IVR Wait Time Calculation Discrepancy in EU1 Architect Flow

Can anyone clarify the methodology behind the “Average Wait Time” metric when displayed in the Performance dashboard versus the raw data exported from the Conversation Detail view?

We are observing a consistent variance of approximately 15-20 seconds in our EU1 tenant. The Architect flow is configured with a standard “Queue” block, followed by a “Gather” block for agent transfer. The business requirement is precise tracking of customer experience, yet the dashboard aggregates data differently than the individual conversation logs.

Specifically, does the dashboard wait time include the duration spent in the IVR menu prior to entering the queue, or does it strictly measure the time from the “Queue” block entry until the agent accepts the interaction? The documentation states it measures “time in queue,” but the business stakeholders are reporting higher wait times based on the dashboard figures compared to the agent-side experience.

The environment is Genesys Cloud EU1, running the latest stable release. The flow version is 42, and the dashboard view is set to “Real-Time” with a 30-second refresh interval. No custom scripting is involved in the queue logic.

**How is the wait time metric calculated in the Performance dashboard when a customer interacts with IVR menus before joining the queue?"