Volume metrics
Tickets created, tickets resolved, response rate, and channel split. The fastest way to spot a volume spike or a resolution gap.
Volume metrics
The volume section of /analytics is the heartbeat of your support operation. One headline KPI plus three charts. If something is wrong, this is usually the first place you see it.
The headline KPI: Tickets opened
Every conversation that started in the selected period — new email threads, new chat sessions, new Slack Connect threads. Reopens don't count, because the conversation already existed.
A flat tickets-opened line over time is healthy. A spike means something happened: a marketing send, an outage, a pricing change. Use the volume chart below to find the day, and the inbox to read the conversations.
The three charts
Volume per day
A bar chart of tickets opened each day in the selected period. The default 30-day view shows daily bars. Switch the period selector at the top of the page (7d / 14d / 30d / 90d) to change the window.
This is the "did anything weird happen" chart. A healthy team has a stable shape; a release day, a marketing send, or an outage shows up immediately.
Resolutions per day
A line of conversations closed each day, on the same time axis. Compare it to the volume bars: if resolutions trail openings for several days, your backlog is growing. If the two move together, you're keeping up.
A conversation closed today doesn't have to have been opened today. So in any given day "resolutions" can exceed "volume opened" — that's a backlog burndown, which is good.
Channel mix
The share of opened conversations in the period, by channel: email, chat, Slack Connect, and any custom channels you have configured.
Use it to spot drift. If chat suddenly accounts for 60% of volume when it was 30% last month, your widget is doing something different — a launch, a footer change, a new help-center entry pointing at the widget.
What "opened" actually means
A few edge cases:
- Imported conversations from a migration count from their original date, not the import date. A Zendesk export from January will show up as January volume.
- Reopens don't add to opened — they continue the original conversation.
- Bot-created conversations (uptime alerts, internal pings) count, unless you've flagged them as system tickets in routing.
What "resolved" actually means
A conversation counts as resolved the moment it moves to status='closed'. That includes manual close, auto-close after inactivity, and AI-driven resolution under autopilot.
Two things to watch:
- Auto-close can inflate the resolved number if your inactivity window is short. See Auto-close.
- Reopens don't subtract from a previous day's resolved count. They add a new resolved event when the conversation is closed again.
The period selector
The same picker that drives the rest of /analytics (7d / 14d / 30d / 90d) drives these charts. Changing the period recomputes everything.
Related
Was this article helpful?