The increasing complexity of today's contact center isn't only changing the way that businesses interact with customers, it's also redefining the way that the centers themselves measure success.
Ten years ago, who would have thought e-mail or chat would be such important service tools? Who would have guessed that the government would impose marketing regulations forcing companies to make the most of incoming customer inquiries to market new products and services?
All this means businesses need different ways to measure whether a contact center is succeeding, experts say.
"In the olden days, we were measuring quantity," said Connie Smith, a 20-year industry veteran who has managed several 24x7 contact centers. Today, a movement is afoot for centers to become more "quality-driven," according to Smith, who now serves as chief evangelist for Envision Telephony Inc., a Seattle-based work force and quality management software vendor.
For example, many centers find it useful to measure average call value, which is calculated by dividing the revenue the center generates by the number of calls it answers. Others gauge customer satisfaction by listening to a sampling of recorded calls, surveying customers at random or having the agent end each call by soliciting feedback.
"Ask the customer, 'how did this call go for you?'" Smith recommended.
She also advocates measuring the average cost per call by factoring in technology investments, telephone costs and agents' salaries.
These additional metrics, however, aren't a substitute for tracking service levels. Call centers still must understand their traffic so they can properly manage workloads to meet customer demand.
That means they should still figure out the number of inquiries that come in, the number of calls they respond to and the number of calls that are abandoned. They should also measure a customer's average hold time and the length of the average call. Most automated call distributor systems generate reports that track these statistics.
Long calls might mean agents need more training or that the technology is sluggish, Smith said.
She also recommended tracking first-call resolution rates, which give a center an understanding of how quickly they're solving customers' problems.
Metrics in action
Borders Group Inc., the Ann Arbor, Mich.-based bookseller, uses a combination of quality- , productivity- and service-based metrics to determine if its agents are making the grade. Its contact center handles about 1.4 million inquiries each year, fielding calls and e-mails concerning book orders, store questions and loyalty programs.
"We look at the ability to balance quality and productivity -- and both have equal weight," said Charlie Moore, Borders' director of customer service.
Borders evaluates its 40 agents by randomly monitoring about a half dozen of their calls every month. Supervisors meet with agents monthly to evaluate the quality of the interactions. Borders also tracks service levels, such as the number of contacts an agent makes per hour. When agents sign on to the system, they can compare their stats with the center's averages and can access updated performance information using dashboards on their screens.
Adherence to schedules accounts for a smaller part of the agents' reviews.
In the last 15 months, incoming e-mail volume has shot up from 15% to 40% of Borders' contact center volume. As a result, agents are also evaluated on their ability to handle e-mail effectively.
Smith said many companies make the mistake of tracking calls, but not gauging how well they're handling other communication channels, such as e-mail or Web chats. They should shoot to hit common service levels for each channel, she said. (see sidebar)
Training is key
Experts agree that success begins with good agent training and continues with ongoing feedback.
For instance, Borders recently took its agents off site to brush up on their skills.
"The way they were interacting with customers wasn't generating loyalty," Moore said. Before the training class, Borders recorded samples of its ideal customer communications and then shared them with agents.
Becky Carroll, senior consultant with Norwalk, Conn.-based consultancy Peppers and Rogers Group, said companies should think hard about the metrics they set. They then need to align agent training with those goals.
"[Agents] are going to do what they get paid to do," Carroll said at the recent Smart CRM conference. "If they're paid to get off the phone quickly, they're going to get off the phone quickly."
That said, if a company is trying to gather more data on its customers, it shouldn't measure agents on the sales they make or the speed with which they dispatch calls, she said. Instead, it should judge them on the quality of the information that they enter into the system.
Envision's Smith added that if a business wants to up-sell or cross-sell its customers, calls will generally run longer. That detracts from the value of a metric like average call time.
Experts generally agree that contact centers track far too many metrics and should really concern themselves with a handful of performance indicators that they deem important.
"I like to say, 'don't measure too much and only measure what counts,'" Smith said.
At Border's, Moore's team gave its center six months to ramp up on new performance goals. During that time, he asked supervisors and agents to comment on the metrics and then he tweaked some levels based on feedback.
After that, he said it's important to adhere to the metrics set.
"Stick to it," Moore said. "We've had times when we haven't hit our service level and missed it by 1%. As hard as it seems, [if that happens] you don't get the goal."
Smith said the blame for falling short on metrics doesn't only reside with the contact center agents and supervisors. Occasionally, the IT department could be bear the blame for a slow server, or marketing could be responsible for launching a campaign that impacts call volumes.