The client is a large supermarket chain with over $20 billion revenues.
The client was facing a significant challenge with store stockouts while also experiencing excess inventory situations, particularly for general merchandise and import items. The root causes of these challenges were high forecast errors and disconnects between forecasting and replenishment.
Core Compete conducted diagnostics to identify process changes and solution capabilities needed to improve forecasting and replenishment. It deployed SAS Forecasting to automate the statistical forecasting process. It also developed a retail forecasting methodology for store-SKU-week forecasting.
Initial diagnostic engagement to identify the true root causes for stockouts, excess inventory, and forecast error
Completely automated statistical forecasting and forecast model selection process
A parameter driven automated weekly process to integrate forecasting and replenishment
Core Compete in Action
Utilized weekly sales, warehouse and store inventory and master data to develop the store-SKU-week forecasts.
Developed time-series models for forecasts at different levels of the hierarchy for the 52 weeks horizon. Models accounted for seasonality, price, product-lifecycle, and trend elements. Utilized store-level inventory and sales patterns while developing rolled-up vendor order quantities.
SAS Forecasting and SAS Data Integration Studio were utilized to develop a highly automated solution that integrated with the Teradata warehouse, and generated weekly forecasts that supported general merchandise ordering.
The entire solution was deployed within 13 weeks with many critical milestones achieved in just the first 3 weeks.
Through the new solution, the client was able to realize a 10.5% reduction in forecast error in many key product categories with an attributed inventory reduction benefit in excess of $112 million.