This chapter examines the business impact of predictive analytics. It argues that in order to understand the potential business impact of a predictive model, an organization must first evalute the model with technical metrics, and then interpret these technical metrics in terms of their financial business impact. This chapter first reviews a set of technical metrics which can assist in analyzing model quality. The remaining portion of the chapter then shows how to combine these technical metrics with financial data to study the economic impact of the model. This know-how is used to illustrate how a business can choose the best predictive model from among two or more candidate models. The analysis techniques presented are illustrated by various sample models from the domains of insurance fraud prevention and predictive marketing.