Workforce Management Demystified

Workforce Management Demystified
By Turaj Seyrafiaan
Do you have a Workforce Management team in your call center? If the answer is no, then how do you properly plan for the incoming calls and schedule your agents? If the answer is yes, is the process efficient and effective? What is Workforce Management anyway? In this article we try to answer some of these questions.

Workforce Management, often referred to as WFM, is an essential part of the operation in any contact center. The effective use of the available staff can enhance the service level realized and minimizes the wait time and frustration for the customers. At the same time, a timely and flexible schedule can increase employee satisfaction and ensure that staff are available for various tasks when needed while minimizing the overall operating costs by reducing labor costs during the down times. A complete WFM process is required to create planning document, call volume forecasts, agents’ schedule and intra-day adjustments. In short, Workforce Management is all about having the right number of agents at the right time!!

So, what is involved in a complete WFM process? A complete process includes four major stages: Planning, Forecasting, Scheduling and finally Intra-Day. It is imperative for call center managers to understand that all these activities are essential in creating a robust WFM process and one step cannot be a substitute for another!

As the name implies, the Planning stage is where it all begins. Using high level forecasting techniques, call centers can come up with their expected annual work-load (work-load is the total time required to handle the customer calls, i.e. call volume times average handle time). From here, the center can calculate the overall staffing requirements (which includes shrinkage), hiring timelines, training requirements and timelines as well as vacation allocation and last but certainly not least, the total budget.

Shrinkage is the difference between the total number of staff versus the number of agents being on-line answering the calls. In North America, shrinkage typically ranges between 20% to 40%. This means that a typical contact center must employee up to 40% more agents to ensure adequate coverage for the incoming calls. At first, to an outside observer, this number may appear to be high since this represents the ratio of staff that are not on-line and therefore are not providing value to the center!! But let’s consider what is included in shrinkage: shrinkage includes any and all off-line activities that a contact center has to endure daily throughout the year.

Typically these include lunch and coffee breaks, training, coaching, team meetings, holidays, vacations, ad hoc projects and finally absenteeism. Shrinkage factor varies from center to center depending what is included in the calculation as well as overall operation of the center and we may observe higher shrinkage factor In European contact centers as agents receive more vacation days.
While discussing the planning stage, it is important to talk about the planning document. This is where all the information (relevant to the Workforce Management) is documented. A comprehensive planning document provides a high level picture of the operation which includes monthly work-load, overall staffing (as well as hiring and training plans) and the expected operating expenses.

One factor that is missed in many centers planning is the impact of other departments (such as Marketing or Finance) and their activities in generating incoming calls. The time-lines for any major activities that would impact the work-load should be discussed and its impact must be considered during planning stage.

Forecasting is a combination of art and science!! During this stage, the historical call patterns are analyzed and combined with the other relevant information, such as sales growth, to generate detailed forecast for the incoming call volumes and AHT. A typical forecast, predicts call volumes for any given day-part intervals (majority of centers use 15 or 30 minutes intervals) for the forecasting period. While forecasting stage is based on the historical call patterns, there may be certain requirements for a final adjustments based on the latest information as well as changes in the environment (such as economic downturn or increased activities due to a special event).
The forecasting analysis is typically based on one or combination of the two major forecasting methodologies: Time Series and Explanatory. As the name indicates, Time Series methodology assumes that there is a certain cyclical pattern (derived from historical analysis) that can be used to predict the future call volumes. Explanatory, on the other hand, is based on regression analysis assuming linear relationship between call volumes and certain external drivers (e.g. billing calls are directly related to the number of bills being sent out).
The details and complexity of these methodologies have forced many contact centers to look at outside vendors to purchase a WFM software. Although for larger centers the expenditure can be easily justified, the risk is in not truly understanding the methodology being employed which will result in poor forecasts. Smaller centers, on the other hand, have a different problem! The high cost of the software cannot be justified, while the center does not have the expertise and qualification to develop their own forecasting procedure. As a result many small centers simply do not include forecasting in their overall WFM portfolio or perhaps rely on very crude estimates based on last year data and expected growth!!
Another reason that smaller centers use to justify their lack of forecasting process is that their call volume is very volatile and cannot be accurately forecasted. We should note that best in class call centers can forecast their call volume within 3% of actual for 95% of their daily intervals regardless of their size.

