Essentials of Contact Center Staffing: Calculations and Staffing Models for Workforce Planning
By Penny Reynolds
Workforce management plays a critical role in every contact center. Successful call center operations depends on getting the right number of people for each period of the day to answer incoming contacts while accomplishing two objectives – meeting desirable service levels and minimizing personnel costs.
In this article, you will learn about how important it is to get the right number of staff in place and what the implications can be for overstaffing and understaffing. You will also hear about the unique characteristics of inbound call center staffing and how it impacts staffing requirements. You will also learn to identify various traffic engineering and staffing models that apply to inbound call centers, as well as how to modify these calculations for other types of contacts like outbound calls, emails, and chat contacts.
The objective of workforce management is to get the exact right number of people every period of the day. If the numbers are off – either too many or too few – there are serious implications. When considering the impact of overstaffing or understaffing, it is important to consider the impact that staffing levels have on the three main stakeholder groups in the center – customers, frontline staff, and senior management. Let’s take a look at how overstaffing or understaffing affect each of these groups.
Impact on Callers
The most obvious set of people impacted by improper staffing is the customer base. Customers placing calls to the center are directly impacted by the number of frontline staff in place. If there are more staff than actually required, that can actually have a positive impact on callers, as their calls will get answered faster with shorter delays. However, if customers call during overstaffed times and experience shorter than normal delays, it might set an unrealistic expectation of service for future calls.
Understaffing has a much more negative impact for callers. Fewer frontline staff mean longer delays for callers. Also, as queues get longer and frontline staff feel the pressure of many calls waiting, the quality of service may also be affected. If delays become too long, callers may give up and abandon the queue.
Think about the wait in queue from the customer’s perspective. It helps to compare a caller’s wait in a telephone queue to a wait in an in-person queue.
When waiting in line at the grocery store or almost any other type of queue, the following are true:
- The customer can see the queue and perhaps make a decision about whether it’s worth waiting or not.
- There may be options where the customer can select which queue to enter.
- Once in line, the customer can see progress moving forward.
- Finally, based on the rate of progress and the number of people in front of him, the customer can make an estimate of how much longer the wait will be.
In this in-person queue, frustration levels may be high upon entering the queue, but then goes down as progress happens.
Contrast the in-person queue to waiting in a telephone queue. While some call centers provide announcements about estimated wait time, most call centers do not. A customer simply receives an announcement to “please hold for the next available agent” and typically hears music or perhaps additional announcements. The queue is an invisible queue where the caller cannot tell what is happening. There are limited choices upon entering the queue and no signs of progress along the way, so the caller has no idea how much longer the wait will be.
While frustration levels may be low upon entering the queue, the longer the wait time, the higher a caller’s frustration level. If there is a very long queue, the caller may be very frustrated by the time he reaches a live agent. Therefore, it is critical to staff to make sure these long queue times don’t happen.
Impact on Frontline Staff
Frontline staff are also impacted by the number of staff in place. If too many staff are scheduled, it means more idle time between calls for all staff. While this might be desirable at some level, it does impact overall productivity numbers and if there are far too many staff, too much idle time can cause boredom.
On the other hand, understaffing has a more negative impact. Too few staff in place means getting call after call with not much idle time in between the calls. If these low staffing levels continue over several hours, this higher level of workload can be very stressful. The stress of the increased pace of calls can be worsened by the fact that callers have been waiting in queue longer and may be frustrated as the call begins.
Impact on the Bottom Line
The final stakeholder group to be considered is senior management. This group is looking out for the bottom-line financial picture of the center. For this group, overstaffing can mean spending needless dollars on staffing and lower productivity than expected.
On the other hand, understaffing can mean lower personnel costs, but have the negative impact of unhappy callers and staff. And from a financial perspective, longer delays in queue mean higher telephone bills and the potential for lost revenue if callers abandon due to long wait times.
It is clear that every center should take the proper steps to get the right number of staff in place to protect the best interests of customers, frontline staff, and senior management. The first step in this staffing process is to calculate the call workload for which to staff.
Workforce management is about getting the right number of people in place – not just overall – but for each period of the day. Therefore, workload is calculated for each interval of the day so staffing levels can be increased or decreased throughout the day to get the just right number in place.
Let’s take a look at how workload and staffing are calculated on an interval basis.
Workload is defined as the call volume multiplied by the average handle time or AHT. Workload is typically expressed as the hours of work in a one-hour period.
Since workload is expressed as hours of work to do, if AHT is listed in minutes, divide by 60 to get hours or if AHT is in seconds, divide by 3600 to get hours of workload.
Here are two examples:
In Example 1, assume 400 calls per hour with an AHT of 180 seconds.
Multiply 400 by 180 and then divide by 3600 (the number of seconds in an hour) to arrive at a workload of 20 hours or erlangs.
In Example 2, assume 300 calls per half-hour and an AHT of 2.5 minutes
Multiply 300 calls by 2.5 minutes and then divide by 30 (the number of minutes in a half-hour) to arrive at a workload of 25 hours or erlangs.
Once workload has been calculated, it’s time to determine the number of staff required. The number of staff required depends on the type of work and arrival rate or pattern of the work.
A different number of staff would be required to handle 20 hours of paperwork in one hour versus 20 hours of incoming call center work. With 20 hours of paperwork, the work is stacked up and each task can be completed in a back-to-back fashion. With sequential workload like this, 20 hours of work can be accomplished with 20 staff, since each person can do back-to-back tasks with no idle time in between.
