Two managers can observe similar performance and assign very different ratings. One may reserve the highest score for exceptional results; another may use it for anyone who meets expectations. If those scores influence salary increases, manager inconsistency can become a compensation problem.
A compensation calibration meeting helps leaders review recommendations against a shared standard. It should improve consistency, not become a room where the loudest manager wins.
Before the meeting: Establish the rules
Define the purpose. Decide whether the meeting is calibrating performance ratings, salary recommendations or both. If the organization combines these steps, make the relationship between rating and pay explicit.
Publish the criteria before managers assess employees. Managers should know the rating definitions, evidence requirements, budget limits and approval process before submitting recommendations. Do not introduce new standards during the meeting.
Prepare a consistent data pack. Include employee role and level, current base salary, salary-range position, performance evidence, proposed rating or increase, relevant changes in responsibilities and any documented exceptions. Restrict access to confidential compensation data.
Train managers on rating definitions. Explain what each rating means and provide examples of evidence. Managers need a shared interpretation of terms such as “meets expectations” and “exceeds expectations.”
During the meeting: Use a repeatable agenda
- Confirm the standard. Revisit the criteria and remind participants that they are assessing evidence, not personal popularity or visibility.
- Review patterns. Compare distributions by team, role and manager. Large differences are prompts for questions, not automatic proof that a manager is wrong.
- Discuss outliers. Ask what evidence supports unusually high or low ratings, proposed increases outside policy or inconsistent outcomes for comparable roles.
- Check context. Consider opportunity, role scope, access to resources, leave, team outcomes and changes in responsibilities. Avoid penalizing employees for factors outside their control.
- Record decisions. Note the recommendation, final decision, reason, approver and any follow-up action.
Questions that improve calibration quality
- What observable outcomes support this rating?
- Were the employee’s goals appropriate for the role and level?
- Are we applying the same standard across teams?
- Is the manager evaluating the full review period or relying on recent events?
- Are similar contributions being described differently depending on the employee or team?
- Does the proposed increase fit the organization’s stated pay policy and budget?
- Is a pay difference explained by consistent, job-related factors?
How can HR reduce bias?
Use evidence-based criteria and train managers to recognize common evaluation errors, such as recency bias, halo effects, similarity bias and unequal access to high-visibility projects. Compare decisions across groups where legally appropriate and protect sensitive personal information.
Calibration itself can introduce bias if participants defer to senior voices, rely on stereotypes or use forced rankings without a sound reason. The goal is not to force every team into the same rating distribution. It is to test whether the same standards are being applied fairly.
CIPD notes that performance-related pay depends heavily on how performance is measured and how managers distribute awards. It also highlights the need to balance objective and subjective measures rather than assuming either one is sufficient on its own.
After the meeting: Close the governance loop
HR should reconcile the final recommendations against the approved budget, check exceptions and complete required approvals. Managers should receive the information they need to communicate outcomes, but not unrestricted access to other employees’ confidential pay details.
Keep a decision log. Record overrides and the reason for them. If the meeting reveals that a rating definition or salary policy is unclear, fix the policy for the next cycle rather than quietly changing the rules mid-cycle.
A simple calibration meeting checklist
| Before | During | After |
|---|---|---|
| Criteria and rating guide shared | Outliers challenged with evidence | Final budget reconciled |
| Managers trained | Comparable cases reviewed | Exceptions approved |
| Data checked and access limited | Bias and consistency discussed | Decisions documented |
| Budget parameters confirmed | Changes and reasons recorded | Managers briefed for communication |
Frequently asked questions
Should HR change a manager’s rating during calibration?
A rating may be changed when the evidence shows that the original assessment did not apply the agreed standard. The reason should be documented and the manager should understand the decision.
Should companies use forced distribution?
Not automatically. Requiring a fixed percentage of employees to fall into each rating category can distort the assessment when team performance does not match the assumed distribution. Use a method that fits the organization’s goals and evidence.
How often should compensation calibration happen?
Many organizations calibrate as part of a formal review or merit cycle. Additional calibration may be useful when teams are reorganized or when rating inconsistency becomes a recurring issue.
Who should attend?
Typically HR or compensation leads, relevant business leaders and managers responsible for the recommendations. Keep attendance limited to people who need the information to make or review decisions.
Make calibration traceable, not theatrical
A meeting can improve consistency, but it cannot replace clear criteria, reliable data and documented approvals. MeritCyc supports structured increment cycles with locked criteria and an audit trail so approved rules and final decisions remain easier to review.