The Data Warehouse Toolkit: The Complete Guide to Dimensional Modeling

Chapter 14. Electronic Commerce

Web-intensive businesses have access to a new kind of data source that literally records the gestures of every Web site visitor. We call it the clickstream. In its most elemental form, the clickstream is every page event recorded by each of the company's Web servers. The clickstream contains a number of new dimensions—such as page, session, and referrer—that are unknown in our conventional data marts. The clickstream is a torrent of data, easily being the largest text and number data set we have ever considered for a data warehouse. Although the clickstream is the most exciting new development in data warehousing, at the same time it can be the most difficult and most exasperating. Does it connect to the rest of the warehouse? Can its dimensions and facts be conformed in a data warehouse bus architecture?

The full story of the clickstream data source and its implementation by companies, such as those involved in electronic commerce, is told in the complete book on this subject, The Data Webhouse Toolkit, by Ralph Kimball and Richard Merz (Wiley, 2000). This chapter is a lightning tour of the central ideas drawn from The Data Webhouse Toolkit. We start by describing the raw clickstream data source. We show how to design a data mart around the clickstream data. Finally, we integrate this data mart into a larger matrix of more conventional data marts for a large Web retailer and argue that the profitability of the Web sales channel can be measured if ...