Stalled / Low

Higher engagement after the redesigned pilot launched

Abandonment

High

Lower abandonment

CSAT

Low

Higher customer satisfaction

Treatment adherence

Low

20%+ increase among app participants

Scalability

Single vendor-built pilot

Designed as a reusable template for additional condition-management experiences

Why the Approach Worked

• Evidence before solutioning: We did not start with a redesign concept. We first established the baseline and used patient input to define the actual sources of friction.

• A smaller product with a clearer job to do: Reducing scope gave the team room to concentrate on value instead of preserving every feature in the existing application.

• Patients and clinicians inside the product loop: Continuous input from end users and clinical partners helped the team validate both usability and clinical relevance.

• A decision model leadership could support: Framing the choice as a resource and value tradeoff made it easier to move beyond sunk cost and align around a focused rebuild.


• Evidence before solutioning: We did not start with a redesign concept. We first established the baseline and used patient input to define the actual sources of friction.

• A smaller product with a clearer job to do: Reducing scope gave the team room to concentrate on value instead of preserving every feature in the existing application.

• Patients and clinicians inside the product loop: Continuous input from end users and clinical partners helped the team validate both usability and clinical relevance.

• A decision model leadership could support: Framing the choice as a resource and value tradeoff made it easier to move beyond sunk cost and align around a focused rebuild.


What I Learned

• End users need to be involved from beginning to end, not brought in only to validate a nearly finished solution.

• There are times when the fastest responsible move is to keep the learning, stop patching the existing product, and start over with a clearer scope.

• A smaller MVP that solves the most important problems well can create more value than a feature-heavy product that users abandon.


• End users need to be involved from beginning to end, not brought in only to validate a nearly finished solution.

• There are times when the fastest responsible move is to keep the learning, stop patching the existing product, and start over with a clearer scope.

• A smaller MVP that solves the most important problems well can create more value than a feature-heavy product that users abandon.


LEADERSHIP TAKEAWAY

The turnaround was not driven by adding more features. It came from using evidence to make a harder portfolio decision, narrowing the product to its highest-value purpose, and keeping patients, clinicians, design, product, and engineering tightly connected through delivery.

The turnaround was not driven by adding more features. It came from using evidence to make a harder portfolio decision, narrowing the product to its highest-value purpose, and keeping patients, clinicians, design, product, and engineering tightly connected through delivery.

Back to Case Studies