Metric of the Quarter – Average Handle Time

Metric of the Quarter – Average Handle Time

One of the key elements of workload is the time it takes to complete a contact or the average handle time (AHT). Forecasting requires that the volume of contacts be multiplied times the AHT to calculate the workload that must be accomplished in a defined period of time. That makes AHT an equal partner in the forecasting process, but it is often given much less consideration and analysis than the volume.

The calculation of AHT is based on the total time that the agent (or other resource) will be occupied doing the work involved with each contact. In the case of incoming phone calls, that would include the time starting when the agent answers the call, the time of conversation with the caller (including any time that the caller was placed on hold), and the time the agent is performing the after-call work associated with that call and is not yet ready to take the next contact. It does not include any time that the caller spent in the queue waiting for the agent or any time the agent spends in a work state other than talk, hold or after-call work. For emails and other types of interactions, the AHT needs to include all of the elements that require the agent to be unavailable for another contact. For example, an outbound call may require some preparation time before the dialing and some calls will not be answered or will reach a busy signal. This is all time the agent is occupied and must be included in the calculation as well as the ringing time, conversation and any after-call work.

It is important to ensure that the agents use their work state buttons correctly and only place themselves in the after-call work state to complete work related to the contact just disconnected. It is for work that must be done immediately after the call and cannot be delayed. This ensures that the conversation and after-call work are planned for in the same interval. Other types of work, such as research, call-backs, filing, meetings etc. should not be done while the agent is in the after-call work state as these tasks can generally be done when ever the time permits or when it is specifically scheduled. That work will not increase and decrease with the rise and fall of the call volume the way after-call work does. Training the agents and supervisors on the importance of using work states correctly takes constant vigilance but it is critical to an accurate forecast.

But what is in it for them, they might ask. When the forecast is more accurate, the schedules will be better matched to the workload resulting in fewer last minute schedule and break changes as well as demands for overtime or time off without pay.

By definition AHT is an average, but the question is averaged over what period? Some centers use the same AHT for an entire day, week, or even longer. This is based on the assumption that it is an average so if there are  variations along the way, it is to be expected. However, it is common for the AHT to have patterns just like volume. These patterns might have predictability based on the time of day, day, or week, or as a result of the specific driver such as a marketing campaign or weather event.

One way to improve the accuracy of the forecasting process is to analyze the AHT patterns and identify the differences that occur by interval (half-hour typically). Perhaps the early morning calls are relatively short while those in the evening tend to be longer. Maybe the calls during busy periods tend to be longer than during low volume periods. Whatever your pattern is should be part of your forecasting calculation. Should you discover there is a definite pattern of longer times during certain intervals or days, searching for the cause can be quite useful. Perhaps the challenge is that there are not enough supervisors or senior agents to assist with difficult calls. It might be that this is the time when the most junior agents are scheduled such as the most unattractive shifts. It is also common for longer calls to be a function of a caller behavior such as saving the long interactions for after work hours or weekends. If the reason is one that can be identified and addressed, the AHT might be reduced as a result. But if it is not likely to change, then it needs to be considered in the forecast and the number of staff scheduled for those periods to ensure that the service goals can be met consistently.

An accurate forecast is the foundation of the entire WFM process. Spending a bit of extra time to ensure that the AHT by interval is accurate will pay significant dividends in consistency of service for callers as well as a more manageable environment in the center.