Consumer Demand Modeling
With over 500,000 items sold in more than 9,000 stores, this Top 10 US retailer faces the weekly task of accurately predicting consumer demand to ensure it has the right products at the right stores at the right time. With its prior forecasting system, the retailer was unable to accurately forecast store-level promotions, resulting in a double loss: lost sales where forecasts underestimated demand, and excess inventory when it was overestimated. With billions of dollars in promotional sales, the retailer understood that it was potentially leaving hundreds of millions of dollars on the table.
Its traditional solution, from an large software vendor, simply could not handle the complexities of promotional forecasting or the scale of data analysis required. Its forecasts were rudimentary, given they were designed before the era of the cloud and Big Data. Finally, the retailer’s inaccurate forecasts prevented it from executing the effective sales, specials and discounts needed to drive higher margins and sales while minimizing inventories.
With the LogicBlox smart database, a tailored promotions forecasting and management application is driving significant benefits for the retailer:
- $150 million in reduced inventory
- 25% increase in store-level inventory productivity
- 50% higher accuracy than prior forecasting system
The retailer now operates with higher forecast accuracy on 70% of promoted items and a 25% increase in store-level inventory productivity. One major supplier to the retailer reported that the new forecasts are 50% better than the prior forecasting system – a tremendous advantage in their supply chain and vendor management.
The solution provides the retailer with a unified baseline and promotional forecast that drives promotion planning and inventory/purchasing decisions. The application runs on the cloud, deploying up to 900 servers on demand to process terabytes of data. The resulting baseline and promotional forecasts are available to the retailer’s merchandising and supply chain teams for them to manage and run simulations on. By harnessing the power and elasticity of the cloud to perform forecasting and analytics, buyers and merchandisers benefit from improved real-time decision making with next-generation response times measured in seconds.