Can anyone clarify the calculation methodology for ‘Script Completed’ versus ‘Script Abandoned’ metrics within the Performance Dashboard when utilizing complex Agent Scripting templates in a Genesys Cloud EU-West BYOC deployment?
Our organization relies heavily on structured Agent Scripting to ensure regulatory compliance across voice interactions. The current Architect flow design mandates specific node traversal for compliance acknowledgments. However, recent audits reveal a significant variance between the expected script completion rates and the actual data presented in the Agent Performance views. Specifically, agents who navigate through the mandatory compliance nodes but subsequently transfer the call to a supervisor queue are frequently flagged as ‘Script Abandoned’ in the daily reports, despite the critical compliance steps being executed.
The environment is running the latest stable release of the Genesys Cloud platform. The Architect flow includes a dedicated ‘Script Progress’ variable that updates upon node completion. We have verified that the variable is correctly populated before the transfer occurs. Despite this, the Performance Dashboard does not reflect the partial completion status accurately. The metric seems to binary: either the entire script is marked as complete, or it is marked as abandoned if the call does not terminate within the original script context.
This discrepancy creates a significant reporting challenge for our quality assurance teams, who need to distinguish between agents who skipped compliance steps and those who followed the procedure but required escalation. The current aggregation logic appears to ignore the granular progress data captured by the Architect flow variables.
Is there a configuration setting or a specific reporting filter that allows for the recognition of partial script completion based on node traversal, rather than relying solely on the final disposition of the interaction? We require a solution that aligns the dashboard metrics with the actual procedural adherence observed in the call recordings.
Is there a configuration setting or a specific reporting filter that allows for the recognition of partial script completion based on node traversal, rather than relying solely on the final disposition of the interaction?