Outbound Routing Metrics Divergence in Analytics Reporting Dashboard

General Discussion
1

Error: Metric Mismatch in Conversation Summary Reports

Why does this setting in the Analytics Reporting dashboard produce inconsistent call duration totals when filtering by specific BYOC trunk groups? I am managing fifteen distinct BYOC trunks across AP-SG and EU-FR regions, and the outbound routing logic recently underwent a failover test. The test triggered a shift from the primary carrier to the secondary SIP trunk due to a simulated 503 Service Unavailable response.

When querying the /api/v2/analytics/reporting/summary endpoint with a filter for routing.trunk.id, the conversation.duration metric aligns with the CDRs exported directly from the carrier portal. However, the interaction.duration metric in the same report shows a variance of approximately 150ms to 2 seconds per call. This discrepancy compounds significantly across high-volume trunks, leading to inaccurate SLA calculations in our performance dashboards.

The environment is running the latest Genesys Cloud release candidate. I have verified that the SIP registration remains stable during the failover, ruling out packet loss as the cause. Is there a known latency in how the analytics engine aggregates timestamps when a call is bridged across different carrier endpoints mid-session? I need to reconcile these figures for the monthly carrier billing audit.

2

You need to verify the metricId in your Analytics v2 query matches the digital channel schema exactly. Use conversation:duration instead of generic call metrics. Ensure the groupBy array includes routing:queueId and telephony:trunkId to isolate BYOC trunk data accurately. Misaligned schemas cause these duration discrepancies.

3

This has the hallmarks of a classic case of expect… Zendesk tickets don’t have “duration” in the same way, so you might be missing the telephony:mediaType filter. Try adding that to your groupBy to ensure you’re only counting voice interactions on those trunks.