| name | goal-setting-okrs |
| description | Use when designing goal-setting frameworks (OKRs, SMART, Locke-Latham) โ covers cadence, level integration, and common failures. |
Goal Setting and OKRs
What the Evidence Says
Edwin Locke and Gary Latham, decades of research: specific, difficult, accepted goals outperform "do your best" goals on simple-to-moderately-complex tasks. For complex creative tasks, goals have nuanced effects โ they can narrow attention and reduce exploration if poorly framed.
Effective goals are:
- Specific: not "improve customer experience" but "reduce P1 ticket response time from 4 hours to 1 hour"
- Difficult: stretch beyond comfortable, not impossible
- Accepted: the person doing the work owns the goal, not just the goal-setter
- Measurable: progress visible
- Time-bounded: with a horizon
OKRs (Objectives and Key Results)
Originated at Intel (Andy Grove); popularized by Google.
Objective: qualitative, aspirational, time-bounded
Key Results: 3โ5 measurable outcomes that, if achieved, mean the objective is achieved
Example:
- O: Become the most trusted partner for our enterprise customers
- KR1: NPS from 32 to 50
- KR2: Renewal rate from 88% to 95%
- KR3: 3 customer references published
OKRs work when:
- Few of them (3โ5 per team max; 1โ3 per IC)
- Quarterly cadence
- Public within the company
- Reviewed regularly (weekly check-ins, quarterly retros)
- Decoupled from compensation (a Google specific principle: aim for 70% achievement; 100% means the goals weren't ambitious)
SMART (and Its Limits)
Specific / Measurable / Achievable / Relevant / Time-bound. Useful as a checklist; not a complete framework.
Limits:
- "Achievable" can soften ambition
- Doesn't address goal acceptance or alignment
- Can encourage measurable-but-trivial goals over hard-but-meaningful ones
Cadence Patterns
- Annual: company-level direction; team strategic priorities
- Quarterly: team OKRs / objectives
- Monthly: progress check-ins
- Weekly: 1:1 progress; blockers; adjust
Levels of Goal-Setting
| Level | Goals |
|---|
| Company | 3โ5 objectives; aligned to strategy |
| Team | 3โ5 OKRs; supports company; team-determined within strategy |
| Individual | 1โ3 OKRs; supports team; IC-determined within team |
Cascading goals work when bidirectional: leadership sets direction; teams interpret to local context; individuals choose how to contribute. Pure top-down cascading produces compliance theater.
OKR / Goal Anti-Patterns
- 15 OKRs per team (= no priorities)
- Compliance theater: copy company OKRs verbatim
- Compensation tied directly (incentivizes sandbagging)
- KRs that aren't outcomes (e.g., "ship feature X" โ output, not outcome)
- Vague KRs ("improve performance")
- Goals set in January for the year, never revisited
- Quarterly retros that celebrate but don't learn
- Personal goals replaced by performance ratings
Linking to Performance
OKRs should NOT be the rating. They are too gameable, too narrow, and too context-dependent.
Better: OKRs as input to performance discussion. Did the person make progress on what mattered? What did they contribute beyond OKRs? What would have made the goals more useful?
Common Failures
- Goal-setting season as theater
- Fighting for whose KR includes what
- "Stretch goals" used punitively
- Goals invented because "we have to set goals," not because they clarify direction
- No follow-through; quarterly retros skipped under deadline pressure
- Senior leaders not setting clear company OKRs, leaving teams to guess
Cross-References
performance-management-systems skill
change-management skill
people-strategy-roadmap skill
Key References
- Locke, E. A., & Latham, G. P. (2002). Building a practically useful theory of goal setting and task motivation.
- Doerr, J. (2018). Measure What Matters.
- Grove, A. (1983). High Output Management.