Rice University | Explore Rice

1/15/2007

The key to business has long been keeping customers satisfied, but a new, better way to profitability may be to nurture the right customers. It can cost anywhere from hundreds to thousands of dollars for businesses to identify and retain a single customer. But where should those dollars be spent, and on whom? Companies find these answers by using a metric called customer lifetime value (CLV). While existing models in the marketplace evaluate CLV, Rice University assistant professors Sharad Borle and Siddharth Singh have built one that is outperforming the others.

Which potential customers should a company go after? Which customers are worth going the proverbial extra mile for? With shrinking marketing dollars and increased competition, companies have been spending a lot of money trying to hit the right targets. According to Sharad Borle and Siddharth Singh, both assistant professors of management at Rice University’s Jesse H. Jones Graduate School of Management, if companies don’t know what each customer is worth to them today and in the future, they’re just shooting in the dark.

At the heart of any company’s initiative to attract customers, retain them, and increase their purchases is the ability to measure each customer’s lifetime value (CLV). “Our model outperforms existing ones in determining this metric,” Singh said. “It does a better job of predicting customer lifetime value, and it does a better job of targeting valuable customers. Our model can also tell businesses which characteristics are most valuable in terms of how much money the customer is expected to spend with them.”

Why the better success rate with their model? “We do a better job of looking at past patterns and data, sequences of purchases, amount of purchases, and the risk of the customer leaving the company,” Borle said. “We look at these outcomes jointly and what they imply for the future; this joint approach gives us better predictions.”

The idea behind Borle and Singh’s work was to create a user-friendly model that the industry could adopt and implement immediately to evaluate the cost of retaining relationships and make marketing decisions. The model’s framework was designed with a lot of flexibility, to incorporate each company’s desired values and variables. Borle and Singh estimate that their model could potentially achieve a savings of 10 to 20 percent on customer identification and retention programs.

“The information our model provides allows you to compare the value of one customer in relation to others,” Borle said. “It’s also very useful in guiding marketing and promotional dollars, targeting customers, figuring out whom to call on, and who should get more attention. After all, if you have a higher customer lifetime value than I, the business should go the extra step for you.”

The ability to target customers who have higher value potential has proven very useful to firms that want to promote new products, Singh added. “It costs a lot of money to send out promotions to everyone. So, by selecting the right customers, you save money and you’ve targeted the customers who are more likely to spend now and make purchases in the future.”

The context for Borle and Singh’s model is membership-based firms like movie clubs, book clubs, and retailers that know when a customer actually starts and stops shopping with them. Very few models address this context, and theirs was one of the earliest to do so. “These businesses spend a lot of money trying to figure out which customers to acquire,” Borle said. “Our model tells them the potential value of a particular prospect.”

The whole arena of CLV is relatively new, and many companies in the marketplace continue to use more simplistic measures like averages to determine customer value. Some sophisticated models are being used by businesses, especially in direct marketing. “But we have compared ours to these,” Singh said, “and ours is much better.”

For more information, contact Sharad Borle at sborle@rice.edu, Siddharth Singh at sssingh@rice.edu or Laura Hubbard of the Jesse H. Jones Graduate School of Management at lhubbard@rice.edu.