Two employees do equally strong work. One gets a glowing review and a raise. The other gets "meets expectations" because their manager rates everyone a notch lower than the manager down the hall. Nobody did anything wrong — the managers just calibrate differently in their heads, and nothing forced them to compare notes.
That's the problem calibration exists to fix.
What Calibration Actually Is
Calibration is a structured conversation — usually a meeting, sometimes software-assisted — where managers compare ratings across their teams before reviews are finalized. The goal isn't to force everyone into a bell curve. It's to catch the manager who rates everyone a 4 out of 5 no matter what, sitting next to the manager who reserves a 4 for truly exceptional work.
Without it, a rating means whatever the individual manager thinks it means. With it, a rating means roughly the same thing across the company.
How a Calibration Session Works
- Managers submit draft ratings first. Nobody walks in blind, but nobody's final either.
- Ratings get grouped and discussed. Often by department, sometimes by level, comparing similar roles across managers.
- Managers explain the evidence behind outlier ratings. If one manager's team is rated unusually high or low compared to peers, they walk through specific examples.
- Ratings get adjusted where the evidence doesn't hold up. Not arbitrarily — based on the actual discussion, with the manager still in the room.
The evidence part is where most calibration sessions succeed or fail. A manager arguing from memory loses to a manager who shows up with specific, documented feedback collected over the year.
Do You Actually Need This?
If you have one or two managers and everyone's ratings already get compared informally, probably not — the overhead isn't worth it yet.
Once you're at the point where a handful of managers run their own review cycles somewhat independently, and especially once ratings start feeding into raises or promotions, calibration stops being optional. Inconsistent ratings at that point don't just feel unfair — they quietly become a pay equity and retention problem.
Common Calibration Mistakes
- Doing it once a year, cold. If managers haven't been tracking performance all along, calibration turns into a debate about memory instead of evidence.
- Forcing a curve. Calibration should catch real rating drift, not artificially manufacture "low performers" to hit a distribution.
- Skipping the follow-up. If ratings change in the room, the employee needs to hear the real reasoning — not just a number that moved with no explanation.
Making Calibration Easier
Calibration is only as good as the evidence managers bring into the room. When feedback is collected continuously throughout the year instead of reconstructed from memory right before the meeting, managers walk in with specifics instead of gut feelings — and the whole conversation gets faster and fairer.
If your team is ready to make ratings mean the same thing across every manager, see how WorkStory can help.
FAQ
What is calibration in performance reviews?
A structured process where managers compare ratings across teams before finalizing them, to catch inconsistent rating standards between managers before they turn into unfair raises or promotions.
When should a company start doing calibration?
Once multiple managers are running review cycles semi-independently and ratings start affecting pay or promotion decisions — before that point, informal comparison is usually enough.
What makes calibration sessions go badly?
Managers arguing from memory instead of evidence, forcing ratings into an artificial distribution, and changing ratings without explaining the reasoning back to the employee.
