The Customer Service Survey
VocaLabs' weblog providing news and commentary on the challenges of providing good customer service.
Help Me, Call Center Guru!
Monday - December 19, 2005 02:21 PM in
Call Center Manager: Help! My boss is asking me for statistics on how we are serving our callers, and I'm afraid she is going to base my performance bonus on the data. What should I do?
Call Center Guru: You've come to the right place. Lets start with the basics. I'm sure your system delivers a lot of telephone call statistics right? The first thing you need to do is pay me for some performance data from other call centers. We'll put some positive sounding words around this information like "Benchmarks" and "Best Practices" because that sounds so much better than "average".
Starting with basic measurements such as "call length" for example, you'll want to show how good a job you are doing at keeping each customer call as short as possible to control telephone expense.
Manager:Ok but what if our data shows that we are taking longer that these benchmarks to finish each call?
Guru:No problem. You have two choices. The easiest, is to point out how you are different that the competition and so the benchmark data doesn't apply. But this approach carries some risk, because once the benchmark data is called into question it will be harder to use in other areas. So a better solution is to alter how you measure. Lets say for example that a caller making a phone payment takes you 3.2 minutes to complete and the benchmark time is 3 minutes. Two tenths of a minute is only 12 seconds. Surely there is more than 12 seconds of ringing time or time on hold while the caller is waiting for an agent right? So just deduct an appropriate amount from each call to show you are outperforming the benchmark.
Manager:Ok, but isn't that kind of cheating?
Guru:Heck no. We don't know whether the companies reporting their statistics are including ring time, time on hold or even whether internal transfers are counted as one call or whether the clock starts over with each transfer, so you can feel justified in subtracting these elements from your data. And if those strategies don't work, you can also time the part of the call it takes to get customer info from your computers and blame any problems on the IT department.
Manager:Ok, I see how I can make the data look good on these types of call statistics, but what about other measurements like the number of calls handled by our IVR versus agents?
Guru: Piece of cake. All you have to do is make it hard for the caller to opt out to a live agent. If you do it right, customers will call back several times trying to get their business taken care of and the percent of calls that stayed inside the IVR will skyrocket. If you present it right as in "Look how we contained a higher percent of calls in the IVR while call volumes doubled!", you might even be able to finagle a raise.
Manager: Cool, but what if she asks about single call resolution?
Guru: We head this one off proactively. You show the boss how you instituted a customer satisfaction survey, and how you had the company in mind because you saved money by conducting a study that you administered yourself.
Manager:How do I do that?
Guru:You begin by asking each caller if they would like to participate in a survey at the end of their call. That way any frustrated caller that hangs up before getting their business done won't mess up the satisfaction scores.
Manager: Sounds good, but what about those that hang on and actually get through to the agent?
Guru: Simple. You give the agent a screen pop up alerting them that this customer has chosen to take the survey. That way, if the call is a problem, the agent will "forget" to transfer the caller to the survey. If you set it up right you can get satisfaction scores of almost 100%.
Manager: I see. But what if the boss wants to relate call center performance to company revenues?
Guru: One of the more enlightened bosses eh? Still no problem. Just make a question or two in the satisfaction survey relate to loyalty and purchase plans. Since you've filtered out the troublemakers already, you can deliver some impressive data on how customer care is adding to the revenues of the company.
Manager:Thanks, you've helped.
Guru: No problem, my consulting bill is in the mail.
Posted by Rick Rappe
Starting with basic measurements such as "call length" for example, you'll want to show how good a job you are doing at keeping each customer call as short as possible to control telephone expense.
Manager:Ok but what if our data shows that we are taking longer that these benchmarks to finish each call?
Guru:No problem. You have two choices. The easiest, is to point out how you are different that the competition and so the benchmark data doesn't apply. But this approach carries some risk, because once the benchmark data is called into question it will be harder to use in other areas. So a better solution is to alter how you measure. Lets say for example that a caller making a phone payment takes you 3.2 minutes to complete and the benchmark time is 3 minutes. Two tenths of a minute is only 12 seconds. Surely there is more than 12 seconds of ringing time or time on hold while the caller is waiting for an agent right? So just deduct an appropriate amount from each call to show you are outperforming the benchmark.
Manager:Ok, but isn't that kind of cheating?
Guru:Heck no. We don't know whether the companies reporting their statistics are including ring time, time on hold or even whether internal transfers are counted as one call or whether the clock starts over with each transfer, so you can feel justified in subtracting these elements from your data. And if those strategies don't work, you can also time the part of the call it takes to get customer info from your computers and blame any problems on the IT department.
Manager:Ok, I see how I can make the data look good on these types of call statistics, but what about other measurements like the number of calls handled by our IVR versus agents?
Guru: Piece of cake. All you have to do is make it hard for the caller to opt out to a live agent. If you do it right, customers will call back several times trying to get their business taken care of and the percent of calls that stayed inside the IVR will skyrocket. If you present it right as in "Look how we contained a higher percent of calls in the IVR while call volumes doubled!", you might even be able to finagle a raise.
Manager: Cool, but what if she asks about single call resolution?
Guru: We head this one off proactively. You show the boss how you instituted a customer satisfaction survey, and how you had the company in mind because you saved money by conducting a study that you administered yourself.
Manager:How do I do that?
Guru:You begin by asking each caller if they would like to participate in a survey at the end of their call. That way any frustrated caller that hangs up before getting their business done won't mess up the satisfaction scores.
Manager: Sounds good, but what about those that hang on and actually get through to the agent?
Guru: Simple. You give the agent a screen pop up alerting them that this customer has chosen to take the survey. That way, if the call is a problem, the agent will "forget" to transfer the caller to the survey. If you set it up right you can get satisfaction scores of almost 100%.
Manager: I see. But what if the boss wants to relate call center performance to company revenues?
Guru: One of the more enlightened bosses eh? Still no problem. Just make a question or two in the satisfaction survey relate to loyalty and purchase plans. Since you've filtered out the troublemakers already, you can deliver some impressive data on how customer care is adding to the revenues of the company.
Manager:Thanks, you've helped.
Guru: No problem, my consulting bill is in the mail.
Posted by Rick Rappe
Posted at 02:21 PM by | | | |