Integrating Your Data Warehouse into the World of E-commerce

Abstract

In the new economy lots of things change. How we do business changes. How we manage customer relationships changes. And how we build a successful data warehouses changes. Service levels for performance, data freshness, and high availability are escalating at a rapid pace in the e-commerce marketplace for data warehousing. The design and implementation of data warehouse architectures for meeting these new service levels demand advances in deployment techniques that are beyond the reach of most databases. Thankfully, however, not everything has changed. Much of what we have learned about successful data warehousing in traditional implementations is amplified in the world of e-commerce. Access to detailed data is more important than ever. The ability to answer ad hoc questions is critical in the quick paced new economy. Data mining and the quantitative techniques used to develop models for predicting customer behavior play an even more critical role when customers make purchase decisions without any human intermediaries. When we design for e-commerce, we merely need to deliver these traditional capabilities more rapidly, with more up-to-date data, and around the clock. The evolution of existing data warehouses and the implementation of new data warehouses will typically mirror the stages of evolution in the Web site of an organization. In the remainder of this paper we examine the prototypical evolution of e-commerce within an organization and discuss the implications of each step upon data warehouse design and implementation.