Workforce Engagement Management

 View Only

Discussion Thread View
  • 1.  Forecasting AHT and Queue Performance View AHT calculations

    Posted 12-01-2024 23:01
    AHT in the Queue Performance View is made up of (cumulative talk time + cumulative hold time + cumulative after-call work + cumulative dialing time + cumulative contacting time) / Number of interactions handled while the AHT in the WFM is calculated from average talk time, average hold time, and average after call work time
    You will see that with the Queue Performance, dialing and contacting time is included in the calculation and WFM is strictly looking agent activity time.  Also, with the Performance View, it takes the total divided by the Handled.  Not sure where abandoned calls comes into play in this. 
    This difference in variables can cause a 20-30% difference in the AHT between the two.  What should be used for determining the real AHT for our contact center?  Having two calculations is bad enough, but when the difference is greater than our shrinkage, it makes for some interesting discussions.  

    #Forecasting

    ------------------------------
    Robert Wakefield-Carl
    ttec Digital
    Sr. Director - Innovation Architects
    Robert.WC@ttecdigital.com
    https://www.ttecDigital.com
    https://RobertWC.Blogspot.com
    ------------------------------


  • 2.  RE: Forecasting AHT and Queue Performance View AHT calculations

    Posted 12-02-2024 04:19

    Hi, Robert! 

    I guess the root cause of the issue is that Avg Hold Time and Avg ACW are calculated only for the interactions where it appeared. AHT should be measured EITHER as Total talk + Total Hold + Total ACW divided by Handled volume OR Avg Talk Time + Avg Hold Handled + Avg ACW Handled. These metrics are available in the queue performance views in workspace performance.

    Hope this helps=)



    ------------------------------
    Ekaterina Kononova
    Product Specialist | Data, Analytics & Quality Management
    Sweden
    ------------------------------



  • 3.  RE: Forecasting AHT and Queue Performance View AHT calculations

    Posted 12-02-2024 13:21

    Exactly the problem.  If you just take these values for face value and add them up, they don't match.



    ------------------------------
    Robert Wakefield-Carl
    ttec Digital
    Sr. Director - Innovation Architects
    Robert.WC@ttecdigital.com
    https://www.ttecDigital.com
    https://RobertWC.Blogspot.com
    ------------------------------



  • 4.  RE: Forecasting AHT and Queue Performance View AHT calculations

    GENESYS
    Posted 12-02-2024 08:49
    Edited by Jay Langsford 12-02-2024 08:50

    For WFM we utilize (tHandle.sum / tHandle.count) from analytics for AHT.
    https://developer.genesys.cloud/analyticsdatamanagement/analytics/metrics



    ------------------------------
    Jay Langsford
    VP, R&D
    ------------------------------



  • 5.  RE: Forecasting AHT and Queue Performance View AHT calculations

    Posted 12-02-2024 13:15

    Here is the problem:

    average handle time (AHT) - Genesys Cloud Resource Center

    Shows  (cumulative talk time + cumulative hold time + cumulative after-call work + cumulative dialing time + cumulative contacting time) / Number of interactions handled

    Forecasting metrics - Genesys Cloud Resource Center

    Shows:  Average Interaction Time + ACW

    Now, if you take a Queue Performance and add up the top one then add up the bottom one, there is a 20% to 30% difference - usually higher for forecasting. 



    ------------------------------
    Robert Wakefield-Carl
    ttec Digital
    Sr. Director - Innovation Architects
    Robert.WC@ttecdigital.com
    https://www.ttecDigital.com
    https://RobertWC.Blogspot.com
    ------------------------------



  • 6.  RE: Forecasting AHT and Queue Performance View AHT calculations

    GENESYS
    Posted 12-02-2024 14:30

    Analytics defines tHandle as "The complete time an agent spent on an interaction, includes time spent contacting, time spent dialing, talk time, hold time, and after call work."

    How does that differ from what AHT shows in the resource center?

    I am not dismissing a disparity in different views (which could have other reasons like one view not being able to exactly duplicate the 'filter' another view has), but I am not seeing much difference in what AHT in the resource center states and what analytics says about tHandle.



    ------------------------------
    Jay Langsford
    VP, R&D
    ------------------------------



  • 7.  RE: Forecasting AHT and Queue Performance View AHT calculations

    Posted 12-02-2024 15:02

    I am just going by the metrics shown in the Queue Performance View.  If you go strictly by the calculations in the Resource Center links, you get entirely different numbers.  The problem is that clients will look at the AHT in Forecasting and see a number 25% off from the number in the Queue View and question one or the other.  We are talking just Voice calls to one queue with no skills or other media.  



    ------------------------------
    Robert Wakefield-Carl
    ttec Digital
    Sr. Director - Innovation Architects
    Robert.WC@ttecdigital.com
    https://www.ttecDigital.com
    https://RobertWC.Blogspot.com
    ------------------------------



  • 8.  RE: Forecasting AHT and Queue Performance View AHT calculations

    GENESYS
    Posted 12-02-2024 16:02

    Forecasting (and intraday and elsewhere in WFM), for actuals, simply queries analytics as does the performance views. I think across the app AHT would be (tHandle.sum / tHandle.count) which makes sense just on a quick read as well as the descriptions of the metrics. If we are trying to explain a difference in numbers across views by comparing what is stated in resource center articles, I just don't think that will yield anything. I can update the forecasting metrics resource center, but again WFM simply uses (tHandle.sum / tHandle.count). And the description of tHandle seems the same as the description you posted from the other resource center article on AHT.

    As far as disparities, there could be multiple reasons. Trent Vance had a post noting two that he had run across: https://community.genesys.com/discussion/intra-day-not-matching-queue-performance#bmdd0effc9-b24f-47d7-baf5-0188e84e34e9

    In escalations I have seen about disparities they have generally all been an issue of one view not being able to exactly mimic the filter conditions and time ranges. To troubleshoot specific scenarios I would recommend opening up a ticket.



    ------------------------------
    Jay Langsford
    VP, R&D
    ------------------------------



  • 9.  RE: Forecasting AHT and Queue Performance View AHT calculations

    Posted 12-02-2024 16:38

    Please update the Resource articles.  That would help at least to give a consistent message.  Yes, I know about the None selection in the planning groups.  Again, we looked at 3 or so clients with just one queue, no skills and no other media and the differences were too much to disregard.  



    ------------------------------
    Robert Wakefield-Carl
    ttec Digital
    Sr. Director - Innovation Architects
    Robert.WC@ttecdigital.com
    https://www.ttecDigital.com
    https://RobertWC.Blogspot.com
    ------------------------------



Need Help finding something?

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