Business · Goals & performance
The Best Prompt for Writing OKRs
The best prompt for OKRs keeps objectives, key results and initiatives apart, because mixing them is how most OKR sets go wrong. This prompt drafts a small number of objectives from your priorities, writes key results as measured changes from a baseline to a target, moves project work into a separate initiatives list, and then reviews its own draft against a short quality checklist. Baselines you did not supply are marked to be measured, not guessed.
OKR Drafting and Review Prompt
Drafts objectives and measurable key results from your priorities and current metrics, then audits its own draft for tasks disguised as results.
- Works with
- ChatGPT, Claude, Gemini
- Difficulty
- Intermediate
- Complexity
- Multi-part reply
- Variables
- 6
- Version
- 1.0
Customize this prompt
Fill in what you know. Anything left blank stays as a [bracketed placeholder]. Nothing you type leaves your browser.
6 variables · fields marked * matter most
# Role You are an OKR coach working with a team lead. You care more about a small set of goals people will check weekly than a complete list of everything the team does. # Context - Team or organization: [Team or organization] - Period: [Period] - Priorities from the level above, or the strategy this must support: [Priorities or strategy to support] - Current metrics and baselines: [Current metrics and baselines] - Constraints (headcount, budget, dependencies, things that must not slip): [Constraints] - Maximum number of objectives: [Maximum number of objectives] # Definitions to apply - Objective: a qualitative statement of what will be different by the end of the period. One sentence, no numbers, understandable by someone outside the team. - Key result: a measurable outcome that would prove the objective was met. Written as 'Metric from X to Y'. It describes a change in the world, not work we did. - Initiative: a project or task we believe will move a key result. Initiatives are bets and can be swapped; key results stay fixed. # Task 1. Draft the objectives, up to the maximum. Each must trace to one of the priorities; name which. 2. For each objective write two to four key results. Use my baselines. If I gave no baseline for a metric, write 'Baseline: to be measured' and name how to measure it. Do not make up a starting value or a target built on one. 3. Label each key result Committed (we expect to hit it fully) or Stretch (ambitious; partial achievement is acceptable). 4. Under each objective, list two to four candidate initiatives, kept visibly separate from the key results. 5. Review your own draft and report, per key result, whether it passes these tests: - Outcome, not output: 'launch', 'ship', 'complete', 'hold' and 'deliver' signal an output. Rewrite or move to initiatives. - Within the team's influence during the period. - Measurable with data the team can get without a new project. - Has no obvious way to be hit while harming something else. If it does, add a paired quality or guardrail key result. 6. List what is deliberately left out: work the team will keep doing that does not belong in the OKRs. # Rules - Fewer is better. Do not use the maximum number of objectives unless the priorities require it. - Respect the constraints. If a target looks unrealistic given them, say so and offer a lower alternative. - No vague verbs in key results: improve, enhance, optimize, drive. # Output format For each objective: the objective, a table of key results (key result | baseline | target | type | how measured), then initiatives as bullets. After all objectives: the review findings, then the left-out list.
Use it when you are
- Drafting quarterly OKRs for a team from company-level priorities
- Reviewing an existing OKR set for key results that are task lists
- Helping a new manager or a team adopting OKRs for the first time
- Preparing a first draft to bring to a team goal-setting workshop
Why this prompt works
Definitions are in the prompt, not assumed
People and models use 'OKR' loosely. Spelling out what counts as an objective, a key result and an initiative, including the 'Metric from X to Y' form, gives the model a standard to write against and gives you a standard to hold the draft to when you edit it.
Initiatives get their own place
The usual failure is a key result such as 'launch the new onboarding flow'. Giving project work a legitimate home under each objective means the model does not have to squeeze it into the key results, and the team can see the bets separately from the measures.
No baseline, no invented target
A target only means something relative to where you are. When you have not supplied a starting value, the prompt requires 'Baseline: to be measured' and a measurement method. That is an honest draft, and it tells you what instrumentation work has to happen in the first weeks.
The draft is audited before you see it
Step 5 makes the model test each key result for output wording, team influence, measurability and perverse incentives. The guardrail check matters most: a response-time target with no quality measure beside it invites rushed answers, and the prompt asks for a paired key result when that risk exists.
How to use it
- Paste the priorities exactly as leadership wrote them so the objectives can be traced back.
- Enter current values for the metrics you already track. Do not estimate; a missing baseline is handled by the prompt.
- Set the maximum number of objectives lower than feels comfortable. Three is plenty for most teams.
- Read the review findings in step 5 before the OKRs themselves, and ask for a rewrite of any key result that failed a test.
- Take the draft to the team. Targets and the Committed or Stretch labels should be argued over by the people who will own them.
- Agree how and how often each key result will be measured before the period starts.
AI model compatibility
| Model | Status | Notes |
|---|---|---|
| ChatGPT | Expected to work | If the self-review in step 5 is skipped or brief, ask for it as a separate follow-up message: 'Now run step 5 on every key result and show the result in a table.' |
| Claude | Expected to work | To review OKRs you have already written, paste them inside <existing_okrs> tags and ask the model to begin at step 5. |
| Gemini | Expected to work | No model-specific adjustment needed. Paste the prompt as written. |
“Expected to work” means the prompt uses plain, model-independent structure and has been reviewed for that — not that it has been run on that model. A model is marked Tested only when a test record is published below.
Worked example
Scenario: A support team lead drafting quarterly OKRs from two company priorities. Team and metrics are invented for illustration.
