Grade a New Hire's First Ten Tickets Against a Rubric

Write the rubric from the task

Start with the definition of done for the queue, not with general qualities such as attention to detail. List the observable parts of a correct ticket: the customer question restated, the source checked, the response sent, the status updated, and the follow-up scheduled. Add the fields that must never be left blank. Each line should be something a reviewer can mark yes or no. If the team cannot agree on a line, the task definition is not ready for a rubric.

Keep the rubric to a single page and use the same wording as the team's procedure. Add a short example of a passing ticket and one that fails a specific line. The examples do the teaching that a scoring table cannot. Review the rubric with the trainer before the new hire starts so the expectations are aligned from day one.

Sample the first ten closed tickets

Pull the first ten tickets the new hire closed without another person rewriting them. Read them in order, because the sequence shows whether a correction carried forward. Include any ticket that was reopened within a week, even if it was closed correctly the second time. The reopened items are often the most informative.

Do not choose only the clean examples. A sample of easy tickets flatters the result and hides the gap that will appear under real volume. If the first ten are unusually simple, extend the sample until it includes a normal mix. Record the ticket identifiers and the date of each review so the finding can be checked later. Keep the sample with the onboarding record rather than in a chat thread.

Separate knowledge from judgment

A missed field may mean the new hire never learned the field, or it may mean they knew the field and decided it was optional under pressure. The first is a training gap and the second is a judgment gap, and they need different responses. Write the reason beside each failed line before deciding what to change. Ask the new hire to explain the choice before assigning the label.

Look for patterns across the ten tickets rather than isolated mistakes. One missed attachment may be a busy day, while three in a row suggests the checklist is unclear. A repeated wrong tone may mean the voice sample was never discussed. The pattern points to the instruction, the example, or the review step that needs repair.

Score the same ticket twice

Have the trainer and the team lead score the same three tickets independently, then compare. Disagreements show where the rubric is ambiguous or where the reviewers hold different standards. Rewrite the line that caused the disagreement before using the score to make a decision. A rubric that only one person can apply is not yet a shared standard.

Keep the scoring notes short and factual. Quote the ticket and name the rubric line rather than describing the person. The record should be usable in a coaching conversation without becoming a character judgment. Store it with the onboarding file and review it at the end of the first month. A short note about what changed is enough.

Feed one fix into the next shift

Choose the single highest-value correction and apply it before the next shift. That may mean adding an example, rewriting a checklist step, or pairing the new hire with a reviewer for one more day. Do not change five things at once, because then you cannot tell which change helped. Tell the new hire what will be different and why.

Recheck the next five tickets against the same rubric. If the correction worked, the specific line should improve while the rest stays stable. If it did not, inspect whether the new instruction was read and understood. The sequence of small tests is more useful than a monthly score that arrives too late to change the training.

Decide when to loosen review

Set the condition for reducing review in advance, such as ten consecutive tickets that pass the critical lines without a rewrite. Name the lines that always need a second read, even after the probation period. The condition should reflect the risk of the queue rather than a fixed number of weeks. Write it down so the decision does not depend on how busy the week feels.

Reduce review in steps rather than removing it at once. Move from full review to a sample, then to a weekly sample, and keep the rubric in use for coaching. Record the date of each change and the evidence behind it. If quality drops, the team can return to the previous level without treating it as a failure.

Keep the rubric after the review level changes, because coaching still needs a shared reference. A monthly sample of five tickets is enough to catch drift without returning to full review. Give the new hire the sample result and one specific next step. The conversation should take less time as the rubric becomes familiar. If the sample fails the critical lines twice, return to full review for a week.