🎯OKRs — Deep
Past the textbook: the OKR patterns that work at scale, the anti-patterns that kill teams, and the senior moves.
OKRs are everywhere and mostly badly implemented. The senior version — the OKRs that actually drive ambitious outcomes — requires nuance that most playbooks skip.
Great OKRs separate aspirational stretch goals from commitments, cascade across the org without being mechanical, and are reviewed weekly with real adjustment. The most common failure: OKRs as set-it-and-forget-it theater.
The senior OKR patterns
1. Separate aspirational from commitment. Aspirational OKRs target 70% achievement; commitments are 100% no-fail. Don't conflate them.
2. Cascade thoughtfully, not mechanically. Team OKRs should ladder to company OKRs, but the team should choose how — not be told. Mechanical cascade kills ownership.
3. Weekly check-ins. Update progress. Discuss what's working. Adjust if a KR is clearly wrong. Quarterly grading is too late.
4. Score honestly. Hit 100% on every KR = sandbagged. Hit 40% = either bad luck or bad planning. The score is information.
5. Carry forward learnings. Each quarter's grading informs the next quarter's KRs. The team learns what's plausibly achievable.
The anti-patterns
- OKRs as performance review. Once OKRs become firing criteria, people sandbag.
- Too many. 5+ OKRs per team is fake commitment.
- Output as KR. "Ship X" isn't a result.
- Set in January, looked at in December. Theater.
- Lift KRs from a template instead of deriving from strategy.
The Q1 reset
January-February of each year is when OKRs are calibrated. Use the December review to honestly assess what worked. Did stretch KRs at 70% mean we set them right, or are we just rationalizing failure?
The team that does this honestly gets sharper every year. The team that doesn't grades themselves favorably and never improves.
Cross-team OKRs
For initiatives that span multiple teams: one shared OKR with explicit role for each team. Document who owns what KR. Run a weekly cross-team sync against the shared OKR.
Without explicit ownership, cross-team OKRs become 'someone else's problem.'
Real-world examples
Google has run OKRs since 1999. Their version distinguishes 'aspirational' OKRs (graded at 0.7 = success) from 'committed' OKRs (graded at 1.0 = expected). The distinction is preserved in the OKR doc itself — anyone reading sees which is which.
Go deeper — recommended reading
Interview questions (1)
Q1What's the difference between aspirational and committed OKRs, and why does it matter?executionsenior▼
Aspirational OKRs are stretch goals — designed to push the team beyond what's clearly achievable. Target grade: 0.7 = success. You'd expect to hit 50-80%.
Committed OKRs are must-do — things the business is depending on. Target grade: 1.0. Missing these is a real problem.
The distinction matters because conflating them causes one of two failure modes:
- All committed. Team sandbags every OKR (sets the bar low so they hit 100%). Ambition disappears.
- All aspirational. Critical operational work doesn't get done because it wasn't 'a big swing.'
Healthy OKRs have both: 1-2 committed (table stakes) and 1-2 aspirational (ambitious bets). The team and leadership both know which is which. Score them differently.