Case Study: PepsiCo boosts retail sales with Sigmoid’s ML-driven recommendation engine

PepsiCo boosts market share by 2% with Sigmoid

PepsiCo, a leading CPG company, faced challenges with inventory management across its retail stores in Columbia. Their stores lacked a unified methodology for forecasting, leading to frequent understocking and overstocking issues. They also had no scalable way to identify high-potential product whitespaces. Sigmoid was engaged to develop an ML-based recommendation engine to address these challenges.

Sigmoid developed a solution that pulled data from SAP and a third-party source to create a master data pipeline. Their system used a hybrid recommendation model to suggest new products for each store and a forecasting model to determine optimal order quantities. This solution provided PepsiCo with a 1.5% improvement in profitability, a 2% point increase in market share, and a 66% reduction in the time spent manually generating recommendations each week.

PepsiCo