Calibration · Performance Rating Calibration Meeting
Performance Calibration
What is calibration?
- A meeting to align rating standards, not to judge people
- Submit three lines of rationale in advance; discuss only borderline cases
- Record the reason for each change so you can explain it in the review conversation
Many people think of calibration as a meeting where someone gets marked down to fit a rating distribution. The real purpose is the opposite: bringing all raters’ standards in line with the same yardstick, so that a generous team leader doesn’t give an A where a strict one would give a B for the same performance. For employees, the biggest unfairness is a rating that depends on which team leader they happened to get, so a calibration meeting should be about the standard, not about individuals.
Half the work of the meeting is done in preparation. Each rater brings the rationale for every rating in three factual lines or fewer, and HR puts up a one-page view of the rating distribution by organization and the change from last year. Rather than walking through everyone, the discussion focuses on people at rating boundaries, people on whom raters disagree, and the top and bottom candidates. Five to ten minutes per person is about right, and the rule is that attendees speak only when they have worked directly with the person.
The facilitator is usually an HR manager or HRBP. Their job is not to reach a conclusion but to keep asking "What facts is that judgment based on?" They bring remarks that drift toward impressions and reputation back to facts, and when a rating changes, they record the reason in one line. Without that record, a team leader has nothing to say later when explaining the changed rating to the employee. Agree up front that what is said in the meeting stays among attendees, and that no one will relay who said what.
Three things tend to go wrong in practice. If one remark from a vocal executive overturns a rating, the meeting becomes a contest for airtime rather than calibration. If the group is fixated on hitting distribution targets, people get marked down in turn with no supporting evidence. And the moment a team leader tells an employee, “It was cut from above,” trust in the whole system collapses. A rating agreed in the meeting must be something the team leader can explain as their own judgment, and any adjustment that can’t be explained should be challenged again in the room.
Inwoo LeeExpert in leadership, communication, and performance management
John KimInstructor John Kim, CEO of Nanum Consulting and author of “B2B Marketing Design”The term comes from an engineering word for adjusting an instrument to a standard. In HR, it came to mean aligning raters to the same benchmark. It has no single originator; the procedure took hold from the rating-adjustment practices of companies that used relative ranking and forced rating distributions.
Consider a hypothetical case. Company A, a manufacturer with 480 employees, has a sales division of 5 teams and 62 people. In the first-round ratings, 55% of Team 1 was rated A or higher versus 15% of Team 4. Last year's calibration meeting ran four hours, yet not a single reason for a grade change was recorded.
- One week before the meeting, the HRBP sends the five team leaders a 'three-line rationale' form and distributes in advance a single table combining each team’s grade distribution and goal attainment rate.
- The first 20 minutes of the meeting are not spent talking about people. The division head and team leaders first agree on a standard sentence: “An A is someone who hit 110% or more of their goal and opened new accounts.”
- Of 62 employees, only the 14 at rating boundaries are put up for discussion. When the Team 1 leader says, “They worked hard,” the HRBP asks back, “Please tell us the facts behind that”.
- Two Team 1 candidates for an A are confirmed at 96% goal attainment and adjusted to B, while one Team 4 candidate for a B is confirmed to have won 3 new accounts and moves up to A.
- For each of the 6 changes, the HRBP writes a one-line reason and sends it to the team leader concerned. The team leader restates that sentence in their own words and enters it in the review conversation prep sheet.
- The meeting shrinks from 4 hours to 2 hours 10 minutes, and the gap between teams in the share of A-or-higher ratings narrows from 40 percentage points to 12. Appeals drop from 11 to 3.