Objective

Set which interactions qualify for automated evaluation, how many are selected, and when ZenOps evaluates them.

Before you begin

You need Quality: Manage. The service category must be enabled and have an assigned scorecard with an active published version. Create and publish the scorecard in QA Settings · Scorecards, then open the category in Categories and select QA Analysis to connect it.

Resolution

  1. Open Automated QA. In the category's QA Analysis tab, find Automated QA. The category itself is implicit: this policy considers conversations classified in the category you opened.
  2. Choose eligibility. Select the Zendesk channels to include. Expand the advanced filters only if the pilot needs specific brands, groups, ticket forms, included tags, or excluded tags. Within one filter, selected values can match; filters across different dimensions must all match. An excluded tag removes a conversation from scope.
  3. Choose coverage. Select All, a Percentage, or a Fixed monthly number of qualifying interactions. For a first pilot, choose a volume your reviewers can inspect. Coverage limits selection after eligibility; it does not make an ineligible ticket gradable.
  4. Choose the evaluation trigger. Select when a chosen interaction should be evaluated: first final state, Solved, Closed, after a specified inactivity period, a daily time in company time, or manual only. Ticket finality comes from this trigger, not from a separate status filter.
  5. Review the effective policy. Confirm the summarized scope, coverage, and timing. ZenOps also checks each interaction for an active published scorecard, an enabled category, a public customer message, a public staff response, and an attributable Zendesk staff identity.
    The effective policy summary shows the category’s eligibility, coverage, and evaluation trigger before you save the configuration.
  6. Save and enable. Select Save changes, then turn on Enable QA evaluations. If the switch is unavailable, check the readiness message for a missing or unpublished scorecard. If the category is later disabled or its scorecard loses an active published version, automated QA pauses while the saved policy remains.
  7. Check the first results. Open QA Analysis after new interactions meet the trigger. Automation applies to interactions finished after it was connected. For older tickets, use a ticket test or a bounded batch from the category's QA Analysis area.

For a plain-language explanation of selection and timing, see How Eligibility, Coverage, And Evaluation Timing Work. If the category shows a readiness warning, use Investigate Why A Category Is Not Ready For QA.