OKR Writing
Parent: Writing and Documentation · researched 2026-05-29T19:11:00.052Z· 4 sources · 0 concepts · skill okr-writing
OKRs (Objectives and Key Results) are a goal-setting framework. The form is deceptively simple — one inspirational Objective, three to five measurable Key Results — but most OKRs in the wild are broke
Overview
- OKRs (Objectives and Key Results) are a goal-setting framework. The form is deceptively simple - one inspirational Objective, three to five measurable Key Results - but most OKRs in the wild are broken. They are project plans wearing OKR clothing. [source]
- The Objective is qualitative, inspirational, time-bound. [source]
- Each Key Result is measurable - a numerator over a denominator, by a date. [source]
- The KR measures an outcome (value delivered), not an output (work done) and never an input (effort spent). [source]
2. The measurable-KR test: numerator, denominator, deadline
- Every KR must answer: [source]
- Numerator - what is being counted? [source]
- Denominator - relative to what? [source]
- Deadline - by when? [source]
- Fails: "Improve onboarding." (no numerator, no denominator, no deadline) [source]
- Fails: "Reach 1000 users." (no denominator, no deadline) [source]
- Passes: "Grow week-1 activation rate from 28% to 45% by 2026-09-30." [source]
3. OKR vs KPI — different jobs
- KPI (Key Performance Indicator) - a steady-state health metric you watch all the time. KPIs run forever. [source]
- OKR - a change goal for a bounded period. Says "we are choosing to push this number from X to Y this quarter." [source]
- A KPI can become a KR if you decide to push it - but the KR must specify the from-to delta and the deadline. [source]
4. Input → Output → Outcome hierarchy
- Input - effort, headcount, money spent. ("hired 5 engineers") - never a KR. [source]
- Output - work produced, artifacts shipped. ("shipped the new dashboard") - almost never a KR. [source]
- Outcome - change in the world, value delivered. ("dashboard adoption reached 60% of paying teams") - this is what KRs measure. [source]
- Felipe Castro's "so what?" test: read the KR and ask "so what?" If the answer is another metric (the outcome), the original was an output. [source]
5. Ambitious-but-not-impossible: the 70% rule
- For aspirational / stretch OKRs: [source]
- Pick a target where your confidence of hitting 100% is about 5 out of 10 (50/50 at draft time). [source]
- Score around 0.7 on average at the end of the quarter is healthy. [source]
- Consistently scoring 1.0 means you sandbagged. Scoring < 0.4 means you set fantasy targets. [source]
- For committed OKRs (operational must-haves - SLAs, compliance deadlines), the target is 1.0 and anything less is a problem. [source]
6. The rollup pattern across org levels
- Company sets 3–5 Objectives for the quarter. [source]
- Each team picks ~3 Objectives that contribute to the company set. [source]
- An individual contributor may have 1–2 personal Objectives that map to a team Objective. [source]
- Key insight: roughly half of OKRs should be set bottom-up. Pure top-down OKRs kill engagement. [source]
7. Mid-quarter check-in: traffic-light language
8. End-of-cycle scoring (0.0 to 1.0)
9. Common bad-OKR patterns
- "Ship feature X by date Y" - that is a task, not a KR. The KR is the outcome the feature is supposed to produce. [source]
- "Do our normal job well" - operational baselines are KPIs, not OKRs. [source]
- Activity counts - "publish 12 blog posts" measures effort, not outcome. [source]
- Sandbagged targets - KRs you are 95% confident in. [source]
- Too many KRs - more than 5 per Objective is a wishlist. [source]
- Set-and-forget - OKRs scored only at quarter end. [source]
References
Children
- No children recorded.