Retail/Consumer

E-Commerce Product Return Reason Dataset — 6.8M Returns with NLP-Classified Reasons

Return records from 3 mid-market e-commerce platforms covering apparel, electronics, and home goods. Each return includes product category, original price, return reason (free-text + NLP-classified into 42 categories), and resolution outcome. Essential for return prediction models and product quality analytics.

Retail Store Checkout Line Video — 14K Hours, Anonymized Faces, Wait Time Annotated

Overhead camera footage from 280 retail store checkout areas with automated face anonymization (Gaussian blur), bounding box person detection, and queue wait time annotations. Each frame includes person count, queue length, and estimated wait in seconds. Built for queue management AI, staffing optimization, and loss prevention models.

Grocery Store Basket-Level Purchase Data — 240M Transactions, Loyalty Card Linked

SKU-level basket data from 1,800 grocery locations across the Midwest and Southeast US. Each transaction includes full item list, unit prices, promotions applied, payment method, and anonymized loyalty card ID. Ideal for demand forecasting, recommendation engines, and CPG brand analytics.

Commercial Foot Traffic Heatmaps — 2.8M POIs, Weekly Cadence, Census Block Resolution

Anonymized mobile device foot traffic patterns for 2.8 million US points of interest (retail stores, restaurants, offices, hospitals). Measured at census block group resolution with weekly visit counts, median dwell time, and visitor home census tract distribution. Powers site selection AI, competitive intelligence, and commercial real estate analytics.