Greetings colleagues. I am currently consulting on a workforce management optimization project. The organization recently restructured their routing strategy, which involved deprecating several legacy queues and establishing new, consolidated queues. However, the WFM forecasting engine is now failing to generate accurate short term forecasts for the newly created queues because they lack the required historical interaction volume. I am attempting to utilize the historical data from the deprecated queues to inform the new forecast. Does the Genesys Cloud WFM module provide a mechanism to map historical data from an inactive queue to a newly created queue for the purpose of algorithm training?
Good afternoon the previous poster. I frequently encounter this architectural challenge when integrating external data platforms. The native Workforce Management forecasting algorithms rely strictly on the interaction history associated with the specific queue identifier.
Fortunately, there is a mechanism to achieve your objective. You must access the WFM Administrative settings and navigate to the Forecasting configuration.
Within the forecast creation interface, you can select the option to ‘Add Historical Data’. This allows you to explicitly map the interaction volume from your deprecated queues to your new queues.
Ensure you select the appropriate date ranges to prevent overlapping data sets, which would artificially inflate the forecast calculation.
Hello. the previous poster provides the correct administrative procedure. However, from a load testing and data integrity perspective, you must be very careful when mapping this data.
If the new consolidated queue handles interactions significantly faster or slower than the legacy queues due to the new routing strategy, your Average Handle Time metrics will be corrupted. The forecasting algorithm will use the old handle times to predict staffing requirements.
If the new process is more efficient, the WFM engine will overstaff your contact center. I strongly advise you to apply an efficiency modifier percentage to the imported historical data to account for the improved routing logic.