leadershipproject-management
OKR Writing
Write OKRs (Objectives and Key Results) with measurable outcomes, scoring rubrics, and alignment checks across team and company levels.
okrobjectiveskey-resultsgoal-settingalignment
Works well with agents
Works well with skills
$ npx skills add The-AI-Directory-Company/(…) --skill okr-writingSKILL.md
Markdown
| 1 | |
| 2 | # OKR Writing |
| 3 | |
| 4 | ## Before you start |
| 5 | |
| 6 | Gather the following from the user: |
| 7 | |
| 8 | 1. **What level are these OKRs for?** (Company, department, team, individual) |
| 9 | 2. **What time period?** (Quarterly is standard; annual for company-level only) |
| 10 | 3. **What are the company or department OKRs?** (Team OKRs must align upward) |
| 11 | 4. **What was the outcome of last quarter's OKRs?** (Scores, lessons, carryovers) |
| 12 | 5. **How many objectives?** (3-5 per level is the practical maximum) |
| 13 | |
| 14 | If the user wants to write 8+ objectives, push back: "OKRs are about focus, not coverage. If everything is a priority, nothing is. Choose the 3-5 things that would make this quarter a success if you achieved nothing else." |
| 15 | |
| 16 | ## OKR structure |
| 17 | |
| 18 | Each OKR set follows this format: |
| 19 | |
| 20 | ``` |
| 21 | OBJECTIVE: [Qualitative, inspiring, time-bound statement of what you want to achieve] |
| 22 | |
| 23 | KR1: [Quantitative metric] from [baseline] to [target] |
| 24 | KR2: [Quantitative metric] from [baseline] to [target] |
| 25 | KR3: [Quantitative metric] from [baseline] to [target] |
| 26 | ``` |
| 27 | |
| 28 | ### Objective rules |
| 29 | - Qualitative and motivating — describes a desired end state |
| 30 | - Ambitious but not impossible — should feel uncomfortable at 70% confidence |
| 31 | - Time-bound (implicitly by the OKR cycle, e.g., Q2 2025) |
| 32 | - No metrics in the objective — that is what Key Results are for |
| 33 | - Starts with a verb: "Establish," "Accelerate," "Reduce," "Build" |
| 34 | |
| 35 | ### Key Result rules |
| 36 | - Quantitative and measurable — a number that can be verified |
| 37 | - States a baseline ("from") and target ("to") |
| 38 | - 2-5 Key Results per Objective |
| 39 | - Measures outcomes, not tasks — "Ship feature X" is a task, "Increase activation rate from 30% to 45%" is a Key Result |
| 40 | - Each KR is independently valuable — achieving 2 of 3 should still matter |
| 41 | |
| 42 | ## OKR template |
| 43 | |
| 44 | ``` |
| 45 | ## Q2 2025 OKRs — [Team Name] |
| 46 | |
| 47 | ### Objective 1: [Verb + desired end state] |
| 48 | Alignment: Supports [Company/Dept Objective name or number] |
| 49 | |
| 50 | | KR# | Key Result | Baseline | Target | Owner | |
| 51 | |-----|-------------------------------------------------|----------|--------|----------| |
| 52 | | 1.1 | [Metric description] | [X] | [Y] | @person | |
| 53 | | 1.2 | [Metric description] | [X] | [Y] | @person | |
| 54 | | 1.3 | [Metric description] | [X] | [Y] | @person | |
| 55 | |
| 56 | ### Objective 2: [Verb + desired end state] |
| 57 | Alignment: Supports [Company/Dept Objective name or number] |
| 58 | |
| 59 | | KR# | Key Result | Baseline | Target | Owner | |
| 60 | |-----|-------------------------------------------------|----------|--------|----------| |
| 61 | | 2.1 | [Metric description] | [X] | [Y] | @person | |
| 62 | | 2.2 | [Metric description] | [X] | [Y] | @person | |
| 63 | | 2.3 | [Metric description] | [X] | [Y] | @person | |
| 64 | ``` |
| 65 | |
| 66 | ## Scoring rubric |
| 67 | |
| 68 | Score each Key Result at the end of the cycle using this scale: |
| 69 | |
| 70 | ``` |
| 71 | | Score | Meaning | |
| 72 | |-----------|-------------------------------------------------------| |