On the other hand, with incoming calls, the work arrives randomly. While there may be 20 calls at some point in the hour, there may be more than 20 or fewer than 20 calls at another point in the hour. At points with fewer than 20 calls, there are some staff in the idle state just waiting on the next incoming call. Therefore, this person cannot handle a full hour’s work, simply because of the way the workload arrives. That is why an important rule of call center staffing is that the staff hours will always be greater than workload hours. This is due to the random nature of the work within the hour.
It has been established that for 20 hours of work, more than 20 staff are needed. How many more than 20 will depend on how fast the calls should be answered.
Staffing requirements are based on speed of answer goals.
Speed of Answer Considerations
There are two basic ways to define speed of answer or delay time in a call center.
These definitions are service level and average speed of answer (ASA).
The most common way to define speed of answer in a call center is by service level, with the level denoting a percentage of calls to be handled within a defined number of seconds. It is typically stated as x% of calls handled in y seconds or less, with a common service level goal in the call center industry being to answer 80% of calls in 20 seconds or less. Service level is typically reported as either the cumulative average for the day, or as a weighted average number based on the percentage distribution of calls. It could also be reported as what percent of the half-hours of the day the service goal was met.
Another common way to describe queue time or delay time is average speed of answer or ASA. This statistic represents the average delay of all calls for the period, including those calls that experience no queue at all. For example, if half the calls go into queue and wait an average of 60 seconds, and the other half go to an agent immediately and wait 0 seconds, the ASA would be 30 seconds.
Defining a speed of answer or service objective is an important part of the workforce management process as it has a direct influence on how many staff will be needed each half-hour. There is absolutely no such thing as an “industry standard” for speed of service. Each call center’s service goal should be based on many different factors, including the following:
Corporate Objectives. Part of setting strategic goals for the call center will be a speed of service goal. For an organization known for speedy service, the service goal may be ambitious while other centers are willing to live with longer delays.
Caller Captivity. Part of setting a service goal will depend on what other alternatives are available for a customer. In a competitive environment where the caller has other options, it may be critical to staff for shorter delays so callers don’t go elsewhere.
Customer Perceptions. To a large degree, the setting of speed of answer goals, should be based on customer needs and expectations. Customers’ expectations are today being based on a myriad of service experiences, and it is important to consider these in defining a service goal to meet customer expectations. Customers should be surveyed regularly to see what their service expectations are in terms of both quality and speed of service. It may be appropriate to have faster speed of answer goals for some customers than for others rather than having one goal that applies to all calls.
Competitive Benchmarking. Many call centers also benchmark against what similar companies are doing and how quickly they are responding to customer contacts.
Once workload has been calculated and service goals defined, staffing requirements can be calculated. In the example with 400 calls and a 180-second AHT, there is workload of 20 erlangs and the speed of service is an ASA of 30 seconds.
The next table shows what happens with varying number of staff in place to handle 20 erlangs of workload.
This table outlines the service to be expected with varying numbers of staff handling 20 hours of workload. This table uses an Erlang C model to arrive at the staffing and service numbers.
Call center staffing involves the use of detailed mathematical models that replicate the unique staffing issues of the call center. There are several mathematical models that are used in telephone traffic engineering applications. Some of these are particularly suited to the unique operational aspects of a call center. The primary traffic models associated with call center operations are:
- Erlang C
- Equivalent Random Theory
These three models represent traffic engineering models that are used in a variety of telephone-related engineering situations. Most of these use an Erlang traffic model as a foundation. A. K. Erlang, a Danish mathematician, developed these techniques. These mathematical models are based on the arrival rates and usage patterns in telephone calling and there are variations of the Erlang formulae to equate with different types of telephone situations.
Erlang C Model
Most call centers use a model called Erlang C to determine resource requirements to handle incoming calls. The assumptions behind the Erlang C model are that the events (or calls) arrive randomly within the work period (hour or half-hour). The Erlang C model also assumes that if a call attempt is made and no resource (in this case, a call center agent) is in place to handle the call, the call will go into a queue and wait there until there is a resource to handle it. It is the model used most frequently in a simple call center scenario where a caller is asked to hold for the first available agent when entering the queue.
A is the total traffic offered in units of erlangs
N is the number of servers
PW is the probability that a customer has to wait for service
Other Erlang models include Erlang-Engset and Equivalent Random Theory. Determining which of these models to use depends upon the arrival pattern of the incoming contacts.
One of the factors that can affect staff numbers is the arrival rate of calls. Nearly all the time we assume that calls arrive in a RANDOM fashion. However, calls might arrive in a SMOOTH or a PEAKED pattern. Calls that arrive in a smooth pattern (one right behind the other like paperwork) will require fewer staff than Erlang C might suggest. This smooth pattern of calling can use a model called Erlang-Engset. Calls that arrive in a peaked pattern – an “all or nothing” demand – will require more staff than Erlang C might suggest and staffing requires another model called Equivalent Random Theory.
In the next article in this series, we’ll learn more about workload calculations and how to determine the right number of people to handle incoming calls, as well as other types of contacts like outbound calls, chat, and email. Stay tuned…
Penny Reynolds was Co-Founder of The Call Center School and is a popular speaker and writer in the area of call center operations. Recently retired, she serves as an Educational Advisor to SWPP, continuing to provide thought leadership and training to the workforce management community. She can be reached at firstname.lastname@example.org or at 615-812-8410.