# Role You are an OKR coach working with a team lead. You care more about a small set of goals people will check weekly than a complete list of everything the team does. # Context - Team or organization: Customer support team of eight at a B2B software company - Period: Next quarter - Priorities from the level above, or the strategy this must support: Company priorities: reduce churn among small accounts; prepare for launch in the German market - Current metrics and baselines: Median first response time 6 hours; customer satisfaction 86%; 31% of tickets are how-to questions already covered in the help center - Constraints (headcount, budget, dependencies, things that must not slip): No new hires this quarter; help center migration must finish in the first month - Maximum number of objectives: 2 # Definitions to apply - Objective: a qualitative statement of what will be different by the end of the period. One sentence, no numbers, understandable by someone outside the team. - Key result: a measurable outcome that would prove the objective was met. Written as 'Metric from X to Y'. It describes a change in the world, not work we did. - Initiative: a project or task we believe will move a key result. Initiatives are bets and can be swapped; key results stay fixed. # Task 1. Draft the objectives, up to the maximum. Each must trace to one of the priorities; name which. 2. For each objective write two to four key results. Use my baselines. If I gave no baseline for a metric, write 'Baseline: to be measured' and name how to measure it. Do not make up a starting value or a target built on one. 3. Label each key result Committed (we expect to hit it fully) or Stretch (ambitious; partial achievement is acceptable). 4. Under each objective, list two to four candidate initiatives, kept visibly separate from the key results. 5. Review your own draft and report, per key result, whether it passes these tests: - Outcome, not output: 'launch', 'ship', 'complete', 'hold' and 'deliver' signal an output. Rewrite or move to initiatives. - Within the team's influence during the period. - Measurable with data the team can get without a new project. - Has no obvious way to be hit while harming something else. If it does, add a paired quality or guardrail key result. 6. List what is deliberately left out: work the team will keep doing that does not belong in the OKRs. # Rules - Fewer is better. Do not use the maximum number of objectives unless the priorities require it. - Respect the constraints. If a target looks unrealistic given them, say so and offer a lower alternative. - No vague verbs in key results: improve, enhance, optimize, drive. # Output format For each objective: the objective, a table of key results (key result | baseline | target | type | how measured), then initiatives as bullets. After all objectives: the review findings, then the left-out list.
What a good response should contain
- Two objectives written without numbers, each naming the company priority it supports
- Key results in 'from X to Y' form that start from the 6-hour, 86% and 31% baselines
- A German-market key result with 'Baseline: to be measured' because no German-language data was supplied
- The help center migration listed as an initiative, not as a key result
- A review note pairing the response-time key result with a satisfaction guardrail
This is an illustrative scenario showing how to fill in the prompt and what to look for. It is not a recorded model output. Recorded outputs will be published here once this prompt has a test record.
Testing and editorial review
Status: Untested. Reviewed for structure and clarity by the editorial team. No recorded model test yet.
No test record has been published for this prompt yet. When it is tested, this section will show the model and version, the date, the inputs, the actual output and notes against our seven-point rubric. We do not publish numerical scores.
| Author | TheBestPromptFor Editorial Team — The editorial team writes, reviews and maintains the prompts and guides on TheBestPromptFor. Prompts are drafted with AI assistance and reviewed against published editorial standards. |
|---|---|
| Published | November 15, 2026 |
| Last reviewed | November 15, 2026 |
| Last tested | Not yet tested |
| Version history |
|
Limitations and things to check
- The model cannot judge whether a target is ambitious or easy for your team. Calibrate targets with people who know the work.
- OKRs written by one person and handed down tend to be ignored. Use the output as a discussion draft.
- Organizations apply OKRs differently, for instance on scoring or how stretch goals are treated. Adjust the definitions to match your house rules.
- A key result is only useful if the metric is actually tracked. Confirm the data exists before committing.
Alternative prompts
Productivity
Weekly Planning Prompt
Checks your task list against the hours you really have, picks the few outcomes that matter and lays out a day-by-day plan with a cut list.
- ChatGPT
- Claude
- Gemini
Business
Evidence-Based SWOT Analysis Prompt
Builds a SWOT around a specific decision, ties each point to evidence you supplied, and converts the grid into strategic options.
- ChatGPT
- Claude
- Gemini
Business
Business Plan Draft Prompt
Turns your idea, customer and known numbers into a structured plan draft, with every unverified figure flagged instead of invented.
- ChatGPT
- Claude
- Gemini
Productivity
Structured Decision-Making Prompt
A two-round prompt: the model questions you first, then weighs your options against your criteria and argues against the one you already favor.
- ChatGPT
- Claude
- Gemini
Related resources
Guide
How to Improve AI Responses
A diagnostic approach to weak AI answers: name the fault, trace it to the gap in your prompt, then follow up precisely or rewrite and start again.
Technical guide
Prompt Chaining: Breaking a Task Into Linked Prompts
How to split a large task into a sequence of focused prompts, design the handoffs between them, and debug the chain when a step goes wrong.
Frequently asked questions
What is the difference between a key result and an initiative?
A key result is a measured change you want to see, such as median response time falling from six hours to two. An initiative is work you think will cause that change, such as introducing a triage rota. If the initiative does not move the number, you replace the initiative and keep the key result.
How many OKRs should a team have?
Few enough to remember without looking them up. The prompt caps objectives at four and tells the model to use fewer where it can, with two to four key results each. A long list usually means routine work has been included, which is what the left-out section is for.
Can I use this prompt to check OKRs my team already wrote?
Yes. Paste the existing OKRs under the context section and tell the model to skip the drafting steps and run the review in step 5. You will get a pass or fail against each test and suggested rewrites for key results worded as outputs.
Sources and references
- Prompt engineering guide — OpenAI
- Prompt engineering overview — Anthropic
Was this prompt useful?
No account needed. Do not include personal information.