The OKR grading and scoring session closes a cycle with an honest number on every key result. Scoring on a 0.0 to 1.0 scale turns a quarter of effort into a clear read on what was achieved, and it only works when the room treats the score as data rather than a personal grade.
Run this at the end of each cycle, once the final metrics are in. It typically precedes the retrospective: grade first to establish what happened, then reflect on why. Keep it separate from planning the next set so scoring stays honest and is not bent to make the next cycle look easier.
Bring the objective owners and team leads, four to ten people. Everyone should arrive with final numbers, not estimates. The tone matters as much as the math: if low scores are punished, future OKRs get sandbagged and the whole system loses its honesty.
Reset the ground rules first, especially that a 0.7 on a genuine stretch goal is a good outcome, not a failure. Score each key result from 0.0 to 1.0 against its target, then roll those up into objective scores. Spend real time on the drivers behind both the wins and the misses, since the explanation is more valuable than the number. Close by capturing the lessons, including which goals were mis-scoped, so the next cycle is sharper. Grading is a learning tool, not a performance review.
Close the cycle with honest numbers. Run it in OrgTP and keep your key result scores and lessons in one place cycle over cycle.
75 minutes total · 5 sections
Stop copying agendas into a doc every week. OrgTP runs your meetings live — scorecard, rocks, issues, and to-dos all in one place, with your AI agents in the room.