For too long, quality assurance in call centers has taken on a perversely Pavlovian approach. Supervisors review recordings of calls to listen for correct behaviors. For agents, correct behaviors yield rewards like recognition, more pay and eventually, promotions. When agents achieve the status of newly-minted managers, they train others to adopt the same behaviors that led to their rise in rank. And the cycle continues.

How useful are checklists of behaviors? Some of the actions supervisors hope to detect within a recording are easy to document, like how many times an agent says a customer's name, but may have no effect on the call's outcome. Perceptions of other behaviors, like conveying empathy, are subjective and tend to reflect a supervisor's, rather than a customer's, point of view.

What's more, the application of call monitoring to quality assurance often depends on a flawed premise to begin with. When you use a recording to evaluate an agent, you're presuming that the agent can control the outcome of the call. By this logic, a poor outcome, like an irate customer, is evidence of poor performance from the agent.

This assumption undermines the value of call monitoring to quality assurance. There's a lot you're not hearing if you only listen to a recording to find out what an agent is doing wrong.

Sometimes, the agent is the scapegoat for problems that originate with a company's products.

Larry Hennessey, director of call center technology with Data Collection Resources (Colchester, CT), recalls that was the case with one client, a manufacturer of camping stoves. "Someone would buy something, have a bad experience and pound the agent," he says. (The pounding, even if verbal, was still painful.) It was not until the company improved its products that agents stopped receiving angry calls.

Recordings reveal customers' perceptions of a company's products, services and policies. If you factor in customer surveys, recordings can serve as powerful gauges of how much an agent's communication skills affect customers' views of a company. Coupled with images of agents' screens, recordings of calls also show whether issues involving phones, computers or software contribute to problems with communicating with customers.

To have a positive effect on your center, call monitoring should be a tool to enable agents, and your company, to serve customers better. And, where possible, agents should have a say in whether monitoring, evaluations and training are helping them achieve this goal.