Computational Finance Techniques for Valuing Customers

Understanding the value a customer has to a business is a fundamental problem. Accurate valuations are critical for setting appropriate levels of investment for targeted marketing and for the setting of individual customer service levels. Traditionally semi-qualitative methods using results from surveys and simple extrapolation of historical profit data have been used for this purpose. In this work we consider customers as assets, producing a time series of profit events. Taking this view, concepts from financial engineering become natural means to analyse customers. We make novel adaptations to approaches from quantitative finance to create more accurate customer valuation techniques. Using these ideas, we show that the uncertainty in the future profits created by customers provide an extra dimension to their value. Application of computational approaches using the mathematics of options analysis are presented. These can be used to gain a true picture of the expected profitability of a customer and the potential value a customer has as a cross or up sell opportunity. We show how these techniques can be developed into computer based decision support tools for use in customer contact environments for choosing appropriate responses to customers, such as attempts to prevent customer churn or selecting the type of promotion to offer a customer. We also discuss how ideas markets, portfolio theory and agent based modelling can be used to augment the options analyses, to give a more complete picture of the value of a customer.