Find ticket handoff delays between support queues

A ticket can spend most of its elapsed life between teams even when each specialist's logged work is brief. To investigate, reconstruct where it resided and when responsibility changed. Call the result queue residence time unless the data also establishes that no work occurred during the interval.

Support systems use queues to organize requests for agents, as described in Atlassian's service management feature guide. That organizational role does not make queue assignment a reliable measure of active effort or an automatic SLA pause.

Choose interval-level records

Use one row per ticket and continuous queue residence interval. A ticket entering the same queue twice has two intervals, not one blended stay. Preserve a separate transfer relation with the sending queue, receiving queue, transfer event and acceptance evidence where available.

Inputs include ticket key, event ID, timestamp with time zone, previous and next queue, event sequence, extraction cutoff, ticket status and any assignment acceptance event. Work logs are a separate table. Define whether queues represent teams, named agents or mixed routing buckets before comparing them.

Follow the ticket-history reconstruction guide for event ordering and provenance. A final current-owner export cannot reconstruct earlier residence intervals by itself.

Reconstruct the path without inventing events

Sort valid events by timestamp and source sequence. Open an interval at a recorded queue entry and close it at the next recorded exit. The final known interval closes at the snapshot cutoff for an open ticket and is labeled ongoing.

If the first observed event is already a transfer, the earlier queue's start may be unknown. Do not substitute ticket creation unless the source proves initial assignment occurred then. Missing exits, equal-time conflicting transfers and duplicated imports need exceptions rather than guessed durations.

Calculate wall-clock residence in minutes or hours. If business-time residence is also useful, produce a separately labeled measure with a named calendar. Neither is equivalent to the qualifying clock in the SLA business-hours method.

DEMO: one transfer path

This synthetic ticket uses Eastern time on October 6, 2026. The report cutoff is 4:00 p.m.; the recorded assignment history is complete for the illustrated interval.

Residence totals 420 minutes: 30 + 150 + 240. Logged active work totals 65 minutes. Their 355-minute difference is unlogged elapsed time, not proven idle time. Staff may have worked without logs, or the ticket may have awaited information within the queue.

The immediate decision is to confirm Fulfillment's current next step and inspect whether Specialist-to-Fulfillment acceptance is recorded. The longest interval identifies a place to investigate; it does not establish blame or prove a team needs more staff.

Compare routes and accepted handoffs

Across many tickets, group by ordered path and request type. Report transfer counts, repeated returns, known residence distributions and the number of ongoing intervals. Completed intervals and current open ages answer different questions, so display them separately.

Where acceptance events exist, measure transfer-to-acceptance duration as its own interval. Without them, the transfer timestamp only shows routing, not when the receiving team acknowledged responsibility. Sample the underlying events before labeling a path an avoidable handoff.

Stratify materially different severity and request types. Compare like-for-like routes before attributing a long stay to a queue that handles the hardest cases.

Keep handling effort and contractual time separate

The effort per account analysis concerns work consumed. Queue residence concerns elapsed placement. SLA attainment concerns the applicable obligation. A useful review can show all three with distinct definitions and reconcile their source records without treating them as interchangeable.

Do not count overlapping agent assignments twice in a single queue timeline. If parallel teams legitimately hold the same ticket, model parallel assignments explicitly rather than forcing a fictional sequential path.

Use the evidence in an operational review

Start with one recurring path and inspect examples with both teams. Record whether the next change concerns routing clarity, acceptance evidence, missing information or actual handling capacity. Compare later snapshots under the same definitions and retain unresolved exceptions.

Contact Covirage to discuss a bounded handoff review using the event fields available. The first useful result is an explainable path and a shortlist of intervals to examine, with uncertainty shown alongside the timing.