How Coca-Cola uses CARTO & Google BigQuery to Optimize Sales of 700k Vending Machines
Coca-Cola Bottlers Japan Inc. (CCBJI) produces and supplies approximately 90% of the Coca-Cola system’s products in Japan to serve the needs of their customers and consumers spread widely across the country. CCBJI is not only the largest Coca-Cola bottling company in Asia, but also a leading player among over 250 Coca-Cola bottlers operating around the world.
With a network of over 700,000 vending machines across Japan, CCBJI collects a huge amount of data regarding not only the overall sales performance of each machine but also how individual products perform per machine and location.
Historically CCBJI would need to extract the data necessary for analysis from the core system, build their own mechanism to create a data warehouse using ETL tools, and perform various analyzes. The sheer size of data being produced exposed a number of challenges for the company including the length of time needed to return the results of a simple query and complex maintenance of such a legacy system.
CCBJI built an analytics and machine learning platform as a layer on top of existing systems centered on Google Cloud BigQuery with results visualized using CARTO.
With their new platform they are able to use different types of data such as POIs & foot traffic data to understand gaps & opportunities in the network.
They are also able to predict performance by location, providing insights to key stakeholders in sales, marketing, & operations.
CARTO’s fully cloud native platform enabled CCBJI to work natively with their spatial data in BigQuery. The Spatial Extension for BigQuery allows them to eliminate ETL complexity and scalability limits previously encountered with their legacy GIS system.
Our BigQuery Tiler is the only solution to visualize massive spatial datasets without having to move data outside of BigQuery, and it works in minutes without ETL, using SQL.
The Data Science team at CCBJI are able to search, visualize, and subscribe to thousands of relevant locations datasets through our Data Observatory with frictionless access to these data streams in BigQuery.
Learn more here (in Japanese).