Why the “Old” Way of Workforce Management isn’t Good Enough

blog feature image
Published:  February 24, 2015

Having been in this business over 40 years and in over 1,000 call centers in five continents, I’ve seen some major changes in workforce management.

For example, back in the early days, we didn’t have ACDs – we had PBX “hunt groups.” Calls were simply given to the next available agent. There were no queues, no service levels, and the only metrics tracked were how many calls an agent handled during the day and occasionally how long they talked.

“Erlang C” wasn’t invented until around 1975. Instead, we took “Erlang B” tables, which were engineering tables used to calculate how many trunks you needed in the center, and then modified them to Erlang C.

If you could somehow figure out about how many calls you would have during the day, how long they would last, and what you thought your service levels would be, you could then predict how many agents you needed on the floor at any given time. This was all done on paper and with a calculator.

Schedules were also done by hand and intraday management was either nonexistent or at best, minimal. Workforce managers would try to figure out if agents needed to take shorter or longer lunches and breaks and would decide if they should send agents home, but everything was primarily based on instinct, not data.

Sophisticated Situations, Archaic Solutions

Then in came technology and centers moved from PBXs to ACDs, which were more sophisticated and allowed the creation and management of different channels and queues. This enabled companies to have multiple queues feeding into one center or even multiple queues feeding into multiple centers. (In 2001, I had one contact center with 83 queues. Agents were trained on 20, 30, or even all 83, and all were prioritized.)

Along with the multiple centers, channels and queues came more advanced forecasting and scheduling systems. Some of these are still around and there are many variations depending on the complexity of the business.

With all of this sophisticated real-time data coming in, dashboards were created so that people can make business decisions based on the data. Centers began monitoring metrics like service level, speed of answer, average handle time and revenue per call while at the same time, coaches, trainers and quality teams are competing for agents’ time and scheduling and forecasting is being done in silos.

With all of these sophisticated contact centers scenarios – multi-channel, multi-site and multi-queue – agent development continues to take a backseat. Most centers I see today have inefficient agent development programs in place or they don’t do it at all.

How can centers today analyze their data and convert it into action that will actually improve the customer experience and the bottom line?

Out of Control

As a consultant, most contact centers call me because they need some help.

Consider this example: A center I recently evaluated has been experiencing high attrition rates. They offer voluntary time off in the mornings and afternoons because their peak call volume is between 7:00 PM and 9:00 PM.

The center is measured on service level and its levels in the morning – when volume is slow and agents are sitting idle, waiting for the next call – is around 97%. Supervisors try sending agents home and offer long breaks and lunches, but when agents don’t get a full 40-hour week, they are more likely to leave. At the same time, agents who are scheduled to come in at 12:00 PM are often asked go stay until 10:00 PM to cover peak volumes in the evenings.

The result? Service levels get blown out of the water. The center experiences eight hours of 90% service level and two hours of 20% service level, which gives them an average service level of around 60%. No one’s happy. (Customer satisfaction, by the way, has no correlation to service level. It depends on who you call and when they called. You could call a customer who says they got right through, and then call another who had to wait two hours. Those two things don’t match up if you don’t have consistent service levels every half hour.)

In the center I was evaluating, direct expense was higher than revenue. And this doesn’t even include indirect expenses like the lack of agent training and development because there is no time in the afternoons when everyone is answering calls and when volume is low, agents are sent home. This center has three people in workforce management who spend all of their time just entering exceptions to the schedule. Adherence hovers around 60%.

To further complicate things, supervisors have the ability to move agents from one queue to another to cover spikes in volume, but not all of the agents moved are equipped to handle interactions in multiple queues. And when an agent is moved from one queue to another, the schedule created for that agent is no longer accurate. And, when the agent’s shift is over, he or she must then be moved back to his original queue to start the next workday.

Game Changer

I wish I could say this particular center is an outlier, but these circumstances are common in most centers I see.

Forecasting is done in silos and then “thrown over the wall” to the center. Scheduling is just as haphazard and intraday teams are faced with KPIs they cannot possibly meet.

The truth is that in most centers, it requires people with expertise to make all of this work and some are better at it than others. And this is just one center with 250 agents – multiply that out to 2,000 agents in multiple centers and it would require as many as 10 people dedicated to intraday management to maintain control.

The next “big thing” in the contact center will be technology like Intraday Automation that integrates all of the disparate technology within the contact center and enables effective decisions to be made automatically.

Today’s contact centers need to take all of this data, analyze it, and then use it to make good decisions based on business rules. Data is coming in faster and faster – you just can’t keep doing it like you were doing it in the 1970s.

Categories

Archive