About skill settings

There are a couple of settings in the skill properties that require some extra consideration; Midnight break and Abandon rate.

Midnight break

The midnight break is the break point between the workdays for a skill. Use midnight as the midnight break if the workloads close before or at midnight. Define the midnight break to the actual time the workday ends if the workloads close after midnight.

The purpose of the midnight break is to achieve a separation of the dates that is more realistic than to assume that midnight is the breakpoint. For example, calls coming in just after midnight are sometimes more related to the previous day. Moving the midnight break makes it possible to create an intraday distribution pattern that extends past midnight.

  • The open hours are 07:00 to 00:00. The midnight break is set to 00:00.
  • The open hours are 07:00 to 02:00. The midnight break is set to 02:00.
  • The skill is open 24/7. The midnight break is in most cases set to 00:00.
  • The skill is open 24/7 and the flow of calls is still high past midnight but dies down around 03:00. The midnight break can be set to 03:00.
  • Possible values for the midnight break are between 00:00 and 08:00.
  • The midnight break for a skill cannot be changed. If you need to change it, a new skill must be created.
  • The midnight break is used for forecasting purposes only. Reports are still based on dates.
  • Agents' shifts belong to the date they start in their local time zone. The open hours for a workload cannot extend past the midnight break.

Abandon rate

The abandon rate is the percentage of the total volume that is abandoned.

Define the predicted abandon rate when creating a telephony or chat skill. The entered abandon rate is used in the Erlang A model to estimate the callers’ average patience, which is the average time that a caller will stay in the queue before abandoning the call. This is then used while forecasting the number of agents needed and the predicted service level.

Analyze what the abandon rate is at times when the service level is reached or when it is acceptable. Enter that abandon rate in the skill properties. Do not use the average abandon rate.

Possible values for the abandon rate are between 0%-30%. If you use an abandon rate of 0%, that means that the forecast is created based on the assumption that no one will leave the queue. This gives the same result as using the Erlang C model.

IMPORTANT   Use the abandon factor with caution. Start with very low values and adjust upwards, if needed. An abandon rate that is too high will negatively impact the staffing calculations.

Related topics