Nationwide CSRs Fired: Hang Up on 50% of Customers

The Des Moines Register reported on Monday that four employees of Nationwide Insurance's Des Moines call center were fired for hanging up on customers. At least one of the fired CSRs stated that she was trying to keep her stats up:

Gillum testified that she hung up on policyholders to boost her job-performance statistics."I didn't think about the fact that it could be someone that was needing help right then and there, that their daughter may have just got in an accident and was in the hospital and they were needing help," she testified. "I wasn't really thinking about the customer. I was thinking about myself and my stats."

The fired employee went on to claim that hanging up on customers to keep up their call metrics is a common occurrence in the call center.

While this may simply be an angry ex-employee blasting her former employer, the reality is that call center managers and supervisors often use metrics and phone stats (e.g. Talk Time, Average Call Time, Average Handle Time, Abandon Rate, Average Hold Time) to hold front-line CSRs accountable and confuse it with "quality". They do so because the statistics are easily retrieved from the phone system and it requires little effort to look at the number.

I remember talking to a local call center supervisor who told me that her company had just "spent a lot of money on a 'Quality Program'." When I asked her to tell me about the program she replied, "It's great. We get a report telling us how many calls the CSRs are taking, how long they spend on the phone, how often they put the customer on hold. That kind of thing."

The problem is that a statistic like  Average Handle Time is not a measure of service quality.  Holding CSRs accountable to a number will result in situations like the one alleged at Nationwide, where reps focus on their "numbers" instead of the customer. The only way to measure quality is to measure what is actually happening and what the customer is experiencing within the call. That requires listening and objectively measuring the content of the call.

Comments

[David B]

Handle time is a reasonable measure of quality. Of course not in the way you describe it being used. Looking at an individual interaction, the handle time should be appropriate to the situation. There's nothing worse that having a 200 second issue take 800 seconds. Except having an 800 second issue take 200 because the CSR hosed and closed.

[Tom Vander Well]

Thanks for helping me make my point, David. Handle time can be a reasonable indicator of quality, but ONLY when taken in context. The root problem of a "200 second issue" that takes 800 seconds lies within the phone call. Why did it take so long? Did the CSR chat too much? Was there a system issue? Etc.

The problem lies with those who would look at just the number and make a value judgment based on that sole piece of information (e.g. "Your call was 800 seconds. That's poor quality"). You don't know that for sure without listening to the call, qualifying that it should of only take 200 seconds and discovering the reason why it took so long. I can tell you that such leaps to conclusion are quite frequent in the industry, unfortunately.

A caviat would be looking at a CSRs AHT over a significant number of calls and comparing it to a call center average. Again, that can be an indicator of problems, but you won't discover why it's too high until you listen to the calls.

[Rich Mewes]

I guess it also depends on how you quantify things. "Issues" may take some quantifiable/measurable average amount of time, but aren't we really trying to fix "people?" (who have problems of one sort or another)

"People" have problems that we can help them solve. The relationship or rapport that a good CSR builds with clients is something you can only quantify with return business, word of mouth evangelism or as you suggest, direct observation.

Done well, you can get both (minimizing the time it takes, without minimizing the person)

Done well, this (to me at least) is what it's all about.

Cheers,

Rich