Potential increase in revenue
Model accuracy
Potential reduction in operational costs
To strengthen the overall business model of the Coca-Cola bottling refranchising in North America, the six largest Coca-Cola Bottlers announced the formation of an Information Technology services company, Coke One North America (CONA), in 2016. CONA runs huge supply chain operations and manages thousands of vending machines across the United States.
These vending machines are stocked heuristically and require manual effort to properly stock and operate. Product mix is rarely changed and a lot of products are never tried in the vending machines. Therefore, CONA wanted an interpretable model which could recommend products for each vending machine, based on similar vending machines and the most popular products in a particular geography.
Microsoft Azure
Python
SQL Server
Apache Hive
Quantiphi and CONA jointly developed a custom Product Stocking Recommendation Engine, called VendPrime, that combines collaborative filtering and popular demand to generate real-time recommendations for Coca-Cola vending machines across North America. The solution analyzes large volumes of data and provides actionable insights for bottlers, enabling them to function efficiently with accurate forecasting. This also helped drive sales growth and reduce the amount of product that sat untouched in the machines. The recommendation solution further manifests into multiple use cases from small and midsize convenience retailers to full-service restaurants.
