Omnichannel Queue Metrics: Tickets Entering, Exiting, and Backlog by Status
Qvasa keeps a history of every stay a ticket makes in a Zendesk omnichannel queue. This guide covers the metrics built on that history: how much work entered a queue, how much left it, how much was waiting at any moment — and, for each of those, a breakdown by the ticket status the ticket arrived with, so you can tell new demand from re-opens.
Why This Exists
A backlog number on its own tells you that a queue is deep, not why. When the queue spikes at 4am, the question is whether a provider pushed a batch of brand-new tickets or whether yesterday's tickets came back as re-opens. Those two situations call for different people and different fixes.
The queue history answers three questions, and each of them can be split by status:
How much work arrived? Tickets Entering Queue — a volume (flow) metric counted at the moment the ticket entered.
How much work left? Tickets Exiting Queue — the mirror image, counted at the moment the ticket left, whether an agent accepted it, it was transferred, or it was otherwise removed.
How much was waiting? Average Queue Height — a moment-in-time (stock) metric that reconstructs the queue at each point and averages it per bucket.
Every metric comes in two flavours: for a single queue, and for a Queue Collection (a named group of queues you define in Qvasa — see Omnichannel Queue Collections). A collection's number is always exactly the sum of its member queues' numbers.
Before You Start
Your account needs Zendesk omnichannel routing queues enabled and syncing into Qvasa. If you can't find the widgets below, that's the reason — reach out and we'll switch it on.
To use the collection variants, create at least one Queue Collection first (Settings → Omnichannel Queue Collections). Collection widgets hide themselves until a collection exists.
The status breakdown records going forward from the day it was enabled on your account. Queue stays from before that show up as a single "Unknown status" series until they age out of your date range. The unbroken entering, exiting, and queue height charts are complete for all history.
Status names follow your account's ticket status mapping. If you've renamed statuses in Qvasa, the series use those names; otherwise you see the default Zendesk names.
Step 1: Chart What Arrived and What Left
On any dashboard, enter edit mode, add a widget, pick Basic chart, and search for "Entering Omnichannel Queue" or "Exiting Omnichannel Queue". Choose the queue (or Queue Collection) from the dropdown on the right and set the time unit. Time units run from 15 minutes and 30 minutes through hours, days, weeks, and months, so you can watch a queue bucket by bucket during a busy hour or step back and read a quarter. The same metric exists as a Quick stat (Tickets Entering Omnichannel Queue) with a comparison to the previous period and a ticket drill-in.
Read the two charts together: when entering runs ahead of exiting for a few buckets, the backlog is growing and the queue height chart will show it. When exiting catches up, the backlog is draining.
Step 2: Split It by the Status the Ticket Arrived With
Add a widget, pick Multi data chart, and search for "queue". The Queue Backlog by Status group holds six widgets — the entering, exiting, and average queue height metrics, each for a queue and for a Queue Collection. Select one, choose the queue or collection on the right, and pick Stacked Bar or Stacked Area as the chart type.
Each series is one ticket status — the status the ticket held as it entered the queue. In practice that means New is new demand and Open is usually a re-open, since tickets enter queues in one of those two states (Closed never appears — a closed ticket cannot sit in a queue).
The stack always adds up to the plain chart. Tickets Entering Queue by Status for a queue sums, bucket for bucket, to Tickets Entering Queue for that queue. Nothing about the numbers you already use changes; you can now see what's inside them.
Step 3: See the Backlog Itself, by Status
Average Queue Height by Status Over Time is the stock view: for each bucket, the number of tickets waiting is sampled minute by minute and averaged, with one series per entry status. Because it reconstructs the queue as it was, selecting a past date range shows you the backlog as it stood then.
Choose a date range at least as long as the backlog you want to see. The reconstruction only looks back one date-range length before the start of your range, so a 90-minute window won't reconstruct a ticket that has been waiting since yesterday. If your queues carry days-old tickets, look at days, not hours.
Step 4: Do the Same for a Queue Collection
Every widget above has a Queue Collection twin — search for "queue collection" in the Basic chart, Quick stat, or Multi data chart lists. Pick your collection from the dropdown and the chart covers every member queue in one series (or one stack).
