Large-scale survey experiment, N=4,623

Designed and ran a survey experiment testing reward caps, redemption flexibility, and friction trade-offs for a renter-to-buyer rewards program. Found that redemption flexibility, not reward-cap size, was what actually drove engagement, a counterintuitive result that shaped the final program design. Findings were used to secure leadership buy-in for a 2026 pilot. 

Quantitative pricing study, informed a live product price

Led a study on renter price sensitivity that directly informed the decision to set a rental product's price at $39.99 — a repeat of a pricing study I originally co-led in 2022, this time run independently three years later with the same analytics partner. 

Cross-functional initiative, org-wide adoption

Led a team of 7 to build and launch a toolkit of AI-assisted research tools (research-retrieval agents, PII de-identification tooling) adopted across an entire research organization, plus a hands-on workshop teaching non-technical researchers to use AI coding tools in their own analysis. 

Cross-functional initiative, org-wide adoption

Led a team of 7 to build and launch a toolkit of AI-assisted research tools (research-retrieval agents, PII de-identification tooling) adopted across an entire research organization, plus a hands-on workshop teaching non-technical researchers to use AI coding tools in their own analysis. 

2×2 factorial experiment, n≈1,500

Ran an experiment isolating how a generative AI's conversational style (directive vs. inquiring) affected user trust and perceived control — findings directly shaped the tone of a live AI product. 

Large-scale sentiment survey, n≈2,000, supplemented with random forest/decision tree modeling

Ran a large-scale survey on real estate professionals' sentiment toward AI tools, then applied random forest and decision tree models to identify which characteristics most strongly predicted opt-in behavior. Findings helped shape early decisions on where to invest in further AI product development.