A practical guide to fairer performance review calibration, including evidence standards, meeting agendas, rating consistency and privacy.
Two employees can do comparable work and receive very different performance ratings if their managers use different expectations. One manager may reserve the highest score for rare achievements; another gives it to anyone who meets their targets. Performance review calibration helps organisations examine those differences before finalising decisions.
Calibration should make evidence and standards more consistent. It should not be a meeting where leaders force a predetermined number of people into each rating category.
Why ratings diverge
Ratings can differ because roles are unclear, objectives changed mid-cycle, managers observe different work, or individual evaluators interpret the same scale in different ways.
Common risks include recency bias, excessive weight on one memorable event, preference for familiar working styles and confusion between effort and outcomes. Managers may also be influenced by visibility: a person who speaks frequently in meetings can appear more productive than a quieter colleague doing equally valuable work.
The solution is not to remove judgement entirely. It is to make judgement traceable to relevant, job-related evidence.
Define the standards before reviewing names
Write down what each rating means, using observable behaviours and results appropriate to the role. An “exceeds expectations” rating should refer to performance above a defined standard, not to how much the manager likes the employee.
Ask whether the criteria are relevant to the work, whether employees knew about them and whether unusual assignments or changes in scope are considered fairly. A rating system cannot be calibrated well if everyone started with a different definition of success.
For a broader foundation, read our performance appraisal system guide, which explains setting objectives and building a review process.
Prepare evidence before the meeting
Each manager should submit a proposed rating and concise evidence. The evidence might include agreed objectives, work samples, quality observations and constructive feedback given during the review period.
Avoid unverified impressions such as “not leadership material” or “always positive.” Ask for specific examples connected to the role. Consider the employee's opportunity to perform: some roles depend heavily on client demand, equipment or contributions from other teams.
Protect employee privacy. Bring only the information needed for review and restrict the discussion to appropriate participants.
Create a consistent calibration agenda
A focused meeting can follow this sequence:
- Confirm the purpose, participants and confidentiality expectations.
- Revisit rating definitions and role-specific standards.
- Identify proposed ratings that appear inconsistent with the evidence.
- Discuss difficult cases using relevant work examples.
- Record the reason for any agreed change.
- Assign follow-up questions where evidence is missing.
- Confirm how decisions and feedback will be communicated.
A facilitator should question assumptions without allowing seniority or volume of speaking to determine the outcome.
Ask better questions during difficult discussions
When a manager proposes an exceptional rating, ask what work demonstrates performance beyond the agreed standard and whether the evidence spans the review period.
When performance is rated poorly, ask whether expectations were clear, whether feedback was provided and whether barriers outside the person's control were considered.
When two employees receive different ratings for comparable results, examine role complexity and context before concluding that the ratings are inconsistent. Calibration is about fair comparison, not mechanical equality.
Avoid forced ranking and unsupported quotas
A calibration meeting should not automatically assume that a fixed percentage of employees must receive each rating. Such a rule can distort results when teams genuinely differ in performance or size.
If the organisation has formal distribution requirements, review them carefully for fairness, legal considerations and their effect on incentives. Make the rules transparent to authorised decision-makers rather than applying hidden expectations.
Document changes without creating a surveillance record
Keep a concise record of the criteria, evidence reviewed, decisions and outstanding actions. Avoid copying unnecessary sensitive material into shared notes. Give managers guidance about what can be communicated to employees and what must remain confidential.
Employees should receive understandable feedback about performance and development, not simply a final number that appears to have changed in a private meeting.
Review patterns after calibration
Look for systematic differences across managers, roles and review cycles. If one manager consistently rates far above or below others, investigate whether standards or coaching differ.
Where lawful and appropriate, HR may review fairness risks across groups, but such analysis must protect privacy and avoid assuming a discrepancy proves discrimination on its own. Seek qualified advice for sensitive or legally regulated assessments.
Use findings to improve manager training, role definitions and evidence collection ahead of the next review. A well-run calibration process should reduce confusion over time rather than require increasingly long meetings.
One-page calibration checklist
Before the meeting:
- Rating definitions and role expectations are documented.
- Managers have provided evidence, not only numbers.
- Review-period context and changes in objectives are recorded.
- Meeting access and confidentiality are agreed.
During the meeting:
- Challenge vague claims and focus on work-relevant examples.
- Use consistent standards while recognising role differences.
- Record decisions and follow-up responsibilities.
After the meeting:
- Communicate constructive, specific feedback.
- Address gaps in manager assessment skills.
- Review whether the process was fair and useful.
Frequently asked questions
What happens in a performance calibration meeting?
Managers and HR compare proposed ratings against common standards and supporting evidence. They discuss inconsistencies, document justified decisions and identify questions requiring more information.
How can HR reduce rating bias?
Use clear criteria, evidence-based discussion, manager training, consistent processes and review of rating patterns. No single meeting can remove every form of bias.
Should ratings always change during calibration?
No. A meeting that validates well-supported ratings may be successful. The goal is confidence in the fairness of the process, not a quota of adjustments.
Build more credible performance reviews
RAST Consultants supports organisations through performance appraisal services and related HR processes. Contact us to discuss your review framework. This is general management guidance and not employment-law advice.