Data-driven knowledge discovery in retail: Evidences from the vending machine’s industry

The purpose of this study is to investigate new forms of marketing data-driven knowledge discovery in the vending machine (VM) industry. Data of shopping activities understanding were gathered and analyzed by a system technology based on a RGBD camera. An RGBD camera, already tested in retail environments, is installed in top-view configuration on a VM to gather data that are processed and embedded within a knowledge discovery project. We adopted the knowledge discovery via data analytics framework (KDDA) that was applied to a real-world VM scenario. Real-world case tests, based on more than 17.000 shoppers measured in 4 different locations, were conducted. By using this method it was possible to verify the ability of this approach to generate new forms of marketing data-driven knowledge available on the VM industry. Main novelties are: i) the application of a KDDA project to the VM industry; ii) the use of RGBD data sources for the first time for a KDDA process; iii) the contribution to the practice by supporting retailers to carry out knowledge discovery processes more effectively; iv) the real-world testing process based on 4 locations and more than 17.000 shoppers.