PayPal's marketing teams needed a way to reach hundreds of millions of customers across global markets with targeted, personalized experiences — the right message, to the right segment, at the right moment. The tool they were building to do that, Personalization Studio, was mid-construction when the agency that designed it handed it off.
I was brought in as the sole designer to take ownership of everything that came next.
I was brought in as the sole designer to take ownership of everything that came next.
What I inherited
Tallwave, a design agency, had taken the platform through seven major versions before handing off to me directly. What I received was a solid foundation with significant surface area still unbuilt — features that existed in roadmap but not in product, and a user base that spanned campaign operations teams, marketers, and analytics leads with very different levels of technical fluency.
The underlying system was genuinely complex. Interactions were built from layered audience segments, contextual targeting rules, multi-region scheduling, and approval workflows that touched multiple stakeholders before anything went live. My job was to make that complexity navigable without flattening it.
Working embedded with an engineering-led product team, I extended the platform across several fronts. I mapped the full interaction architecture — tracing how campaigns, offers, target groups, creatives, and placements related to each other — and used that structure to identify where the experience broke down for non-technical users.
The most significant addition was an in-house experimentation feature. Rather than routing teams to a third-party testing platform, I designed an A/B testing workflow directly inside Personalization Studio — covering audience splits, treatment assignment, and message-level control. This kept the experimentation loop inside the tool teams were already using and removed a costly external dependency.
I also designed the contextual parameters system: a rule-builder that lets marketers layer AND/OR logic across account attributes, transaction history, and behavioral signals to define precise audience conditions. The challenge was making that level of specificity accessible to people who weren't thinking in code.
This project taught me how to design at the intersection of marketing, data science, and engineering — and how to advocate for usability without losing respect for the technical constraints that made the system work. It was also my first lesson in what it means to own something someone else started: you have to understand it deeply before you can extend it well.