Rating Distributions: The Honest Conversation Nobody Wants
When every manager rates their team excellent, ratings become meaningless. Calibration forces the honest distribution, and honesty is the point.
Left to themselves, managers rate their own people generously, everyone's a top performer, which sounds lovely and renders your ratings worthless. When every rating is excellent, the rating tells you nothing. Distribution and calibration force the honest conversation, and honesty is the whole point of rating at all.
Decide if you want a distribution guide
Some organisations guide managers toward a rating distribution, not a rigid quota, but a nudge against everyone-gets-top. It's contentious. Decide deliberately whether you want it, because it shapes how much the ratings actually mean.
Use calibration to surface the outliers
The manager who rates everyone brilliant, and the one who rates everyone harshly, both distort. Calibration puts ratings side by side so those patterns show, and the group can correct for the generous and the severe.
Keep it about fairness, not quotas
The goal is fair, comparable ratings, not hitting a number. Configure and run calibration to serve fairness across teams, or it becomes a resented box-ticking exercise that managers game.
Real scenario: a client's ratings were useless, 95% of staff rated in the top two bands, so the ratings couldn't inform pay or promotion at all. We introduced calibration focused on fairness across teams, not quotas. The distribution got honest, and suddenly the ratings meant something for the decisions that depended on them. Meaningful ratings need an honest distribution, and that needs calibration.