Roadmapping & planning · Goal-setting system
OKRs
Align a team on a meaningful objective and measurable evidence of change.
Start here · 5-minute visual lesson
State the change you want, then define measurable evidence that would prove the change is happening.
Think of it like this: An objective is the destination; Key Results are the instrument readings that prove you are getting closer.
- 1Write the direction
- 2Choose outcome measures
- 3Separate initiatives
- 4Review and adapt
“Launch new onboarding” is called success even when customers behave exactly as before.
The team aims to make the first week productive and measures activation, time to value, and support burden.
Learn by deciding
The launch that is not a result
A team writes “Launch guided onboarding” as a Key Result. What is wrong?
That is an initiative—work the team controls—not evidence that customer behavior changed.
What should replace it?
A baseline-to-target outcome such as seven-day activation from 42% to 58%.
Activation rises but support tickets double. Did the OKR work?
Not safely. Add a guardrail Key Result so local optimization cannot hide transferred harm.
01 · Understand
In plain English
Objectives state a concrete direction; Key Results state verifiable outcomes that show whether the objective is being achieved.
Origin and context
Andy Grove developed the approach at Intel from management-by-objectives roots; John Doerr later named and popularized OKRs, including at Google.
The problem it solves
Plans can become lists of projects without a shared definition of success. OKRs connect intent to evidence and create a review cadence.
The concepts, explained
Learn the vocabulary first. Each concept changes a decision you will make when using the framework.
Objective
A short, meaningful description of the change the team wants to create. It gives direction without listing the work.
Key Result
A measurable outcome that proves whether the objective is becoming real, with a baseline, target, owner, and data source.
Initiatives
Projects and experiments intended to move a Key Result. They are bets, not evidence of success by themselves.
Committed, aspirational, and learning OKRs
Different goal types set different expectations: must-achieve delivery, ambitious stretch, or deliberate uncertainty reduction.
Check-ins and grading
Regular reviews connect current evidence to decisions. End-of-cycle grading matters only when it produces learning for the next cycle.
02 · Deep dive
Go beyond the definition
Follow the reasoning inside OKRs, then challenge it before you use it.
1 of 4mental model
The mental model
What changes in the way you see the problem
Objective
significant direction
Key Result
measurable outcome
Initiatives
work intended to move KRs
Committed, aspirational, and learning OKRs
Use this concept to examine the problem from a distinct OKRs perspective.
Check-in and grading cadence
Use this concept to examine the problem from a distinct OKRs perspective.
Reasoning trace
Watch each step earn the next one
Objective: “Make the first week feel reliably productive for new workspace admins.” KR1: raise seven-day activation from 42% to 58%. KR2: reduce median time to first shared workflow from 3.2 days to 1.5. KR3: keep setup-related support contacts below 0.25 per new workspace. Projects such as templates and import guidance remain initiatives.
- 01Frame
Start from strategy and the change that matters now.
Decision question: If every initiative ships but KRs do not move, what will we do? - 02Interrogate
Write one clear, time-bounded objective without metric soup.
Decision question: Which KR could be gamed? - 03Interrogate
Choose two to five outcome KRs with baselines and targets.
Decision question: If every initiative ships but KRs do not move, what will we do? - 04Interrogate
Test whether teams can influence each KR.
Decision question: Which KR could be gamed? - 05Interrogate
List initiatives separately; do not disguise them as results.
Decision question: If every initiative ships but KRs do not move, what will we do? - 06Interrogate
Assign owners and instrument measures before the cycle.
Decision question: Which KR could be gamed? - 07Interrogate
Review leading evidence regularly and change initiatives when needed.
Decision question: If every initiative ships but KRs do not move, what will we do? - 08Decide
Grade, explain learning, and decide what carries forward.
Decision question: Which KR could be gamed?
Failure-mode lab
Which mistake is most likely on your team?
Choose one to turn it into a control for your next workshop.
Context adaptation
The canonical version is only a starting point
A startup can run one company and one product OKR; an enterprise may align without rigidly cascading. Platform teams pair adoption outcomes with ecosystem health guardrails.
Signals it fits
- Strategy needs a focused near-term expression
- Cross-functional teams need shared outcomes
- Measures can be instrumented and influenced
Signals to switch tools
- Leadership already mandates a fixed output plan
- Metrics are unavailable and cannot be created
- OKRs will be tied mechanically to individual compensation
Now use the framework. The workbench below turns your own context into a first decision draft.
Work my problem03 · Apply
How to use OKRs
- 01
Start from strategy and the change that matters now.
- 02
Write one clear, time-bounded objective without metric soup.
- 03
Choose two to five outcome KRs with baselines and targets.
- 04
Test whether teams can influence each KR.
- 05
List initiatives separately; do not disguise them as results.
- 06
Assign owners and instrument measures before the cycle.
- 07
Review leading evidence regularly and change initiatives when needed.
- 08
Grade, explain learning, and decide what carries forward.
04 · See it work
A worked example
Objective: “Make the first week feel reliably productive for new workspace admins.” KR1: raise seven-day activation from 42% to 58%. KR2: reduce median time to first shared workflow from 3.2 days to 1.5. KR3: keep setup-related support contacts below 0.25 per new workspace. Projects such as templates and import guidance remain initiatives.
Business case study
Documented case study
What Matters documents the development of OKRs at Intel and their later introduction to Google. The Atlas uses that history only to establish provenance; the worked product OKR is fictional.
Across company types
A startup can run one company and one product OKR; an enterprise may align without rigidly cascading. Platform teams pair adoption outcomes with ecosystem health guardrails.
05 · Work your problem
Turn the framework into a result
Explore the mechanics
Copy the working template
Objective: Why now: KR1 baseline → target / owner / source: KR2 baseline → target / owner / source: KR3 guardrail baseline → target: Initiatives: Weekly check-in: End-of-cycle learning:
06 · Use judgment
Where it helps—and where it breaks
Use when
- Strategy needs a focused near-term expression
- Cross-functional teams need shared outcomes
- Measures can be instrumented and influenced
Do not use when
- Leadership already mandates a fixed output plan
- Metrics are unavailable and cannot be created
- OKRs will be tied mechanically to individual compensation
Advantages
- Creates focus and transparency
- Separates outcomes from initiatives
- Supports learning through check-ins
Limitations & criticisms
- Poor KRs distort behavior
- Cascades can become bureaucracy
- Ambition without strategy creates noise
Common failure modes
- Writing “launch feature” as a KR
- Too many objectives
- Scoring without narrative learning
Often confused with
SMART tests the quality of an individual goal; OKRs are an operating system linking direction, results, and review cadence.
Practical checklist
Reflection questions
- If every initiative ships but KRs do not move, what will we do?
- Which KR could be gamed?
07 · Check
Five-question knowledge check
Key takeaways
- Align a team on a meaningful objective and measurable evidence of change.
- Creates focus and transparency
- Poor KRs distort behavior
08 · Go deeper
Sources and further reading
- What is an OKR? Definition and ExamplesRyan Panchadsaram · What Matters
Objectives, measurable key results, history at Intel, cadence, and common errors.
- Objectives and Key Results (OKRs)Atlassian · Atlassian Team Playbook
Independent operating guide for outcome-oriented Key Results, check-ins, scoring, and quality safeguards.