| 73 | | 0.0 | No progress | |
| 74 | | 0.1 - 0.3 | Some progress but fell significantly short | |
| 75 | | 0.4 - 0.6 | Made meaningful progress — partial achievement | |
| 76 | | 0.7 | Delivered — this is the expected "success" score | |
| 77 | | 0.8 - 0.9 | Exceeded expectations | |
| 78 | | 1.0 | Fully achieved or exceeded — may indicate sandbagging | |
| 79 | ``` |
| 80 | |
| 81 | A healthy average OKR score across a team is 0.6-0.7. If the team consistently scores 0.9+, the OKRs are not ambitious enough. |
| 82 | |
| 83 | ## Alignment check |
| 84 | |
| 85 | Every team OKR must connect to a company or department objective. Use this format to verify alignment. |
| 86 | |
| 87 | ``` |
| 88 | Company Objective: "Become the market leader in developer tools" |
| 89 | └─ Dept Objective: "Accelerate product-led growth" |
| 90 | └─ Team Objective: "Improve new user activation experience" |
| 91 | └─ KR: Increase Day-7 activation rate from 30% to 45% |
| 92 | ``` |
| 93 | |
| 94 | If a team objective does not trace up to a company objective, either: |
| 95 | 1. The team is working on something the company has not prioritized — escalate the misalignment |
| 96 | 2. The company OKRs are missing a priority — propose adding it |
| 97 | |
| 98 | ## Mid-cycle check-in template |
| 99 | |
| 100 | Review OKRs at the halfway point (week 6 of a 12-week quarter). |
| 101 | |
| 102 | ``` |
| 103 | | KR# | Key Result | Baseline | Current | Target | Status | Action Needed | |
| 104 | |------|------------------------|----------|---------|--------|-------------|---------------------| |
| 105 | | 1.1 | [Metric] | 100 | 130 | 200 | On track | None | |
| 106 | | 1.2 | [Metric] | 5% | 5.5% | 10% | At risk | Reassign resources | |
| 107 | | 1.3 | [Metric] | 0 | 0 | 3 | Off track | Scope discussion | |
| 108 | ``` |
| 109 | |
| 110 | Status definitions: |
| 111 | - **On track**: Current trajectory reaches the target |
| 112 | - **At risk**: Behind pace but recoverable with intervention |
| 113 | - **Off track**: Will not reach target without significant change |
| 114 | |
| 115 | ## Quality checklist |
| 116 | |
| 117 | Before finalizing OKRs, verify: |
| 118 | |
| 119 | - [ ] Each objective is qualitative — no numbers in the objective statement |
| 120 | - [ ] Each key result has a numeric baseline and target |
| 121 | - [ ] 3-5 objectives maximum per level |
| 122 | - [ ] 2-5 key results per objective |
| 123 | - [ ] Every team OKR traces to a company or department objective |
| 124 | - [ ] Key results measure outcomes, not outputs or tasks |
| 125 | - [ ] At least one KR per objective would be uncomfortable to commit to (stretch) |
| 126 | - [ ] Each KR has a single owner |
| 127 | |
| 128 | ## Common mistakes |
| 129 | |
| 130 | - **Key results that are tasks.** "Launch the new dashboard" is a task. "Increase weekly active dashboard users from 200 to 500" is a key result. OKRs measure outcomes, not deliverables. |
| 131 | - **Sandbagging targets.** If every KR scores 1.0, the team set easy targets. A 0.7 average means the OKRs were properly ambitious. Normalize this expectation with the team. |
| 132 | - **No baseline.** "Improve NPS to 50" is meaningless without knowing the current NPS. Every KR needs a "from X" starting point, even if it's zero. |
| 133 | - **Too many objectives.** Five objectives with four KRs each means 20 things to track. That is not focus — that is a wish list. Ruthlessly cut to 3-4 objectives. |
| 134 | - **OKRs written in isolation.** If engineering, product, and marketing each write OKRs without coordinating, you get misaligned priorities. Run an alignment session before finalizing. |
| 135 |