---
title: "Performance Calibration"
description: "Performance calibration is when managers compare ratings across teams so the same standard applies to everyone. See how a calibration meeting works."
url: "https://sageo.ai/glossary/performance-calibration"
source: "Sageo (sageo.ai)"
---


# Performance Calibration

Performance calibration is a process where managers meet to compare and adjust employee ratings across teams, so the same standard is applied to everyone. Without it, a "meets expectations" from a tough manager and an "exceeds" from a generous one can describe the same work. Calibration usually happens after managers draft their ratings and before results are shared or tied to pay and promotion.

**Also known as:** calibration meeting, performance review calibration, rating calibration, talent calibration

## Key takeaways

- Calibration checks that ratings mean the same thing in every team.
- It happens after draft ratings and before decisions on pay or promotion.
- Evidence beats opinion: every rating discussed should be backed by examples.
- Record what changed and why, so decisions can be explained later.

## At a glance

- **When:** After draft ratings, before results are shared
- **Who:** Managers of comparable teams, an HR facilitator, a senior leader
- **Output:** Final ratings plus a record of changes
- **Related:** Performance review, 9-box grid, bell curve, talent review

## Why does performance calibration matter?

Managers rate differently. Research by Scullen, Mount and Goff (2000) found that more of the variation in performance ratings came from the person doing the rating than from the employee's actual performance. Calibration reduces that effect. It makes pay and promotion decisions fairer, helps spot bias such as recency or favouritism, and gives leaders a consistent view of talent across the organisation. With pay transparency laws spreading, a clear record of how ratings were set is also becoming a compliance need.

## How does a calibration meeting work?

1. **Agree the standard.** Share the rating definitions and examples of each level before the meeting.
2. **Managers submit draft ratings** with short evidence for each person.
3. **Review the distribution.** Look at how ratings spread by team, level and demographic group.
4. **Discuss the edges.** Spend time on the highest and lowest ratings and anyone near a boundary.
5. **Challenge with evidence.** Ask "what did they do?" rather than "how do you feel about them?"
6. **Record changes** and the reason for each, then share final ratings with managers.

## Calibration vs forced distribution

Calibration compares ratings against a shared standard. Forced distribution (the bell curve) makes managers fit ratings into fixed percentages. You can calibrate without a forced curve, and most modern teams do.

## Worked example

Two engineering managers each rate five people. Manager A rates four as "exceeds"; Manager B rates one. In calibration, the group reviews evidence against the written definition of "exceeds": delivering beyond the role's scope with impact outside the team. Two of Manager A's ratings move to "meets" because the work was strong but within scope. One of Manager B's people moves up after a peer shares that she led a cross-team migration. Each change is logged with its reason.

## Frequently asked questions

### What is a calibration meeting?

A meeting where managers and HR review draft performance ratings together, compare them against shared standards, and agree final ratings.

### Who attends a calibration session?

Managers whose teams are being compared, a facilitator (usually HR or a people partner), and often a senior leader who makes the final call.

### Is calibration the same as stack ranking?

No. Stack ranking ranks people against each other and often forces a fixed share into the lowest group. Calibration checks ratings against a standard; it does not require a quota.

### How long does calibration take?

Plan around 3 to 5 minutes per employee discussed, focusing longer on edge cases. Groups of 30 to 60 people per session are manageable.

### Should employees know about calibration?

Yes. Explaining that ratings are calibrated builds trust that the process is fair, even though individual discussions stay confidential.

### How do you reduce bias in calibration?

Use written rating definitions, require evidence for every rating, review results by demographic group, and rotate who speaks first.

## Sources

- Scullen, Mount and Goff, "Understanding the latent structure of job performance ratings", *Journal of Applied Psychology* (2000)
