At The Home Depot, I developed a Python multi-touch attribution model using Shapley values across Google Ads, Meta, and programmatic display. It informed the reallocation of $3.8M+ in media spend and contributed to an 18% improvement in blended ROAS.
I also evaluated campaigns across 10+ geographic markets using geo-based A/B testing and difference-in-differences analysis, estimating approximately $950K in incremental revenue attributable to paid social. My cohort and LTV analysis informed lifecycle campaigns that improved repeat purchase rates by 13%.
At Procter & Gamble (P&G), I developed a marketing mix model for the Head & Shoulders portfolio that supported a 13% improvement in annual media budget allocation. I also built forecasting and customer segmentation models, and analyzed campaigns including a national Ariel product launch.
