It's always been hard to find research on call center performance, much as it's been difficult to find broad-based research on the size and scope of the industry in general. Call centers tend to rely on benchmarking, which compares their performance to other centers in similar circumstances. And most of the time benchmarking is a comparison of specific callbased activities. In other words, if I am a bank, I try to find out how long callers are waiting on hold at my center, and compare that to hold times at other banks.
There's value to this information, but only to a point. My main problem is it really only tells you one thing: is the center performing as well as other centers? It doesn't tell you anything about two critical business aspects:
How well your center is doing compared with your own company's expectations and How well you are doing compared to your customers' expectations.
In the first case, against your company's expectations, benchmarking comes up against the fact that it's best at measuring callbased activity. That's what two call centers are going to have in common, frankly. It's easy to get two banks to talk about what their call handling times are, and compare them to each other. It's not so easy to get two banks to talk about customer value or revenue expectations, and whether their call centers are hitting the targets.
A lot depends on the tools used to come to your view of customer satisfaction. The better the tool, the more accurate your view will be. The most-used tool to measure customer sat is random monitoring. What's puzzling about that is random monitoring is primarily designed to assess the quality of a call from the agent's point of view. Did the agent ask the right questions? Did the agent follow procedure? In some sophisticated cases, you can even use monitoring to search out examples of calls where the customer was frustrated and try to decipher what went wrong, but really, monitoring is a QA tool — it was designed to assure consistency of service across a broad agent pool, and as a feedback mechanism to ensure certain policies are adhered to and skills acquired. It's not a good way to assess customer sat.
The second most popular tool managers use is phone or mail surveys. On the mail side, these suffer from delay, low participation and self-selection. When you get a survey in the mail asking you how the service the last time you called your bank or a rental car company, you're going to have to scratch your head and try to recall what the interaction was about, and whether it was either outstanding or crappy enough to merit remembering. And by phone; first you have to get the person to agree to stay on the line after their interaction with the agent has ended, and then you're counting on them to give you an honest — or at least unbiased — answer in what might be an emotionally charged situation.
Almost as widely used as these two tools is unsolicited feedback. In more than 40% of call centers, managers are relying on random, over the transom responses from a self-selected sample of customers to tell them how their call centers are doing.
People who provide unsolicited feedback are either very happy or very unhappy. No one ever wrote a letter to a company to praise or complain if the experience was soso. It just doesn't happen.
One way to get around this problem inside the center is to measure — and reward for — outcomes, not just for activity. That way your call center, your agents, even your managers, are motivated to achieve goals that are relevant to the rest of the company. Reward based on revenue achieved, sales opportunities converted, customer duration. Not calls handled, call duration, or even firstcall resolution.
And I urge you to create and experiment with different ways of capturing information about the broad spectrum of customer states from happy to unhappy and everything in between. Capture it from agents, formally and informally. Capture it from customers in every way you can — and encourage technology vendors to help you sort through it to find unseen patterns of customer preference and behavior.