AT&T 555-230-520 manual EWT Algorithm, When to Use Wait Time Predictions

Advanced Vector Routing

The VRU can also announce expected wait time to a caller periodically throughout the time that a call is in queue. In this way, the caller can observe his or her progress up the queue. However, this approach should be used with caution. Circumstances such as a reduction in the number of agents or a sudden influx of higher priority calls could cause the caller’s expected wait time to increase from one announcement to the next.

If the call is not queued or if it is queued only to splits that are unstaffed or splits where all agents are in AUX work mode, the end-of-string character “#” is the only data item outpulsed.

The EWT Algorithm

EWT is calculated using an algorithm that is based on the number of calls in a queue at a particular priority level and the rate of service of calls from the queue at that priority level. It adjusts for many other factors such as multiple split queuing, call handling times, and the impact of direct agent calls on the wait time of other calls to the split. The algorithm adjusts EWT immediately for changes in staffing, such as agents logging in or taking breaks in AUX work mode.

Since changes occur constantly in a call center, and since EWT cannot predict the future, the accuracy of the EWT predictions will be in proportion to the rate at which call s are serviced from the queue and the level of stability achieved in the call center between the time that the prediction is made and the time that the call is serviced from queue.

When to Use Wait Time Predictions

Wait time predictions are best suited for medium or high volume call scenarios. In general, the potential accuracy of a wait time predictor increases as the rate of removal from queue increases. It is recommended that EWT be used when the rate of removal from queue at a given split priority level is at least one call every 30 seconds.

Predictions can be made for a split with multiple priority levels in use as long as the majority of calls are delivered to the lower priority levels. If the majority of calls are queued at the higher priority levels, any predictions made for the lower priority levels may not be accurate.

The following list describes circumstances that will limit the accuracy of the wait time predictions.

■Immediately after a system restart or when a new split is administered.

The EWT algorithm uses a combination of historical and real-time information to make predictions. When no historical information exists, such as when a new split is added or a reset system 3 or 4 is completed, there is the potential for inaccuracies.

6-4Issue 4 September 1995