Neural networks in action: an engine for the marketing information system. Information technology construction of the neural network decision support system

Summary form only given. This paper describes the third phase of a 3-phase study conducted for improving the efficiency of decision making and operations in sales and marketing activities in a bank. The result of the study have been adopted and incorporated into the infrastructure of the bank. The goal was to define the variables that best explain bank potential, profitability, loyalty and risk. Using these variables, models were constructed for the estimation of the classification factors that enabled us to make more efficient selections of target groups for sales and marketing operations. In this construction, the neural network was able to yield better classification results than the participating sales offices did on an average. The neural network modeled for this study was much faster than the human workforce. Using the neural network, it was possible to classify the material used in this study in relation to three segmenting classification criteria in just one day. The goal of this study-a more effective classification process and utilization of consequential effects-was reached. The additional benefits of lower classification costs, smaller amount of marketing material required, and lesser amount of contacting required, show that the use of the model saves a great deal of money