"Honey Gold" meant something to Honey users and nothing to PayPal's hundreds of millions. Every word was up for grabs: the name, the currency, the earn language, the redemption flow and the fine print.

I drove the decision to retire Honey Gold and introduce PayPal Rewards—new for PayPal users, a refresh for Honey users. From there I set the mechanics language ("cash back," two words, sentence case) and wrote the copy that taught two audiences one program.

Points-first earning created escheatment risk, and legal wanted lengthy disclaimers everywhere points appeared.

I negotiated the language down to what the law required—setting precedents like never showing a dollar value for unspent points—then documented every naming call so no one had to relitigate them. That framework is still PayPal's source of truth today.

Stash launched as a debit-and-balance rewards program, but Venmo's rewards had outgrown it. Credit card cash back and crypto incentives floated outside the name with no labeling or unifying principles. The ad-hoc system meant writers, marketers and PMs were each solving naming locally—and users met a different vocabulary on every surface.

I ran a structured portfolio analysis with two goals: figure out how Venmo should organize and position everything living under "rewards," and show how the taxonomy changes under different structural architectures. Then I built three models and pressure-tested each.

Exploration 3 won because it was the only model that unified every reward under one system—and the only one that could absorb new reward types without structural changes. Naming the parts descriptively under one umbrella made the taxonomy coherent and extensible at once.

I made the case that terminology is a product system, not a campaign asset: it has to survive every surface, every legal review and every future feature. Content design won ownership, and with it the ability to keep the taxonomy consistent as the program scales.

Walking the flow as a user made the problems concrete: category tiles with bare labels and no definition, no demonstration of real-life value, a wall of options in an indigestible format and a CTA that didn't say what would happen next.

Listen first. Be the user and flag every pain point. Survey the competitive landscape. Then decide what we can actually test.

I partnered with UX research to run several content tests, and worked closely with experience design to bring merchant logos into the tiles—instant recognition doing the work that copy alone couldn't.

Four content moves, each one testable: give every category tile a definition users could recognize themselves in ("Groceries—the essentials, in store or delivered"), demonstrate concrete value at the moment of choice, restructure the list into a scannable format and rewrite the CTA to name the action—"Set [Month] category" instead of a generic "Next."

The newly added Rideshare & Public Transit category captured 3.4% of selections in its first month and drove the strongest transaction activity of any category for rewarded users.