Web metrics for retailers

In this study, we first propose a set of web usage metrics for multi-channel retailers. We then apply these metrics to data originating from a retailer who operates both an e-shop and a number of traditional stores. We focus in particular on the analysis of k-Means session clusters exhibiting an interest in offline purchases. Web usage metrics measuring the success of Web sites have been proposed before. In the domain of Web merchandizing, success is measured eventually by the number of purchases accomplished. Basic statistics such as conversion and traffic have been proposed to quantify this notion of success. However, when it comes to retailers with multiple distribution channels, there is a definite lack of standard metrics based on web log and transaction data. This paper tries to remedy this void.