Performance Calibration

A structured meeting where managers collectively review and normalize employee performance ratings before they're finalized — ensuring consistency across teams and preventing rating inflation, deflation, or bias.

Performance calibration is a group process in which a set of managers — typically all the managers who report to a common leader — meet to review their preliminary performance ratings before those ratings are finalized and communicated to employees. The core purpose is consistency: without calibration, performance rating scales mean different things to different managers. One manager's '4 out of 5' may represent the same actual performance as another manager's '3 out of 5,' and without a shared reference point, employees doing equivalent work receive different ratings with different compensation and career consequences. Calibration sessions force managers to defend their ratings publicly, compare them against their peers' ratings, and adjust where they've been systematically too lenient (inflating) or too harsh (deflating).

A typical calibration session works as follows: each manager pre-populates a shared matrix or spreadsheet with their team's names, proposed ratings, and brief evidence notes. In the session, the facilitator (usually an HR business partner) may display aggregate rating distributions — for example, 'Manager A has 40% of their team rated at the top level; the company guideline is 15-20%.' Managers then discuss outliers: 'You have five people rated Exceeds Expectations — walk us through your top two.' Other managers in the room provide perspective, particularly when they've worked with the employee directly or have context the rating manager lacks. Ratings are adjusted up or down based on the group discussion. The outcome is a calibrated set of ratings that is more defensible, more consistent, and less susceptible to any single manager's idiosyncrasies.

Performance calibration is distinct from both the performance review process (the formal conversation between manager and employee about their performance) and the nine-box talent review (which evaluates potential in addition to performance). Calibration focuses specifically on the current-cycle performance rating — the number or label that will appear on the review form and influence merit increases, bonus payouts, and promotion eligibility. Calibration typically happens after managers have drafted their ratings but before those ratings are shared with employees. In some companies, employees submit self-evaluations that feed into manager ratings before calibration; in others, self-evaluations come after initial manager ratings and before the final review conversation.

The calibration process has well-documented risks alongside its benefits. The benefits: consistency, fairness, and reduced outlier bias. The risks: social dynamics in the room can work against employees whose managers are less senior, less articulate, or less politically connected — a junior manager advocating for a high rating for their best employee may lose the argument to a more senior manager who pushes back without substantive evidence. Recency bias is also amplified in calibration — the most recent quarter's performance tends to dominate discussion because it's fresh, even when the full year's context would tell a different story. And the process can compress ratings toward the middle, inadvertently disadvantaging genuinely outstanding performers whose ratings are negotiated down in the interest of appearing 'calibrated.'

How Calibration Affects Your Rating

Your manager's advocacy matters: in a calibration room, a manager who can cite three specific, concrete examples of your impact is far more persuasive than one who says 'she's great — trust me.' Help your manager help you: give them clear evidence in your self-eval and regular 1-on-1s.
Rating distributions matter: if your manager already has several people at the top rating, adding you may require displacing someone — or accepting that the math doesn't work. Understanding your company's expected distribution helps you know what you're competing against.
Cross-functional visibility helps: calibration participants often have context about employees from cross-team interactions. High visibility on high-stakes projects means your name carries recognition that supports your manager's advocacy.
Recency bias is real: strong Q4 performance tends to be what calibrators remember. Plan important, visible contributions to land in the final quarter of the performance period.
Your manager's standing matters: a respected manager with strong credibility in the room can push back on calibration pressure more effectively than one who is junior or less trusted.

What Managers Do in Calibration

Prepare a pre-read: most calibration sessions require managers to pre-populate ratings with brief evidence notes — specifics, not adjectives.
Defend outliers: any rating significantly above or below the expected distribution will require justification — expect to be asked 'walk us through your Exceeds Expectations ratings.'
Listen and update: calibration is bidirectional — you may learn that a peer has context about one of your employees that changes your view, or you may successfully advocate for an upward adjustment based on evidence you provide.
Watch the math: if the company guideline says 15% of employees receive the top rating and your preliminary ratings have 35% at the top, you'll need to either adjust down or win the argument that your team genuinely outperformed.
Document your reasoning: after calibration, you'll communicate final ratings to employees. Having clean notes from calibration helps you explain the rationale clearly and avoids 'the committee decided' responses that feel dismissive.

Rating Inflation and Deflation

Rating inflation: the systematic tendency for managers to rate employees higher than their actual performance warrants — driven by desire to avoid difficult conversations, protect team morale, or retain employees by signaling appreciation.
Impact of inflation: when calibration reveals one manager has consistently inflated ratings, adjustments in the room reduce those employees' ratings — often without the employees knowing their manager advocated for more.
Rating deflation: less common, but some managers chronically underrate their teams, either due to high standards or as a strategy to create 'headroom' — employees who could be rated higher but are kept at a lower rating to give the manager budget flexibility next cycle.
Compression: calibration sessions sometimes compress all ratings toward the middle, reducing the differentiation between outstanding and merely good performers — a known dysfunction of forced calibration processes.
Recalibration signals: if you've been consistently rated the same level for several cycles despite growing contribution, ask your manager directly whether calibration is affecting your ratings — sometimes the answer is yes.

Example

A manager submits preliminary performance ratings with 5 of her 8 direct reports rated 'Exceeds Expectations.' In the calibration session, the HR business partner notes the company's guideline is 20% at that level — which would be 1-2 employees on an 8-person team, not 5. The manager's skip-level asks for specifics. The manager walks through each person's impact with concrete examples. For two of the five, the evidence is strong: one led a product launch that exceeded revenue targets by 40%, one rebuilt a critical system that reduced outages by 80%. For the other three, the evidence is more general — 'consistently positive,' 'great team player,' 'hits her deadlines.' The calibration group agrees the first two should remain Exceeds Expectations; the other three are adjusted to Meets Expectations. The manager accepts the outcome but commits to building stronger documentation for those three employees next cycle. The two who retain Exceeds ratings receive differentiated merit increases; the three who are adjusted receive standard increases.