Use case
Project Background
One of our clients, a consumer product company, places more than 10 million items into over 300,000 sales channels weekly. The challenge is predicting which products to sell at each outlet and in what quantities to manage their supply chain effectively and prevent stockout issues. Additionally, the company aims to boost revenue by expanding the number of SKUs sold at each outlet. We were asked to build a predictive model to address stockout problems and enhance revenue through new SKU recommendations.
Our Approach
Our team employs design thinking to grasp the purchasing needs and behaviors of outlets, ensuring our model truly aligns with their business requirements. We identified that the previous SKU recommendation technology was burdened with excessive factors, causing slow processing and inaccurate results. We delivered an enhanced model that prioritizes crucial internal and external factors, significantly reducing processing time from two days to just four hours, allowing our client to run the model daily. Thanks to its speed and accuracy, we swiftly helped our client improve revenue, and achieved payback within only a few months.