Skip to content
Automation Squad
ZdSMade to order

Reporting & dashboards

Zendesk weekly ticket volume → Slack summary

Last week's ticket counts by channel and group, posted to #support before Monday standup.

On Monday morning the workflow runs three Zendesk searches: tickets created last week, tickets solved last week, and everything still open. Counts go to Slack split by channel and by group, with the five most used tags and the change from the week before. Support leads who tally these by hand get the numbers in the channel before anyone logs in. Two apps, six nodes, no sheet to maintain.

How it works

  1. Zd

    Trigger

    Monday at 08:00 (n8n Schedule Trigger), then three Zendesk ticket searches over the previous seven days

    Zendesk

  2. Zd

    Step 1

    Search tickets created last week

    A type:ticket search in the Zendesk node covers created between last Monday and this Monday, and returns channel, group, tags, priority and status for each ticket.

  3. Zd

    Step 2

    Pull solved tickets and the open backlog

    Twice more the same node runs: once for tickets with a solved date inside the week, once for status below solved, so the backlog is counted as of the moment the report runs.

  4. if

    Step 3

    Count by channel, group and tag; compare with last week

    A Code node counts created, solved and open tickets per channel and per group, ranks tags, and reads last week's totals from the workflow's static data to produce the change.

  5. S

    Step 4

    Post the weekly summary to #support

    Slack gets a Block Kit message: created, solved and open with week-over-week change, a per-channel and per-group table, the top five tags and a link to the Zendesk view.

What changes once it runs

  • Created, solved and open for the previous week reach #support before the Monday standup, every week.
  • Each number carries its change from the week before. A rising backlog is visible the first Monday it rises.
  • Channel and group splits show where the week's tickets came from and which team carried them, without anyone opening Explore.
  • Five tags, the most used that week, say what people wrote in about, in the same message.

What you get

  • n8n workflow export with the schedule, the three searches, the counting node and the Slack message
  • Setup guide with the Zendesk API token steps, the three search strings explained, and the Slack Block Kit JSON annotated
  • Walkthrough video of a first Monday run
  • A group id-to-name map you fill in once, so the message shows team names rather than numbers

Before you start

  • Zendesk Support or Suite on any plan, with API token access switched on in Admin Center and an agent or admin account to own the token.
  • A Slack app with the chat:write scope, invited to the channel that receives the summary.
  • n8n Cloud on any plan, or self-hosted n8n 1.x, with the workflow activated so its static data persists between runs.

Built for

Questions before you buy

We get more than a thousand tickets a week. Does the search cope?
Zendesk's Search API stops at 1,000 results per query. Below that, the workflow as shipped is enough. Above it, the guide swaps the two weekly searches for the Search Export endpoint through an HTTP Request node, which pages through any volume with a cursor.
Is first reply time or CSAT included?
CSAT is close. Each ticket the search returns carries its satisfaction_rating, so counting good and bad ratings is a short edit to the Code node, and the guide shows it. First reply time is not on the ticket. It takes one extra call per ticket to Zendesk's ticket metrics endpoint; at 800 tickets that is 800 extra requests each Monday, and the guide shows where to put it.
Can it write to a Google Sheet as well?
Yes. Add a Google Sheets Append node after the counting step and map the same fields; the guide includes the mapping. Slack-only is kept small on purpose so it runs on any n8n plan with nothing else to connect.
Why is the week-over-week change blank on the first run?
Last week's totals live in n8n's workflow static data, which only fills once the workflow has run while activated. The first live Monday stores the numbers; the second shows the change. Manual test runs do not write static data.
Does it run on Make or Zapier?
n8n only, as shipped. Grouping three result sets by channel, group and tag, and holding last week's totals for the comparison, is Code-node work. In Make it takes a data store and a chain of aggregators, and Zapier's steps are not built for it. On n8n it runs on the free self-hosted build or any Cloud plan.

Nearby on the shelf

Whole catalog