Digital-Offer-Targeting-Merchant-Engagement-Analytics

This project is an end-to-end analytics demonstration combining incrementality/uplift modeling on the Criteo Uplift dataset with merchant engagement scoring and GenAI-driven sentiment analysis on the Yelp Open Dataset. It is built to showcase data science skills relevant to digital offer targeting and merchant health monitoring. The Criteo workstream analyzes ad targeting incrementality, finding a full-dataset lift from control 0.194% to treatment 0.309% conversion (+59.4% relative lift, p ≈ 7.3e-179) and identifying four statistically-attractive customer segments. A T-Learner uplift model shows that top-decile customers capture 65% of test-set conversions, with Top 10% targeting beating random by ~4×. The Yelp workstream scores 8,812 restaurants, classifying 48.4% as Stable, 26.8% as Declining, and 24.8% as Growing. GenAI sentiment analysis on 2,193 reviews from 300 merchants agrees strongly with star ratings (Pearson r = 0.92). A priority framework splits merchants into four actionable groups (Expand / Monitor–Intervene / Growth Opportunity / Reassess), each with a distinct recommended action.