Efficiency analysis is a survival tool for bank managers due to severe competition in retail banking industry. Certainly, it is important to identify the inefficient branches of a bank in building better service management and marketing strategies. The problem here is to eliminate the causes of their inefficiencies in order to improve the inefficient branches. In this study, a two-stage methodology consisting of Data Envelopment Analysis (DEA) and data mining is proposed for solving the problem. Firstly, DEA is used for measuring the relative efficiencies of a bank's branches. Secondly, data mining is used to extract useful information from given data consisting of the characteristics of bank branches, thus allowing bank managers to identify the underlying causes of inefficiencies and helping them to improve their service management. The approach is used in the valuation of the one's branches of the biggest banks in Turkey.