The Two Functions
OKRs do two things when they work. Knowing which two clarifies most of what follows.
Objectives radiate intent
An Objective, working, makes a team's intent legible to everyone around it. A stakeholder reads the Objective and knows what the team is for. A delivery partner sees why the work matters. Someone joining the team orients quickly to the question of what they have been hired to help with.
The function is informational. The Objective is a signal sent outward - into the organisation, across functions, up and down hierarchies. It says: this is what we are committed to. It says it clearly enough that a person who is not in the team can act on it.
When this function is working, the team does not have to explain itself constantly. The artefact does the explaining. Stakeholders pre-coordinate around it. Decisions made elsewhere take it into account. The Objective becomes a shared reference point that reduces the negotiation cost of every adjacent activity.
When this function is failing, the team is constantly having to translate. Stakeholders ask what the team is doing. Adjacent teams duplicate work. The Objective exists but does not radiate - because it is too vague, or too internal, or too political, or because nobody outside the team has been given reason to read it.
Key Results anchor decisions
A Key Result, working, makes a trade-off decidable on evidence. When something new lands on the team's backlog, the KR tells you whether it fits. When something the team is doing has drifted away from the intent, the KR tells you to stop. When two stakeholders disagree about what the team should prioritise, the KR is the third party in the conversation.
The function is operational. The KR is a signal used inward - into the team's planning, prioritisation, refinement, and reporting. It says: this is what we will measure ourselves against, and therefore this is what we will and will not do.
When this function is working, the team has grounds for refusal. Saying no to work that does not fit is not a political act - it is an evidence-based one. The KR carries the political weight that an individual saying no cannot.
When this function is failing, the team has no defensible basis for prioritisation. Every new request has to be argued on its own merits. Refusal becomes a function of who is asking. The KR exists but does not anchor - because it is unmeasurable, or because it is decoupled from any actual decision the team makes, or because it was imposed and the team does not trust it.
Why two functions matter for diagnosis
Most OKR teaching describes the artefact - what an Objective should look like, how many KRs to have, how to score them. Very little teaching names what the artefact is for.
This is a problem because, without naming the functions, there is no way to tell whether an OKR is working. The artefact looks the same whether it is radiating intent or not, whether it is anchoring decisions or not. An organisation can have a complete set of well-formed OKRs that perform neither function. The format guarantees nothing.
Naming the functions gives you a diagnostic. For any OKR, two questions:
- Is this Objective making the team's intent legible to people outside the team? Could a stakeholder read it and pre-coordinate?
- Are these Key Results anchoring decisions inside the team? Has any prioritisation conversation changed because of them?
If both answers are yes, the OKR is doing OKR work. If either is no, the OKR is a wallpaper instrument - formally correct, functionally absent.
How the failure patterns map to function
The four signals from the previous section are all failures of one or both functions.
- Used as a stick breaks both. The Objective stops radiating intent and starts carrying performance commitment. The KRs stop anchoring decisions and start anchoring blame. The artefact is no longer doing its work, but doing the work the management system has demanded instead.
- Used for control breaks intent-radiation. The Objective does not carry the team's intent because the team did not write it. It carries someone else's intent, dressed in the team's voice.
- Set and forgotten breaks anchoring. The KR cannot anchor a decision if nobody is consulting it when decisions are made. The artefact has the right shape and the wrong rhythm.
- Used and ignored breaks anchoring in a worse way. The KR is consulted ceremonially but ignored operationally. The artefact has the right shape and the wrong relationship to the work.
A useful test, when you find an OKR system that is not working: ask which of the two functions has broken. The answer determines which constraint is missing.
Why this is not a recommendation
The two functions are descriptive, not prescriptive. They tell you what OKRs do when they work. They do not tell you that every organisation should adopt OKRs.
Many organisations would be better served by simpler instruments - a written strategy that radiates intent without numerical commitments, a backlog discipline that anchors decisions without quarterly artefacts. OKRs are one way to deliver the two functions. They are not the only way.
The question worth asking is not how do we improve our OKRs but how does our system make intent legible and how does it anchor decisions. If the answer to the second question involves OKRs, this playbook is useful. If it involves something else, the something else has to do the same two jobs.
Last updated 15 May 2026
Licensed under CC BY-SA 4.0.