A metric for customer lifetime value of credit card customers

A metric for customer lifetime value of credit card customers Received (in revised form): 8th July, 2008 Harsha Aeron is a doctoral student in IT and systems at the Indian Institute of Management, Lucknow, India. His main research interests are business intelligence, data mining applications, customer relationship management and customer lifetime value. Tarun Bhaskar is working as a lead scientist with GE Global Research in Bangalore, India. He received his PhD from the Indian Institute of Management, Calcutta in the area of Operations Research. His main research interests are customer relationship management, decision-making under uncertainty, soft computing techniques and combinatorial optimisation problems. Ramasubramanian Sundararajan works as a lead scientist with the Computing and Decision Sciences Lab in GE Global Research, India. He received his PhD from the Indian Institute of Management, Calcutta in the area of management information systems. His work concentrates on the application of predictive modelling and optimisation techniques to varied problems in engineering and management. His research interests are machine learning, soft computing and optimisation. Ashwani Kumar is presently an associate professor in IT and systems at the Indian Institute of Management, Lucknow. His current research interests are business intelligence, data mining and knowledge discovery, and computational intelligence and its applications in business. He received his PhD from the ABV-Indian Institute of Information Technology and Management, Gwalior, his MS from National University of Singapore, his MBA from University of Melbourne, Australia and his B Tech (EE) from the Indian Institute of Technology, Kanpur. Janakiraman Moorthy is Professor of Marketing at Pearl School of Business, and on leave from Indian Institute of Management, Lucknow. He specialises in advanced marketing research, new product development, customer value creation and market orientation of firms. His recent publications were in Journal of Database Marketing & Customer Strategy Management and Marketing Science. His current work is focused on reviewing the methodologies for customer valuation and marketing productivity analysis. He received his PhD from Indian Institute of Management, Ahmedabad, India and was Global Research and Project Director of Institute for Customer Relationship Management, Atlanta, USA. Keywords customer relationship management, customer lifetime value, credit card, financial services Abstract Estimating customer lifetime value (CLV) is becoming increasingly important in order for firms to identify and invest on prospective profitable customers. A credit card issuer firm has to take several different decisions regarding a customer throughout her stay with the firm. CLV estimation can help a firm in making some of these crucial decisions. In this paper, we have presented a conceptual model for revenue from a credit card customer and have further presented a metric for CLV. This metric has been designed specifically for credit card customers. We have simulated different states of a customer to demonstrate how the proposed metric works. Journal of Database Marketing & Customer Strategy Management (2008) 15, 153–168. doi:10.1057/dbm.2008.13; published online 15 September 2008 Harsha Aeron FPM-33, Indian Institute of Management Off Sitapur Road Lucknow-226013, India e-mail: harsha.aeron@gmail. com INTRODUCTION Credit cards are replacing currency in many emerging markets and are also nearing saturation in developed economies such as the US. By providing a revolving credit facility, credit cards empower customers © 2008 Palgrave Macmillan 1741-2439 Vol. 15, 3, 153–168 Database Marketing & Customer Strategy Management www.palgrave-journals.com/dbm 153 Aeron et al. to manage their cash requirements with convenience for a fee. As the customers demanding credit are increasing, so are the firms that are ready to satisfy this demand, resulting in tough competition. There are many card-issuing banks and nonbank companies in the market. More than 90 per cent of the market-share is with less than 1 per cent of these units.1 The prevalence of tough competition in the industry and the relatively high costs of acquisition as compared to retention compel card issuers to be customer-centric and make the right decisions for the right customer at the right time. These decisions require customer information and predictions regarding the value of the customer from the card issuer’s perspective. Customer lifetime value (CLV) is a metric that indicates the value of the customer. A credit card firm has to take various decisions throughout the lifetime of its relationship with the customer, that is, from acquisition to attrition or default. The set of decisions starts with deciding which customers to acquire. As acquisition involves cost and there is a fixed budget assigned for it, a firm aims to select customers with high profit potential and low risk. Owing to the revolving nature of a credit card product, however, the relationship between profit and risk in a card is more complex than in a closed-end loan. Similar to other financial organisations, the card issuer bank would prefer customers to pay back the amount that they have borrowed using their card. It may not, however, want customers to pay back the entire amount in the first cycle itself. It may, instead, prefer them to ‘revolve’ and generate revenue for the bank. During the lifetime, the card issuer company needs to decide on the credit limit and price for each customer. At the retention stage, again, a firm has to decide whom to retain and how many resources to allocate for retention. These decisions can be guided by the CLV of the customer. Estimating the CLV of a credit card customer can help a card issuer bank in 154 taking the aforesaid decisions. First, at the time of acquisition, customers with high CLV scores can be given priority and accordingly the channel to acquire can be decided, that is a costly channel for highworth individuals and a cheaper channel for prospects with low CLV scores. Similarly, estimating CLV can help in taking decisions at the retention stage. The firm may aim to retain customers with high CLV scores and can accordingly decide on the cost of retention efforts. Researchers have recommended CLV as a metric for selecting customers, designing marketing programs and taking informed decisions in a structured framework.2–4 Customers selected on the basis of the CLV metric are more profitable as compared to those customers selected on the basis of other widely used CRM metrics such as previous-period customer revenue, past customer value, customer lifetime duration, etc.5 In this paper, we discuss the revenue model of a credit card and use this to propose a conceptual model that captures the CLV of a credit card customer. We use this conceptual model to build the CLV estimation process. We simulate different states (...truncated)