Practical Business Intelligence with SQL Server 2005: Data Mining
As you have seen, Analysis Services enables you to build powerful Business Intelligence (BI) solutions that enable users to really understand the business. However, many business problems rely on the ability to spot patterns and trends across data sets that are far too large or complex for human analysts. Data mining can be used to explore your data and find these patterns, allowing you to begin to ask why things happen and to predict what will happen in the future.
In this chapter, we look at how to use some of the data mining features in Analysis Services 2005 to perform tasks such as customer segmentation and market basket analysis. The data mining results are presented in the form of new dimensions in cubes and are used in Web applications.
Our customer for this chapter is a large music retailer with stores across the country, and which also has an e-commerce site where customers can buy CDs. The retailer has also moved into the broader entertainment market and added product lines such as videos, computer games, and, more recently, DVDs. This latest product line has just been added to the Web site so that customers can buy DVDs online.
The retailer faces strong competition in the online DVD market and is struggling to achieve profitability and gain market share. Its e-commerce system has built-in capabilities for conducting marketing campaigns and performing analysis; however, this is restricted to information learned from customers' online behavior and does not tie back into the retailer's extensive data warehouse, which is populated mostly with information from their stores.
This has led to the following challenges:
- There is currently no way to segment customers by combining the extensive customer profile information with the Internet-usage metrics. This segmentation is needed so that they can target direct mail and other marketing to segments that will potentially use the Internet channel.
- The profit margin on DVD sales is low because of extensive competition. The retailer needs to find ways to increase the value of items sold in a transaction, such as by promoting and cross-selling additional products at the time of the purchase.