The Happy

I heard some upper management types talking about the call center at work today. Now, I’ve never run a call center, but I have (like most techies, getting their start) worked on an ISP helpdesk in the past, providing call-in support to users with problems. My first tech gig, in fact. We grew very very good at walking people through things like windows networking settings, while playing Quake at the same time. So anyway, I have worked on the coal face, so I am not ignorant of call center stuff.

Now here’s the weird bit, the management guys in question were talking about stuff like maximum call times, abandonment rates, and other phone related shit. Which is all well and good, but these are .. really .. fucking stupid metrics for call center performance. The customer doesn’t give a shit about how many calls the techie they’re talking to has answered in an hour. If you start rewarding your call center staff based solely on how quickly they get off a call .. they will find ANY FUCKING EXCUSE to get off that call fast. And then you end up with calls like this:

“Hi, I can’t connect, I’ve double checked my password and username, but I just…”
“Sir, try rebooting. Then call me back. ThankYouForYourCall.” [CLICK]

…time passes ..

“Hi, me again. I rebooted, and I still can’t connect, I don’t think it’s any…”
“Okay, Sir, try reinstalling windows.”
“For serious? But windows runs just fine, I think it…”
“ThankYouForYourCall!” [CLICK]

..because hey, management has very clearly informed staff that what REALLY matters is call duration. Nothing else.

In my day we used to occasionally spend an hour or more on the phone with a customer, but when we were done, everything worked, and the customer understood what was happening. Repeat calls were rare enough to be comment-worthy when they happened.

Which is why if I ever need a support call center for this kinda stuff in any of my companies, I plan to measure performance in only one way – I’ll tell the staff “Stay on the line for however long you want. Hell, have them bring the PC in, if you can’t work it out over the phone. But fix the problem. While you’re there, ask if there’s any other problems you can fix. Video drivers. Printer drivers. Desktop background colour not quite what the customer wants. Whatever.” and I’ll have an auditing team (or person) whose job it will be to track recent tickets for each call center member, and randomly phone up customers that have called in, and ask “Did we help you? ARE YOU HAPPY?”.

And THAT will be my metric for call center performance. Because that’s what matters.


  1. Upper management don’t understand much beyond easy to understand at a glance graphs with an upward trend – and the trend is them getting more bang for their buck – give them a pretty picture which looks good and they’ll be pleased….

    I may be exaggerating, but that’s what our board seems to like..

  2. I think call time is a reasonable enough metric to be concerned with if you’re under-staffed. Customers having to call back several times is better than customers being left listening to the Girl from Ipanema until they give up and take their business elsewhere, because you’re spending too much time with the lucky person at the front of the queue. The obvious thing is to rate your performance based on the average amount of time customers are kept waiting, but the disadvantage of that from a management perspective is that it’s one number that applies to the whole team – you can’t use it to work out which individuals are your best and worst performers, because the call isn’t assigned to any one person until it’s picked up. Call time can be monitored individually, and is related, if a bit more indirectly, to the amount of time other customers are kept waiting.
    So if you need a number to measure how good your customer service is, and you need to measure individual staff members, it probably works.

    If you have enough staff to deal with the calls reasonably comfortably, you won’t have as much problem with people being kept waiting. That would be a good time to stop worrying about call times and start worrying about other aspects of customer service, like whether the customers have to call back and whether they’ve had a joyous and splendid experience. I suspect many companies never get very close to this state.

    Personally, I try to adjust my approach depending on the situation. If we’re busy and customers are queueing up, I try to solve the particular problem the customer has and move on. Other times, I try to stay on longer to solve as many other problems as I can see, and suggest other things they might like to have that they haven’t thought of.

  3. And then there are the companies who out-source their call centers to external companies, companies who are paid per call they take and who therefore have every incentive to get the fucking customer off the line as quickly as possible:

    1. Watch ad to get day pass thingy.
    2. Read.
    3. Weep, weep like a orphaned child.

  4. definately. i’ve talked to a few of the helpdesk guys out at Smales Farm EDS – and they’ve got monitors with calls waiting, waiting time etc… and the main goal is “answer all calls w/in 20secs” – i guess the unspoken, but understood clause here is “even if you haven’t helped the person out you’re CURRENTLY talking to”.

    its pretty sad really.