Agent Producitivity In An Imperfect Scenario

April 26, 2010 9:22 AM

I need to measure my employee's performance however, due to system limitations, I cannot track their not ready time. What is the best method for measuring employee productivity in a call center that has vastly different call types with no good way to actually measure offline time?

I'm looking for a way to effectively measure employee productivity in a quite imperfect scenario. I have numerous system limitations with using a PC based attendant console (Amcom) and a Meridian 2250 Console.

I manage a switchboard for a health care facility. We are 24/7/365. Our calls during the day M-F are mostly answer and transfer but we do have quite a few calls that require operator assistance (dr-dr transfer, physician paging, etc). In the evening it's mostly answering service work as well as intake for the Dial A Nurse program. In addition to filling out templates for physician paging, the operators have to take triage information from patients that are calling for the nurses. Some of these calls are lengthy as they have to locate an available nurse for emergency calls.

I can't pull anything useful from my current software package but I can pull number of calls. Thats about it. And it's not completely accurate as it's only what hits our answering software, not what comes in in through our T1s. I can't measure wait time because my software only records when it actually begins ringing on an operator console. I am unable to accurately measure not ready time as the console was not deisgned with agent monitoring in mind. There are too many ways to trigger an active call state without actually being on a call. And if an agent happens to utilize the physical console for anything, it's not recorded in my software.

My idea was to take my slowest and fastest and run some time trials and then average the two times together on the different types of calls. I would also take those two and average their number of calls per shift to come up with a range for number of calls taken.

Is there another way to measure productivity that will not cost me anything nor take months to implement?

posted by bodgy to Work & Money (5 answers total)

How many people are on the switchboard? My first impression would be to sit in on a few calls with each agent and give them a one-on-one feedback session. I worked in a call-center and despite the tracking and metrics data, this was the one area that improved my performance the most.
posted by Hiker at 9:30 AM on April 26, 2010

Response by poster: I have 35 people with no more than 12 at any one time. We've been doing side by sides however it is time consuming when you are looking at the shift options plus I have a few people who work one day every 2 weeks.
posted by bodgy at 9:55 AM on April 26, 2010

Be very careful here. Time is an important part of the equation, sure, but you are going to get what you measure and loose what you don't measure. If I'm going to be graded on speed, then screw accuracy and what not. This is absolutely not what you want. Here's an article you might want to do a quick read through of.

In your shoes, I'd look at number of calls per shift over the long haul relative to calls on similar shifts. But I'd also talk to doctors and nurses who receive the output of their work. If you get a long series comments about it not being accurate, you have a problem. (Be careful here too because an emergency caller might not be terribly precise or clear in the first place and nothing is going to sap morale quite like chewing people out for faithfully recording the ramblings of someone who is on the verve of hysterics.)
posted by Kid Charlemagne at 10:44 AM on April 26, 2010

You need to start with the basic question: What is your goal? Why are you measuring?

Measurement costs money, so either the benefit of improvement or the pain of not improving needs to be worth the investment.

If you can get a basis of estimate on the potential for improvement, maybe you can justify some well-targeted software improvements. Or maybe you can justify doing some periodic sampling using manual checksheets or observation? Or maybe you can just make the best out of the data you have on hand (calls and call time per employee clock hour?). Or maybe just save yourself some trouble and not measure at all. Lots of money is wasted on measurement for measurement's sake.

Make sure you can quantify the pain/opportunity before you start measuring. And when you do start measuring, use Statistical Process Control to avoid causing yourself more trouble (e.g. Process Tampering).

If your organization has access to a Lean or Six Sigma person (Black Belt), this would be right up their alley.
posted by cross_impact at 12:04 PM on April 26, 2010

Perhaps your best bet is to hire a productivity specialist who interviews and observes your employees and provides you with a report? Something like a corporate anthropologist or a business consultant? Are you looking to measure individual employee productivity (as part of their performance review, for example)? Or are you trying to measure your productivity as a department as a whole? An independent consultant, especially one who makes observations of employees on a more anonymous basis (i.e. not naming them particularly in their report) could be a helpful way to make productivity suggestions for your organization, if you're not worried about individual performance (at least at this time). By being able to interview and discuss with your employees on a more "anonymous" basis (maybe take them for coffee off-site for a half-hour or so interview, surveys, etc. and write up an ethnographic-style report without naming informants directly), you may have some interesting insight from the employees without them getting their backs up or feeling that their jobs are threatened. That's of course only if you're looking at improving productivity as a whole and not individual performance evaluations.
posted by 1000monkeys at 7:04 PM on April 26, 2010