Revealing the retail black box by interaction sensing

Abstract

Today a huge variety of methods to track and analyze the customers' behavior in e-commerce systems is available. However, in traditional retail stores such systems are not widely known and therefore the customers' behavior is considered as a black box in this domain. This paper presents the Smart Shelf technology able to track basic simple actions, such as take, return and remove, which are performed on items by the customer. These actions form the interaction context replacing the black box. We will show that this context can be used to enhance existing data mining and store management systems as well as the customer will benefit from recommendation systems comparable to those used in e-commerce systems of online stores.