After forecasting the number of incoming calls for each interval, the next step is to determine the number of agents required for each interval. Based on Erlang “C” equation, given the number of calls, AHT and target service level, one can easily calculate the number of required on-line agents for each day-part interval. Keep in mind that these are the warm bodies in the seats and does not include staff on breaks, coaching and/or team meetings (i.e. shrinkage).
During the scheduling process, all the available resources are matched against the requirements of the center to create a master schedule. Agent skills, languages etc are all mapped separately and rolled-up into the Master Schedule. The Master Schedule would indicate the start/end/break time for every individual for the desired period. Depending on the availability of resources, a center may have to under-staff or over-staff at certain intervals to create the most efficient schedule. This problem becomes even larger for smaller centers, centers with fixed schedule and centers with limited choice of workforce options (full time / part time / remote). Again, scheduling software, as part of a complete WFM portfolio, can assist significantly in creating efficient schedules and again high cost may prevent smaller centers to try to obtain one (although recently number of vendors are providing hosted services at much more attractive terms).
One major factor in creating effective schedules is to consider all off-line activities such as coaching and team meetings as part of the schedule. There are two main methodologies for doing this: first is to create the master schedule without them, look for time periods where there is surplus of agents (over-staffed) and assigned them to those off-line activities. The second method is to include those off-line activities as part of the basic requirements and allow the software to schedule proper number of agents. In this method, it is imperative that the off-line requirements be added during the intervals with low call volume where surplus of agents is expected. Similarly, centers that complete non-queued work activities such as email or white mail processing can take the same approach.
Final decision about scheduling is to determine what length of time should the schedule cover and how far in advance to publish a schedule. A long schedule published several weeks prior to the actual dates will be much less accurate than a short one being published closer to the date as there are always changes (both in forecasts and available resources) that will impact the effectiveness of the schedule. [The difference between scheduling assumptions and actual data is called ‘melt factor” and obviously it increases as we move further from the date when the schedule is published]. At the same time, earlier publication of schedule can enhance agents’ satisfaction significantly while shorter time-lines would create dissatisfaction. For these reasons, it is important to find a balance between agents’ satisfaction and scheduling effectiveness. In most centers this balance is achieved by providing a two weeks schedule, two weeks prior to its start.

Even with the best laid plans and calculations, it is necessary to track the operation of the queue (call volumes, service level) and adjust the staffing to ensure that the center is providing the best service level possible, while maintaining a reasonable occupancy rate. The Intra-day team (could be just one individual) is responsible for tracking and reporting the operational indicators, re-forecasting the daily volumes (usually twice a day), reassigning staff to and from various off-line activities and maintaining the overall target service level.
A well trained intra-day team will have a significant impact on the daily results and can make a difference between achieving the target results versus long delays for customers and/or high number of idle staff (low occupancy rate). At the same time, we must realize that intra-day activities cannot replace proper planning, forecasting and scheduling but rather complement the entire process.
Workforce Management is an important aspect of any call centers. It is unfortunate that due to lack of full understanding of the process many centers, especially smaller one with fixed shifts, feel that the procedure cannot be helpful to them and as a result simply ignore the process and accept the outcome!!!

One thought on “Workforce Management Demystified

  1. Chad Westover says:

    “We should note that best in class call centers can forecast their call volume within 3% of actual for 95% of their daily intervals regardless of their size.”

    I challenge you to find ANY call center that handles an average of 20 calls per 1/2 hour or less that consistently 95% of their intervals within 3% (that isn’t manipulating the calls in some manor). This means that this center would have to be exactly correct on 95% of the interval, as a miss of a single call is more than 3%.

Leave a Reply

Your email address will not be published. Required fields are marked *