Call Center Agents Take Control

Conventional wisdom states that millions can be made and businesses built through cross selling. But it didn't quite work that way for Spaarbeleg, a Dutch financial services institution.

The firm, which has 1.6 million customers, tried to convert inbound calls into selling opportunities. But returns “were not keeping pace with the growth targets,” said Laura E. Squier, director of North American commercial solution engineers for software vendor SPSS.

There were two reasons this was happening. One was that customers were getting too many messages. That, in turn, led to a high refusal rate, resulting in “increased stress, burnout and higher call handle time” for call center agents, Squier said during a session last month at the National Center for Database Marketing conference in Orlando, FL.

“The agents were not excited about doing this,” she added.

A New Strategy

The company decided on a two-pronged approach, the first involving predictive modeling.

“We would predict the customers most likely to respond to cross selling, but leave it up to the agent to figure out whether they should make that offer or not,” Squier said.

When a customer calls in, the agent gets a recommendation in real time regarding what, if any, offer to make. Cross selling is encouraged only when it appears success is likely.

“We look at response probability, expected value and potential actions,” Squier said.

It's up to the agent to determine whether to proceed with a pitch, a decision based on any number of factors.

For example, “If the customer is angry, that's no time to make an offer,” she said. This strategy led to a vast reduction in agent stress.

And what was the bottom-line result?

Out of 1 million inbound calls received, 180,000 were flagged for cross selling. The agents made 60,000 offers, generating 30,000 quality leads and some 22,000 product sales.

Moreover, the program realized $30 million in additional sales as well as “a nice increase in profit,” according to Squier.