Workforce Engagement Management

 View Only

Discussion Thread View
  • 1.  Intraday Monitoring View

    Posted 01-27-2023 12:56

    We have recently starting using the Intraday Monitoring View but cannot find how it is calculating the stats it is providing (for example: what shrink and occupancy it is using). We have read articles on how to use this feature and are not seeing exactly how this is calculated. 

    We are also wondering if we can edit how it is calculating these number or if the formula is set in stone? 

    #Genesys Cloud CX
    #Workforce Engagement Management
    #Performance Management
    #Interaction Analytics

    Addison Hild

  • 2.  RE: Intraday Monitoring View

    Posted 01-28-2023 09:54

    See the (?) icons at the top of each column? Hover over them for information on how they are calculated. Any calculation/formula is fixed - being able to change these in one part of the system would lead to confusion. The formulas we chose are industry standards (and exist in multiple products for more than 30 years). also has information. One item I would point out is Service Level because you can change how this is calculated system-wide (Admin / Analytics) with the options to include/exclude flowouts, short abandons, abandons. Just a note on that from a WFM-perspective: discounting or receiving 'credit' for flowouts and/or abandoned interactions to gain calculated service level is contradictory to WFM - essentially you are hiding the fact you might have a real service problem with flowouts and abandons.

    Jay Langsford
    VP, R&D

  • 3.  RE: Intraday Monitoring View

    Posted 01-30-2023 10:36
    Hi Jay - Thanks for providing some insight on this. Our organization uses ASA rather than service level to determine how our Contact Center is doing. We understand how the ASA is calculated, we just want more insight on how the ASA is forecasted at the interval level. Does the forecasted ASA use shrink, forecasted calls, etc.? Our organization is wanting us to get more in depth to be able to forecast the ASA for the interval and the day as a whole. 

    The help articles we have looked at don't provide this definition. Is there an additional article or place that we can determine this?

    Mackenzie Becker

  • 4.  RE: Intraday Monitoring View

    Posted 01-30-2023 11:37
    Edited by System 01-26-2024 19:28

    ASA, SL, and ABD are predicted based on the forecast volume, AHT, and planned staff (specifically those scheduled to be On Queue with their capabilities taken into account). Forecasting service performance really does not make any sense since it depends on many other pieces - i.e., doing time-series forecasting for ASA wouldn't factor in volume and AHT forecast, planned staffing, etc.

    Those predicted service performance metrics take planned shrinkage into account (i.e., breaks, meals, meetings, etc. that are actually planned on a schedule). They do not take unplanned shrinkage into account, but we have both predicted staffing levels with and without shrinkage that utilize any unplanned shrinkage % you've configured per day/interval.

    When I say predicted there is lots of tech involved. From a high-level we utilize continuous simulation modeling. This is our custom proprietary, patented, state-of-the-art contact center staffing and service performance modeling that leverages our big data infrastructure to train models on a nightly basis. This is used in conjunction with mathematical optimization to come up with an optimal staffing plan that best meets the service performance goals set for your contact center. It includes capabilities such as:

    • Built-in, automated capability to determine customer patience curve and scheduling inefficiency
    • Optimized model hyper-parameters that best correlates service performance (SL, ASA, ABD) with staffing input (available agents, volume, handle time)
    • Sophisticated mathematical optimization for determining the most optimal staffing plan using the latest-and-greatest operational research/AI algorithms and technologies

    List of algorithms we use (not exhaustive):

    • Mixed integer programming
    • Large Neighborhood Search (LNS)
    • Greedy search algorithm
    • Constraint propagation
    • Column generation (Dantzig-Wolfe Decomposition)
    • Failure-directed search
    • Recursive backtracking search
    • Bender's decomposition
    • Network optimizer
    • Brand & bound
    • Branch & cut
    • k-means clustering
    • Barriers interior point method

    Jay Langsford
    VP, R&D

Need Help finding something?

Check out the Genesys Knowledge Network - your all-in-one access point for Genesys resources