HR & performance glossary
What is a calibration session?
Calibration session: A meeting in which managers review and adjust the performance ratings they have proposed so the same standard applies across teams before ratings are final.
Definition
A calibration session is a meeting in which managers, usually facilitated by HR, review the performance ratings they have proposed for their teams and adjust them so the same standard applies across the organisation. Without it, a rating depends as much on who the manager is as on how the employee performed: some managers rate generously, others rarely give a top score, and the difference flows straight into pay and promotion decisions.
Calibration usually happens once managers have drafted their ratings and before those ratings are shared with employees or used for pay. It is sometimes confused with forced ranking, but the purpose is a consistent standard, and a well-run session can move ratings up as well as down. In a smaller organisation it may be a single meeting; in a larger one, a series of sessions by department that roll up to a final review.
How it works in practice
A good session works from evidence rather than advocacy. Each rating is discussed against goal results, competency assessments and feedback, attention goes to the outliers and borderline cases rather than every name, and changes are recorded with a reason. Distribution guidelines can prompt a discussion, but they work best as a check, not a quota.
Preparation does most of the work. HR circulates the proposed ratings with the supporting evidence and a view of how each manager’s ratings are distributed, so the group can see in advance where one team looks unusually generous or harsh. Sessions are usually grouped by level or function, so participants know enough about each other’s people to contribute, and a facilitator keeps the discussion on evidence and on time.
Afterwards, managers need to be able to explain each final rating in their own words. A rating that was changed in calibration and delivered as ‘HR moved it’ undermines both the manager and the process, so every manager should leave the session knowing the reason for each change to their team’s ratings.
A worked example
An illustrative example: the team described is hypothetical, not a customer.
Consider a hypothetical product department with four managers. Before the session, one manager has rated six of eight people as exceeding expectations, while each of the other three has rated one or two people at that level. The facilitator puts that team first on the agenda.
Walking through the evidence, the group finds that three of the six have goal results and competency ratings in line with top ratings elsewhere; the other three met their goals but did not exceed them. Those three are adjusted to meeting expectations, with the reason recorded. In another team, a quietly effective engineer whose manager had rated them as meeting expectations is moved up after peers describe their contribution to a critical release.
The department’s final ratings no longer depend on which manager each person reports to, and every change can be explained to the employee concerned.
Common mistakes
Advocacy over evidence. The manager who argues hardest should not win; the one with the best evidence should.
Discussing everyone. Sessions that try to cover every name run out of time before the difficult cases. Focus on outliers, borderline ratings and people being considered for promotion.
Treating a distribution as a quota. A target curve can prompt the right questions, but forcing ratings to fit it penalises strong teams and flatters weak ones.
Skipping manager preparation. Managers who arrive without having reviewed their own evidence tend to defend their first draft rather than test it.
Not recording changes. If a rating moves and nobody writes down why, the manager cannot explain the outcome to the employee, and the organisation cannot defend it if it is questioned.
How CLEAR Talent supports it
In CLEAR Talent's performance review software, calibration is a stage in the review cycle, after manager review and before the final rating. Live heatmaps and outlier flags surface inconsistent ratings across managers, forced-distribution simulations are available as an optional guardrail, and every change is written to the audit trail. Because the composite rating keeps goal scores, competency ratings and 360° input visible as separate parts, a session can question one without disturbing the others.