A collection's series is exactly the sum of its member queues' series. If a ticket passes through two member queues, it counts once per stay, so the collection can read higher than the number of distinct tickets — that's expected, and it's the same rule the single-queue charts follow.
Step 5: Click a Bar to See the Tickets Behind It
Every chart here answers "how many". The drill-ins answer "which ones".
Click any segment of a stacked bar on the entering and exiting by-status charts and Qvasa opens the tickets behind exactly that slice: that time bucket, that entry status. A bar showing 14 tickets that arrived as New between 09:15 and 09:30 opens those 14 tickets. This is the fastest path from "something spiked" to "here is what it was about", and it works on the single-queue and Queue Collection versions alike.
The lightbox that opens is the standard ticket inspect, so you get your configurable column sets, sorting, and CSV export on the result. See Ticket and Agent Drill-in Capabilities with Configurable Column Sets for how to shape those columns.
For the whole window instead of one bar, open the widget menu and choose Inspect. The Tickets tab lists every ticket behind the chart across the full date range and all series.
How the Numbers Are Counted
These metrics count queue stays, not tickets. A ticket that enters the same queue twice counts twice on entering, twice on exiting, and is waiting twice as far as queue height is concerned. The inspect drill-in lists the distinct tickets behind the stays.
Entering is dated by when the ticket entered; exiting by when it left. A ticket that entered on Monday and left on Tuesday counts under Monday on the entering chart and under Tuesday on the exiting chart.
A ticket still waiting has not exited and is not on the exiting chart; it does appear on entering and on queue height.
The entry status is not refreshed while the ticket waits. If a ticket entered as New and became Open while it sat in the queue, it stays in the New series for that stay.
"Unknown status" collects stays with no recorded entry status: stays from before the breakdown was enabled, phone-call stays, and stays Qvasa has not reconciled with Zendesk yet (that usually takes a few minutes). The series only appears while such stays exist, so the stack still sums to the plain chart.
All channels are covered — messaging, support (email), and phone. Queue stays Qvasa has not yet matched to a ticket are excluded everywhere.
Ticket filters apply. A tag or attribute filter on the widget narrows every one of these charts, drill-ins included, to the matching tickets' stays.
A ticket status filter selects the status the stay began with. Filter these widgets to New and you get the stays that arrived as New, which is the same population the New series counts and the same tickets a click on that segment returns. Filter, stack, and drill-in always agree.
Tips
Put entering, exiting, and queue height on one row per queue collection. Arrivals, departures, and what's waiting tell the whole story of a queue at a glance.
Use the stacked view to triage a spike. A New spike points at something upstream (a product change, a provider, a campaign). An Open spike points at re-opens — replies that didn't land, or tickets solved too early.
Switch a multi data chart to its table view when you want interval-by-interval numbers, or export it to CSV.
Drop to 15- or 30-minute buckets when you are watching a spike develop. Hourly buckets are right for a shift review; minute buckets show you the shape of the surge while it is still happening, and each bar is still clickable.
Match the date range to the backlog age on queue height charts, as described in Step 3.
FAQ
Can I get the list of tickets behind one bar? Yes — click the bar segment. You get the tickets for that time bucket and that entry status. For the whole date range, use Inspect from the widget menu.
Why do I see an "Unknown status" series? Those stays have no recorded entry status — most often because they started before the breakdown was enabled on your account. It shrinks as older stays leave your date range, and it's never shown when it's empty.
Why doesn't the status stack match a ticket's current status? The series use the status the ticket had when it entered the queue, on purpose: that's what tells new demand from re-opens.
Why is exiting lower than entering today? Tickets still waiting in the queue haven't exited yet. They'll show on the exiting chart on the day they leave.
Why is my queue height chart flat at zero for a short window? The reconstruction looks back one date-range length. Widen the range so it reaches back to when the waiting tickets entered.
Is queue height a live number? No — it's a moment-in-time reconstruction over a date range. For what's waiting right now, use the live Outstanding Tickets by Omnichannel Queue quick stat. See Understanding live, moment-in-time, and range-based metrics for the difference.