ICMI is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.


How Busy Is Too Busy? Create a Realistic Occupancy Goal for Your Business

Are we keeping our agents too busy in an effort to maintain service level? Recently, I was asked to do a “what if” exercise, with a scenario that included a 65% service level and 100% occupancy inputs. In today’s economy, most of us have been asked to squeeze more “blood out of a turnip,” (and often this turnip has already been dehydrated to the point it now resembles recycled shoe leather). As a result, we – as leaders in the contact center – are always looking to creatively drive and increase efficiency, and create quick and delightful customer service, as well. To accomplish this, we fully realize we will be increasing pressure on both our “systems,” as well as on our people. But, do we really consider all of the factors in our equation?

Definitions of Occupancy

I have seen several derivations of the occupancy definition. Some only include talk time. Some include all activity but free time. The industry definition that ICMI uses is: The percentage of time agents handle contacts vs. the time they spend waiting for contacts to arrive. This definition goes on to explain that using full Average Handle Time (AHT), inclusive of hold and After Call Work (ACW), is imperative. I would add that in a world where agents are often expected to document calls, answer e-mails between calls, answer surveys about IVRs and in some cases write knowledge base articles, and we should include ALL work in our equation. (I know. I know. Here comes the argument about the difference between “utilization” and “occupancy.” All of the Ops folks - and I am one - will go into the world of numbers for the sake of numbers, have a great debate, and lose track of what we are solving for!)

In short, shouldn’t we actually be looking at “how much time to breathe” our agents have?

The Cornhusker Approach

Now, here is a Cornhusker’s definition. I recently visited my best childhood friend, now a preacher, in Iowa. While there, I listened to him share a sermon about how hard we worked as boys—working fields, bailing hay, shelling corn, doing “worse jobs,” and most times forgetting to wash our hands at lunch. Being the contact center guy that I am, as I listened to the sermon, I began putting equations and productivity metrics around this work. In particular, I thought of how we used to shell corn. This task involved two kids armed with huge pitchfork-type tools in a giant corncrib (or a bin of whole-ear dried corn) with a conveyor belt beneath the corncrib. The kids would start pulling the little boards that cover the alleyway over the conveyor belt out from under a 20-foot high pile of corn so that the corn would flow into the machine that would then separate the corn, passing the kernels into a wagon and the cobs into a pile to be burned or used for other purposes.

How does this relate to occupancy?

It has to do with the two kids working in the corncrib who, if the corn didn’t move swiftly enough through the sheller, were screamed at and even threatened. The kids, on the other hand, strategize. They decide to let the corn flow until they have a huge wall towering over them before taking the bottom of the pile out with the rakes to fill the conveyor. Once that’s done, they can sit and visit about things teenage boys do for as long as they can… until the yelling starts. The problem with this strategy is that if the wall of corn gets too big, the conveyor can’t keep up and the belt breaks. The moral of this story—just like with occupancy—is that a steady pace serves the customer better and creates a repeatable expected result (as opposed to regular intervals of being “too busy to keep up” followed by intervals of doing nothing.)

Creating a Balanced Workflow

With occupancy, our desired goal is to create a balanced workflow that is conducive to fantastic customer satisfaction. We hope we get a high Net Promoter as a residual. If agents are too busy, they will feel frustrated and results from this are less than desirable. Likewise, if agents are too slow, then we may have lag time. In either case, the result is an environment where realistic efficiency is compromised. In the contact center today, I would venture that most of the time, we are dealing with busier agents.

I have always felt that agents are the smartest employees in the contact center. As leaders, if we are seeing constant queues, we are usually also seeing longer hold times, higher ACW, maybe a few more escalation calls, etc. - our survival instincts kick in. I would challenge you to think about this differently than I know many leaders normally do. The agents want to do a great job. They hate making mistakes. They like helping our customers. They simply have to find a way to “catch their breath” in order to be able to deliver quality service. Factoring increased handle time into a staffing equation means the number of required agents goes up. I have found a direct correlation between occupancy and agent requirements: extreme occupancy will increase the number of required agents. The reason is a compound factor resulting from high occupancy and the impact that it creates.

High occupancy will drive less desirable residuals. First, we’ll see adherence suffer. Next, we’ll see folks starting their shifts a tiny bit later and leaving a tiny bit earlier. Then attendance suffers. We may see the much talked about Least Occupied Agent (LOA) syndrome take hold. We might hear conversations in the halls, know that our agents are looking and then feel the blow of attrition. This quickly compounds as we lose “tribal” knowledge when we begin to populate the contact center floor with new hires, see handle time rise, see occupancy rise even more, see service level deteriorate, and finally get to a point that we wish the circle was only vicious.

One note to consider here: when measuring occupancy in a one-to-many environment (i.e., chat, Twitter, Facebook, online communities) we should consider concurrency. If we use our definition above chances are, by design, we are planning (under the “old” definition) 100% occupancy. For example, if you plan agent staffing at say a 3-concurrent contact ratio, will the agent ever really be “unoccupied” in the traditional sense? How much is too much? Try this exercise: Ask someone to ping you on your internal message system every 2 minutes. Ask someone else to e-mail you every 3 minutes. Ask someone else to send you a PM in Twitter every 7 minutes. Finally, ask someone to post an inquiry to Facebook. Now set a timer for 30 minutes, tell everyone you asked to help with your experiment “Go!” and start typing as fast as you can while trying to ensure that everything is answered in what you deem a timely manner. How did you do? I know most of our chat and social media agents are much better at this than we are. (They also thrive on and enjoy this type of challenge!) My suggestion here is to set a lower concurrency ratio – to set this ratio by interval - and to find creative ways to break up these agents’ days. But, ultimately this needs to be realistic to your center, your people and your business.

Create a Realistic Occupancy Goal

So, what do we do?

1. Involve the team. As you design your goal, ensure that your agents – and their leaders – are involved on from the “ground floor.” Use their input to come up with a reasonable starting point.

2. Look at interval occupancy. There may be realistic intervals (i.e., lunch times, afternoon peaks) where occupancy will necessarily have to be compromised. From there, design a scheduling strategy that emphasizes “all hands on deck” at the busiest parts of the day and allocates off-line activities and other duties to non-peak periods. (Note: this may require a mindset change, as it isn’t “how we’ve always done things.”)

3. Don’t build your goal in a silo. Remember the relationship between occupancy and service level. These metrics are intertwined in such a manner (I won’t get into a math lecture here) that setting independent goals for each without considering the whole picture could be like throwing eggs into a cake bowl without ever bothering to break the shells.

4. Ensure leadership has line of sight. Share your ROI and the benefits of a reasonable occupancy model (increased CSAT, less attrition, higher FCR etc.) with the organization. And don’t forget to quantify the cost savings!

5. Don’t be afraid to adjust for reality.

6. Communicate and test for understanding-repeatedly, often and over and over again.

Finally, as you begin to work with the entire team to understand the dynamics of service level, occupancy and most important, how they relate to customer satisfaction, remember that one size is not going to fit all. Different contact types mean different stress levels for agents, and each agent’s threshold of “How Much is Too Much” varies. Ask your people leaders what a calibrated that agent’s needs are in terms of breathing space. Ask your bean counters the cost of LOA, attrition and absence. Ask your agents if they know that we actually don’t plan (or, we shouldn’t be) for them to take back-to-back calls. Finally, ask yourself if you have the right feel and the right inputs to confidently say you have a “good” occupancy goal that can be measured. And if you can’t answer that question easily, then you need to ask yourself “Why?”