
Pfizer’s Helix Design System is the brand-agnostic foundation used across its vast portfolio of drug and support websites. Each with its own themes, regulatory quirks, and brand constraints. In 2021, Pfizer hired Graphite Digital to migrate all of these sites from Sketch to Figma using the newly updated Helix 3.0 design system, while also re-capturing and reframing their user journeys across in-person, digital, and call-centre touch-points.
I joined the project mid-flight as a Senior UX Designer, one of five seniors embedded in a 20+ designer force supported by three excellent project managers who heroically kept an incredible amount of chaos neatly contained.
We were tasked with producing high-fidelity user flows, holistic journey maps, and interactive prototypes for over 80 websites and 100+ flows, all while the backlog kept growing and last-minute changes were the norm.
While my main responsibility was executing end-to-end redesigns and user flows in Figma, I also helped push the team’s operational efficiency forward. I advocated for consistent use of auto-layout and created reusable scaffolds to accelerate flow creation. I also documented and shared improved design-ops practices across the team.
This case study matters to me because of the design-ops contributions I made to improve the system behind the work, instead of the work itself.
The Challenge
The scale and complexity of this project were unlike anything I’d worked on before. We were systematising workflows across 80+ pharmaceutical brands and 100+ user flows, each with its own constraints, timelines, and review chains. The work moved fast, but it demanded precision. And the only way to keep up was to operate with structure, discipline, and shared standards.
Here’s what the typical delivery loop looked like:
- Receive a detailed brief (e.g. a journey spanning a product site, a visit to a healthcare professional, and a follow-up phone call to Pfizer).
- Create the corresponding user flow in Figma.
- Export and upload that flow to InVision.
- Hotspot the flow for interactivity.
- PM reviewed the output against the brief and narrative logic.
- A peer designer checked for visual consistency and design details.
- The flow was linked into broader journey maps.
- Create the website designs in Figma using Helix components, based on legacy Sketch references.
- One or more designers reviewed the layout depending on complexity.
- Upload the designs to InVision and hotspot them for full prototype fidelity.
- PM reviewed the result again against the brief.
- Link the final design back to the related flows and sitemap.
The structure was strong, but in practice, several friction points slowed us down:
- Manual layout overhead: Everything was positioned absolutely. There were no auto-layouts or responsive scaffolds, which made even small changes time-consuming and error-prone.
- No shared patterns for non-digital interactions: Scenes like doctor visits or phone calls had to be built from scratch using illustrations and boxes, with no consistent visual system across flows.
- No scaffolding for flow diagrams: Each user journey was composed manually, with designers solving the same layout and narrative problems over and over.
- Compliance-sensitive designs: Strict rules around ISIs (Important Safety Information) dictated visual hierarchy, formatting, and placement. Also varied depending on market (U.S. vs. international). These requirements had to be respected at all times.
- Frequent brand switching: Designers typically handled 2–3 brands per week, often jumping into work started by others. That made consistency, structure, and handover quality critical.
- High turnover: New designers were onboarded every few weeks, meaning documentation and shared process weren’t just nice-to-haves.
Key Solutions
1. Introduced the use of auto-layout
I collaborated with another senior designer to advocate for a full shift to auto-layouts across all designs. This replaced the fragile, manually-aligned positioning with structured components and predictable behavior. We mentored other designers on how to use auto-layout effectively, and defined it as a QA checklist item for all page work going forward and retroactively restructured the earlier designs, rebuilding them with auto-layout to bring them up to the new standard.
We also began building larger “organism” level components to reinforce design patterns across brands, cutting down repetition and improving visual consistency.
2. Defined a system for user journeys
To speed up and standardise user flows, I developed a dedicated system for composing journey scenes in Figma. It used:
- Component variables for fast configuration
- A curated, inclusive set of illustrated people, places, and objects
- Material icons and visual storytelling elements (e.g. panel dividers, phone frames)
This turned a previously manual, inconsistent task into a faster, modular workflow. Something even new joiners could pick up quickly.
3. Training and Guidelines
To make all of this stick, I created a set of working guidelines and tutorial videos covering both design best practices and how to use the new systems. These became part of the standard onboarding for new designers. Especially useful given the high turnover on the project.
None of this changed the brief. But it changed how the team worked! Fewer layout issues, faster flow creation, and less ambiguity for anyone picking up a file mid-stream.
Outcomes & Impact
The systems and improvements I introduced were admittedly loud, but they had real downstream effects. Layout issues dropped. User flows became much faster to produce, talking about hours, to minutes. New designers onboarded more smoothly and started contributing earlier. The work felt less fragile, more structured, more predictable.
A last-minute amendment to update content at the top of a completed page used to be a scary, time-consuming task. It turned into a five-minute chore.
While I couldn’t track the results quantitatively myself, the project managers made a clear distinction: after these changes were adopted, overall delivery speed improved by roughly 30%.
The design work didn’t change in scope, but the team’s ability to deliver it improved significantly.
Final Thoughts
What stands out to me about this project is that people often treat speed and structure as opposites, like you can have one, but not both. But in this case, the structure was the speed.
Everything I talk about here was done, while keeping pace (and then some) with my regular workload. In fact, I became the person to go when something needed done quickly.
I wasn’t doing all this in spite of the design-ops work. I was able to do it because of it.