It’s critical to ensure that you have agents handling transactions at the time they come in. Nothing is more annoying to a customer than long waits, whether it’s holding on the phone or waiting for an email response. The ability to forecast and schedule appropriately is both art and science. Using workforce management tools is key to ensuring that your agents are available at the right time.
The goal is to schedule the right number of people, or your costs can increase. Too many and you have bored agents waiting for something to do, and too few means unhappy customers waiting for an answer. Typically the operations team will perform the following basic steps when determining the correct schedule:
1. Gather and analyze historical data.
2. Forecast call workload.
3. Calculate staff requirements.
4. Create staff schedules.
5. Track and manage daily performance.
The science is using historical data to predict behaviors. The art is learning how to become flexible and ensure that changes can be made based on unexpected transaction increases. Basically, your team has to predict the future and then respond accordingly. Not an easy task!
Using Historical Data
Having historical data is critical, but how far back do you go to create an effective predictor of future behavior? Many times the data set considered is too short to actually determine a trend, particularly when seasonal spikes and downturns are at play. To determine trends requires approximately 24 months of data. Much less than two years will not provide the most accurate tracking.
Does historical data really predict the future? It’s a good indicator, but how do you handle those issues that only show up on the day – like absenteeism or unexpected spikes?
Because agents are not always available, taking historical data and applying the agent human factor is needed to create a good forecast. For example, the flu may hit your office, or perhaps critical training is needed for a new product roll-out. Now you have a situation that is not included in historical data.
The trick is to determine the appropriate shrinkage range for your contact center and use that percentage when forecasting schedules. Most centers have a shrinkage range from 20% to 35%, which can easily be accounted for by dividing the Erlang staff requirement by the productive staff percentage (or 1 minus the shrinkage percentage).
Of course, there are also planned absences, such as vacation (particularly this time of year). These are easier to factor in, but on occasion there will be the unplanned or last minute occurrence. It is important to factor this into the shrinkage percentage in order to forecast properly.
Forecasting is not for the faint of heart.
The reason that workforce management tools are popular is because good forecasting is not an easy task. However, the cost savings accrued from properly staffing agents coupled with the increased levels of customer service from having the proper number of agents to handle transactions far outweigh the complexity of utilizing these tools. Solid forecasting is the secret to managing those ever changing schedules and making customers happy!