Case Study 02 — Geisinger Health System
Medicare Advantage
The team dissolved. The deadline didn't. One designer, one PM, the entire Medicare Advantage section — shipped before CMS Annual Enrollment opened. +80% digital plan enrollments.
At a glance
The highest-stakes conversion on the site — and I was the only designer left.
Medicare Advantage is where Geisinger seniors enroll in a health plan. The section serves three very different audiences — prospects figuring out eligibility, current members managing renewals, and brokers shopping plans — all under a hard CMS deadline that doesn't move.
In 2020, the team that owned the section dissolved. The institutional knowledge walked out the door. The Annual Enrollment Period date did not change. I had a project manager, a Sitecore CMS, and an eligibility funnel that wasn't converting.
increase in digital plan enrollments post-launch
pages consolidated — cut what didn't convert
designer end-to-end: strategy, IA, UI, copy direction, rollout
01 — The Constraint
One designer. One PM. A live deadline.
The original team had built the section across multiple cycles. When they left, the institutional knowledge left with them — but the CMS Annual Enrollment Period kept its date.
I had no junior to delegate to, no second designer to pair with, and no time for a textbook process. I had a project manager, a Sitecore CMS, and an eligibility funnel that wasn't converting. Scope was non-negotiable: hero, eligibility flow, plan comparison, broker pages, Newly Eligible — all of it.
Crew
1 + 1
Solo designer. Single PM. No researcher, no second designer, no analyst.
Deadline
Oct 15
CMS Annual Enrollment date is non-negotiable. Miss the date, miss the year.
Scope
Full section
Hero, eligibility flow, plan comparison, broker pages, Newly Eligible — all of it.
Coverage
3 audiences
Prospects, shoppers, and current members — all in one section, with zero collisions.
02 — Strategy
Cut scope to what converts.
Without a research budget, I leveraged what I had: site analytics, existing user research from prior cycles, and direct conversations with the call-center team who fielded enrollment questions every day.
i. Audit, not invent
Inherited 30+ pages. Used analytics to find the 8 that drove 90% of the funnel. Cut, consolidated, or merged the rest.
ii. Talk to the call center
Every question reps fielded was a UX failure. Logged the top 12 question themes and turned each into a content or interaction fix.
iii. Design as you ship
Sitecore + a friendly dev meant I could push iterations weekly. Treated the whole section like a continuously-deployed product.
03 — The Redesign
An eligibility journey that actually finishes.
Before
After
03.A · Audience routing
One section. Three audiences. Zero collisions.
Above the plan types: three entry cards — Understanding Medicare, Ways to Enroll, Already a Member. Prospects, shoppers, and current members each pick their own track on the first scroll.
03.B · Learn about Medicare
Personas as entry points.
"I am turning 65 soon" · "I have original Medicare only" · "Can I switch Medicare plans" · "I am a Caregiver." Four cards, four journeys — instead of one generic learn more funnel.
03.C · Newly eligible timeline
A timeline that turns "what do I do?" into "do this next."
Replaced the form-first hero on the New to Medicare page with a five-stage chevron timeline — Age 64 → 64+9 → 65 → 65+3 → 65+9 — each with the specific action to take in that window. Plus a downloadable Medicare Roadmap.
03.D · Ways to Enroll component
One component, six conversion paths.
A reusable action row — call, schedule a home visit, find a meeting, compare plans, select a plan, enroll — that lived at the bottom of every Medicare Advantage page. Same component, six high-intent CTAs. Reused across the section.
"The team left, the deadline didn't. I shipped it because someone had to — and the metrics said it was the right shape."Veronika — Personnel Log, 2020
04 — Impact
Telemetry after launch.
digital plan enrollments after launch
pages consolidated — cut 22 low-value pages
call-center question themes turned into onsite content fixes
designer end-to-end: strategy, IA, UI, copy, rollout
Reflection
The constraint wasn't the team leaving — it was learning that most "essential" pages aren't. Cutting 22 of 30 pages was scary. The metrics said it was the right call.
If I picked it back up: a moderated test with newly-65 users on the eligibility timeline. The redesign worked, but I want to know how it lands for people who've never thought about Medicare before. The call center became my proxy users — that collaboration outlasted the project.