Objective

Introduce QA in stages so reviewers can validate grades, coaches can turn findings into useful follow-up, and agents know which results and actions are shared with them.

Environment

  • An administrator and a quality owner can confirm QA access, service categories, and the intended pilot teams.
  • A representative category has an assigned, published scorecard and suitable Zendesk conversations.
  • The pilot group has agreed on who reviews grades, publishes agent results, coaches, and resolves appeals.

Resolution

  1. Define the pilot. Choose one category, a small set of agents, and a review period. State what the first cycle should prove: usable evidence, consistent reviewer decisions, clear agent feedback, or a complete coaching follow-up.
  2. Prepare the standard and access. Publish the scorecard in QA Settings, verify the category's evaluation scope, and give reviewers the permissions their tasks require. Confirm agent identities are linked to the right workforce people before planning publication.
  3. Run a controlled evaluation. Test a selected ticket, then use a bounded batch from QA Grade Batches if more examples are needed. Check scored, skipped, and failed runs in QA Analysis. Have reviewers compare criterion grades with the cited conversation evidence and record where the rubric needs adjustment.
  4. Agree on the review and publication rule. Decide who can approve or change grades and when an agent result is ready to share. A scored run is internal until it is published; later corrections also need an explicit publish action.
  5. Introduce My Quality to agents. Explain that agents see their own published results, the current quality summary, shared coaching, and available appeal actions. Tell them who to contact about a result and how the appeal deadline is set.
  6. Start coaching with a small set of cases. Use QA Coaching to assign each case to a coach and agent, share specific strengths and improvement areas, agree on actions and due dates, and record the follow-up. Keep private notes distinct from the material shared with the agent.
  7. Review the first cycle before expanding. Check eligibility and evidence gaps, reviewer changes, publication timing, agent questions, appeals, and action completion. Adjust the scorecard or process, then add more categories or agents deliberately.

Use How To Choose Your First QA Workflow if the first pilot outcome is still undecided.