OKRs die at the same rate every quarter: set enthusiastically in week 1, ignored by week 4, panic-updated in week 12. The fix is boringly simple — a weekly 30-min review that costs less than the alternative. This workflow encodes it as three prompts. Copy each week; the answers accumulate context that makes each subsequent week faster.
Status snapshot
Goal: For each objective, honest position vs where you'd need to be to hit target.
Here are my Q[X] OKRs and where each stands as of today: [paste list]. For each objective: (1) rate on-track / at-risk / off-track, (2) name the specific metric or milestone that determines the rating, (3) note if this is the first week it's shown that status. Be blunt — no marketing-speak.
Blockers audit
Goal: For each at-risk / off-track, name the specific thing blocking progress.
For each objective rated at-risk or off-track above: what's specifically blocking progress? Not 'we need more focus' — the concrete blocker (resource, decision, capability, external dependency). One sentence per objective.
One thing this week
Goal: The single most-leveraged action for this week that moves at least one at-risk indicator.
Given the status + blockers above, what's THE ONE THING I do this week — the most-leveraged action that moves at least one at-risk objective? Not a to-do list — one action. Explain why it's the highest-leverage move.
Repeatability, not cleverness.
What lives where when the workflow finishes.
Each week's output feeds the next week's input. Paste last week's status + this week's status in step 1 to see week-over-week movement.
The same shape, different jobs.
- Team version: everyone runs steps 1-2 individually, brings answers to a 30-min team sync where step 3 is decided together.
- Company version: leadership team does the same on company-level objectives; department leaders bring their own review to a monthly rollup.
- Personal-life version: works identically for annual personal goals if you have measurable ones.
Try this workflow once. Then decide if it earns a recurring slot.
The value compounds after the third run, not the first.