Multitasking in the call center is all the rage now. Over the last decade I’ve seen tools that automate simple work, can route any type of work to any agent and create schedules to match, that will identify idle time and slide in computer based training, that will blend inbound and outbound calls (not new I know), and of course we have had the ability to blend calls forever. The arguments are pretty convincing… better mix of work makes for more engaged employees, more efficient use of time and better service results.
As a WFMer, I am all about being more efficient in the contact center, but I sometimes wonder if we have gone too far.
For example, as one vendor was smoothly describing their routing and planning tools, complete with graphics showing all interaction types dive bombing smiling agents, I felt like I was being whisked through a grocery store in an empty cart-banging around in a mix of thrill and fear. But bumping that aside, I listened to their message; which was—we can schedule and route for multiple interactions and I noticed something missing—how do you hire for this environment?
When I say hire, I’m not concerned so much with the complexities and length of training for multiple channels; though that is a huge consideration. Or even a company’s ability to turn a LOL’er into an effective business communicator. I’m concerned with how you model this stuff in a hiring plan. And specifically how you would model effectively handling randomly arriving immediate response work like calls and chats with blended deferrable work like outbound calls, emails and forms.
My concern is occupancy and service level. I know I can use Erlang for my randomly arriving inbound transactions and basic production forecasting for outbound or deferrable ones. Heck I can even convert items per hour to handle time and calculate in the occupancy of my inbound work as an efficiency factor on my deferrable work. But I’ve never seen it actually work that way.
My real-life experience is that occupancy is lower on blended work as multiple systems need time to work together. And I’m not talking lag time, I’m talking the time it takes to queue up another outbound transaction, force suspend work or even for an agent to put their head around doing something different.
But most importantly, we know that Erlang tells us we need a certain occupancy in order to meet an inbound workload’s service level. Blending in work for the “time between calls” uses up that available time, so how can we hit Service Level efficiently?
That leads me to my long term planning conundrum… What occupancy should I use?
As I sit here, trying to write this article, I weave through a series of distractions, conference calls, texts, email alerts and the occasional drive by question. What should be an hour or so of work has turned into an all-day affair. Maybe I’m just getting old, but this blending thing doesn’t sound so good to me… or maybe it’s just the math I can’t wrap my head around…
If any of you are blending workloads with a high level of occupancy and meeting service goals, please let me know! Let’s talk.
