As banks, telecommunications companies, and cable television operators expand the role of their inbound call centers to include revenue generation,¹ determining how many customer service agents to employ will become more challenging. After all, selling products takes time, and adding agents is the only way to avoid the long waits that customers detest. But at what point do returns diminish? How should companies weigh investments in call centers versus, say, direct mail and outbound telemarketing?

A new application of a familiar tool—regression analysis—can help. By assessing the statistical relationship between service levels (such as the percentage of calls answered within 60 seconds) and top-line indicators (such as revenue per successful call, close rates per call, and abandon rates for customers in the queue), a company can calculate the likely return from investing in additional agents. With this information in hand, it can make straightforward comparisons of competing marketing opportunities. Counterintuitive conclusions often emerge. One telecom company, for example, hired more sales agents for its highest-performing call center when it learned that incremental service improvements there would lift close rates and sales per successful call more than would larger improvements in ordinary call centers.