Turning a Tool Pharmacists Were Gaming Into One That Changed Patient Outcomes

02 The Problem

The Patient Care Portal was built by engineers with no UX involvement. It worked - technically. But the pharmacists using it had turned it into something unrecognizable.

In-store pharmacists were gaming the system. The tool required so many steps to complete a single patient interaction that pharmacists developed workarounds to hit their metrics without actually helping patients. Call avoidance behaviors. Metric manipulation. The tool was measuring compliance with itself, not patient outcomes.

Call center HOPs built shadow infrastructure. Health Outcomes Pharmacists needed to handle high volumes of outbound calls, but the portal wasn't designed for that workflow. They'd built custom button configurations, ran dual monitors with personal spreadsheets alongside PCP, and developed individual configs that made the tool unrecognizable from desk to desk.

Two user populations. Same tool. Neither was using it as intended. Both had legitimate reasons.

05 What I Inherited

The original Patient Care Portal required pharmacists to navigate through 10 distinct steps before they could have a single meaningful interaction with a patient. Every step was a point of friction. Every step was a reason to game the system instead of using it.

06 The Redesign

I collapsed the entire workflow down to two steps. Not by removing functionality - by restructuring how information was presented and when actions became available.

Step 1: Unified Patient Profile

Login, store selection, patient search, identity verification, and history review all collapsed into a single view. The system auto-populated context based on the pharmacist's location and the patient interaction. Everything the pharmacist needed to know was visible the moment a patient appeared on screen - no clicking through tabs, no searching, no verifying what the system already knew.

Step 2: AI-Driven Services Scroller

Instead of manually selecting a service type, entering notes, submitting, and logging - the system surfaced relevant services automatically. An intelligent scroller presented the most likely actions based on the patient's medication history, adherence patterns, and open care gaps. The pharmacist chose and confirmed. Done.

07 The Design Decision That Almost Failed

My first redesign was clean. Plenty of whitespace. Progressive disclosure everywhere. Information tucked away behind expandable sections. It was the kind of design that looks great in a portfolio and tests terribly with pharmacists under time pressure.

They hated it. Not because it was ugly - because it was slow. Every collapsed section was a click. Every hidden detail was information they needed visible at a glance. In a pharmacy context, density isn't clutter. Density is clarity.

08 Navigating Approval

Walgreens doesn't ship fast. Everything passes through an 8-gate approval chain. The design work was maybe 20% of the timeline. The other 80% was navigating the organization - aligning stakeholders, surviving UAT, creating training materials, getting legal sign-off, and managing phased rollout across thousands of stores.

10 What Didn't Work

Not everything landed the first time. Some of the hardest lessons came from designs that were technically sound but organizationally wrong.

Early designs optimized for design principles, not reality. My first pass prioritized progressive disclosure and clean aesthetics - principles I'd internalized from years of consumer UX work. Healthcare pharmacists don't want clean. They want fast. The redesign happened after I spent time watching pharmacists use the early prototypes and realized they were fighting the interface, not using it.

Stakeholder translation friction. Ten business stakeholders meant ten different mental models of what PCP should be. Getting alignment wasn't a design problem - it was a communication problem. I learned to present designs in the language of business outcomes, not UX principles.

Phased rollout surfaced problems at scale. What worked in UAT with 50 stores broke in unexpected ways at 200. Network latency, hardware variation, screen resolution differences across pharmacy locations - all things that don't show up in controlled testing.

11 What I Learned

When your tool is used across 8,700+ locations, every second you save per interaction compounds into millions. A two-click reduction in workflow doesn't sound dramatic until you multiply it by tens of thousands of daily patient interactions. The 40% increase in outbound calls wasn't because pharmacists worked harder - it was because the tool stopped wasting their time.

Navigating complex approval isn't overhead - it's the job. At Walgreens scale, the 8-gate approval chain exists for good reasons. HIPAA, patient safety, legal liability. Learning to work within those constraints - not around them - was the skill that actually shipped the product. The design was done months before it reached patients. The organizational navigation is what got it there.

Tying design to business outcomes changes every conversation. When I stopped talking about user experience and started talking about operating income, patient adherence rates, and labor cost reduction, the same stakeholders who had been blockers became advocates. Design decisions that would have taken weeks to approve took days - because the business case was the design case.