Application of data mining techniques for customer lifetime value parameters: a review

Abstract

Computational and digital advancements with the advent of relationship marketing have changed the land signs of business. Digital revolution led to generation and collection of data in companies and extracting knowledge from this data through knowledge discovery in databases (KDD) process. KDD involves many steps, of which an important step is data mining. Data mining is a process of extracting patterns in data through statistical and other techniques and algorithms. In business, firms are shifting their marketing approach from mass marketing to relationship based marketing leading to an era of customer relationship management (CRM). CRM requires sustainable long term relationship with customers and allocation of resources to maintain these relationships. Customer lifetime value (CLV) is a metric to justify resource allocation by segregating customers on the basis of their contribution to the company. In this paper we review applications of statistical and data mining techniques for predicting CLV and its parameters. The applications of techniques such as logistic regression, decision trees, artificial neural networks, genetic algorithms, fuzzy logic and support vector machines are covered. In the end, a case study is presented to estimate few CLV parameters for a direct marketing company.