This chapter discusses why data mining is relevant today, both to
consumers of data mining results and users of the technology. It then
introduces and explores data mining at a high level, contrasting data
mining with other forms of advanced analytics and reviewing the
basic data mining process. Since the mining metaphor is often used,
we digress to contrast the data mining process with the gold mining
process. To be relevant, data mining must provide value. We discuss
the reliability of data mining results and highlight ways in which
data mining adds value to businesses.
Why Data Mining Is Relevant Today
Today's business landscape is highly competitive. Product margins
are typically low due to increased competition and commoditization.
Consumers have more information about competing offers and more
channels such as the Web to pursue them. Customer loyalty exists
either as long as the customer experience [Shaw/Ivens 2002] remains
positive or until a better alternative comes along.
Savvy businesses have long taken advantage of advanced analytics
and data mining to give them an edge in the marketplace. Major
retailers know which customers to target with ad campaigns. Manu-
facturers know how to determine which aspects of their manufactur-
ing process are yielding inferior results and why. Financial services
providers, such as banks, know which customers are a high risk for a
loan [Davenport 2006].
Companies that do not leverage data mining in their business
processes are not likely to realize their revenue and profit potential.
Their customers' experiences may be inferior as they become
fatigued by what appear to be random solicitations or irrelevant
offers. Companies may be missing key insights into ways to deter-
mine why customers are leaving or what customer profile yields the
highest customer lifetime value.
From another angle, various regulatory compliance measures
(e.g., Sarbanes-Oxley [SOX 2006], Basel II [BIS 2004]) require the
keeping of large quantities of historical data. As such, multi-terabyte
data repositories are becoming commonplace. Many companies
make dramatic efforts to collect almost everything about their
businesses, and to ensure that the data are clean. Consequently, cor-
porate executives want to put this costly asset to good use. One such
use is in the area of Business Intelligence (BI), which traditionally
involves extracting information, generating reports, and populating