Researcher: Call Monitoring Needs FixesNEW YORK -- Call monitoring practices used in the inbound teleservices industry are due for some changes that could help contact center managers improve agent performance, a researcher affiliated with Purdue University said yesterday.
Call centers typically sample an average of five of 1,000 calls that contact agents handle during a month, then use those samples to score agent performance and determine pay and promotions, said Jon Anton, director of benchmark research at Purdue University's Center for Customer-Driven Quality. However, the practice rarely provides a true picture of agent performance and is more likely to stir agent resentment.
Anton spoke at a seminar here sponsored by customer contact auditing firm HyperQuality, Seattle, and the Society of Consumer Affairs Professionals. Since 1995, Anton has been the principal director of the Purdue University Call Center Benchmark Research survey, data from which is warehoused at BenchmarkPortal.com.
Call centers typically limit the number of samples they use to review an agent's performance because of cost issues, Anton said. But to obtain a statistically reliable result, call centers would need to sample more than 300 calls from 1,000, he said.
Instead of random sampling, a better solution for monitoring and evaluations would be to monitor calls that can be identified as potentially in need of improvement, Anton said. For example, calls that take longer, are on hold longer or have longer "dead air" time than normal can be flagged as in need of review and discussed with the agent.
Agents likely will balk at this practice if review of these samples, which could reflect negatively on them, is tied to performance reviews and pay, Anton said. A better option is to do away with this tie, he said.
Call centers typically grade agents based on a "scoring sheet" and check off whether they comply with rigid rules such as, "Use the customer's name three times in a conversation," Anton said. Those standards do little to help performance.
Instead, call centers can use other parameters, such as calls per shift and customer satisfaction ratings, to judge agent performance, he said. Using these standards, agents will view monitoring not as a punishment but as an opportunity to improve the quality of their calls and a chance to increase their pay.
Call center managers often struggle to obtain money from their organizations to support call monitoring, Anton said. Direct costs for monitoring range from $5 to $10 per scored call, and indirect costs can raise the total to $20 to $30 per call.
It's important to emphasize that call centers affect company profits and costs as well as market share, wallet share and customer satisfaction, Anton said. In a survey of 1,000 consumers conducted by Purdue in the past 12 months, 92 percent said their contact with a company's call center affected their image of the company.
"Sometimes we don't hang a big enough hat on what we do," Anton said. "And yet, when you think about it, 80 percent of next year's business comes from this year's customers."