How to set OKRs that actually work

Vision and strategy tell you where you are going and why. OKRs are how you take the first serious bite. Done well, they translate direction into measurable focus. Done badly, they become a task list with better branding.

A surprising number of companies know, broadly, what they want, yet still fail at the next step: turning that intent into a small number of measurable priorities that people can actually execute against.

That is where OKRs matter.

McKinsey's organisational-health research argues that healthy organisations translate vision and strategy into actionable and measurable objectives that are clearly articulated and shared at all levels. Its 2024 performance-management research adds that 72% of employees cite goal setting as a strong motivator, especially when goals are measurable and clearly linked to company priorities.

That is the real job of OKRs.

Not to create more language. Not to create more reporting. Not to make work look tidy on a slide. To make strategy executable.

“Vision is the destination. Strategy is the route. OKRs are the first serious bite.”

Vision is the destination. OKRs are the first bite.

This is the distinction many companies muddle.

Vision defines the destination: the what and the why. Strategy defines the current best route: the how. OKRs define the next meaningful stretch of movement: what progress should look like now.

That sequencing matters.

If vision is vague, OKRs become arbitrary. If strategy is weak, OKRs become performative. If OKRs are absent, vision and strategy remain intellectually attractive and operationally inert.

Google's re:Work guide is useful precisely because it treats OKRs as a way to communicate strategy and how success will be measured. McKinsey makes the same point from a broader organisational angle: strategic clarity only matters when it is turned into measurable objectives people can actually use.

Good OKRs focus the organisation on outcomes, not activity

This is the most common failure mode.

Teams write OKRs that are really project plans:

• Launch feature X
• Run campaign Y
• Complete migration Z

Those may all be important pieces of work. They are not, by themselves, strong key results.

Google's guidance is blunt on this. Key results should describe outcomes, not activities. If the KR contains words such as “consult”, “help”, “analyse” or “participate”, it is probably describing work rather than impact. Google also says OKRs are not a checklist and should not become a master task list of everything a team plans to do in the quarter.

That distinction matters because activity is easier to complete than value is to create.

A team can finish the work and still miss the point. A feature can ship and still fail to move the business. A campaign can run and still fail to create demand. A migration can complete and still fail to improve speed, quality or customer experience. Good OKRs keep the team honest by measuring whether the work mattered.

“A task completed is not the same thing as value created.”

Revenue is a useful proxy for value, but it is rarely enough on its own

Revenue matters. It is one of the clearest lagging proxies for value creation a company has.

But it is not sufficient as the only lens.

By the time revenue moves, many other things have already happened first: customer behaviour changed, adoption shifted, conversion improved, retention strengthened, trust increased, or some friction in the experience was reduced. Good OKRs usually recognise that chain.

That is why the best goal-setting systems do not rely on one financial number alone. McKinsey's 2024 research found employees are most motivated when goals are measurable and clearly linked to company priorities, and when those goals include a mix of individual and team-level outcomes. Google's OKR guide similarly emphasises measurable milestones that provide credible evidence of completion and direct progress toward the objective.

In practice, that means a good quarterly objective may have key results such as:

• A commercial result
• A customer-behaviour result
• A quality or speed result
• A trust or retention result

Revenue is the signal everyone sees. The stronger teams also track the leading indicators that explain it.

“Revenue is a useful proxy for value. It is rarely enough on its own.”

Fewer OKRs usually means better OKRs

Another mistake is trying to pour the whole roadmap into the OKR framework.

Google warns against this directly. It recommends three to five objectives with roughly three key results each, and notes that more than that can lead to overextended teams and a diffusion of effort. For team-level planning, Google also asks whether there are more than three priorities, which is a useful practical test.

This is one of the reasons most OKRs feel weak: they try to carry too much.

Good OKRs require choice. They force the leadership team to say:

• These things matter most now
• These are the measurable signs of progress
• And these other things, while still real, are not the quarter's defining bets

That is where the power comes from. Not from the framework itself, but from the quality of the prioritisation behind it.

The best OKRs are specific, challenging and jointly owned

There is a reason the goal-setting literature keeps returning to the same point: vague goals do not motivate or guide behaviour as well as specific, demanding ones.

Google's re:Work guide explicitly notes that research shows more specific goals can result in higher performance and goal attainment. Its team-effectiveness guide adds that goals should be specific, challenging and attainable. McKinsey's 2024 research reinforces the same principle from an employee perspective: measurable goals linked to company priorities are the ones people find most motivating.

That gives you a very practical standard. A good objective should be:

• Directional
• Ambitious enough to matter
• Clear enough that an outsider could tell whether it had been achieved

A good key result should be:

• Measurable
• Outcome-oriented
• Directly relevant to the objective
• Few enough that the team can actually focus

And the process matters as much as the wording. McKinsey found employees were more motivated and saw the process as fairer when they were involved in setting the goals and when goals were updated throughout the year to reflect changing priorities. Google similarly recommends a mix of top-down and bottom-up suggestions.

That is a much better model than executive-only goal writing followed by organisational compliance.

A simple way to set them properly

If I were setting OKRs for a quarter, I would do it in this order.

First, confirm the vision and the current strategy. If those are unclear, stop there. Writing OKRs on top of blurred direction usually creates false precision.

Second, identify the one to three priorities that genuinely matter this quarter. Not everything important. The things that most need to move now.

Third, write objectives as end states, not tasks.

Fourth, define around three key results per objective, making sure they describe measurable outcomes rather than activity.

Fifth, pressure-test the set:

• Do these KRs really prove progress?
• Are they connected to strategy?
• Are they too easy?
• Are they too numerous?
• Are they all lagging indicators with no leading signal?
• Will the team know what to stop in order to pursue them?

Sixth, make them public. Google's re:Work guide says OKRs should be public within an organisation so everyone understands what matters and how success will be measured. It also says team OKRs should connect clearly to at least one organisational OKR.

That point is more important than many leaders realise. Public goals reduce the amount of private interpretation happening inside the company.

“OKRs should not be a checklist. They should be a visible statement of what impact matters now.”

Communicate the OKRs well — especially when you miss them

This is one of the most underrated leadership disciplines.

Most organisations are willing to communicate goals when things are on track. Far fewer communicate missed goals with the same clarity. That is a mistake.

McKinsey's 2025 work on transformation argues that performance transparency — clear visibility into goals, metrics and progress — is what most directly enables disciplined execution and continuous improvement, yet only 4% of the transformation plans it analysed explicitly prioritised it. Gallup's trust research points in the same direction: trust grows when leaders communicate clearly, explain where the company is going and how it will get there, and highlight progress along the way. Gallup also found that when employees strongly agree they have opportunities to provide honest feedback about organisational changes, they are 7.4 times as likely to have confidence in leadership's ability to manage emerging challenges.

That is why missed OKRs should be communicated well, not hidden. If you miss:

• Say that you missed
• Explain what moved and what did not
• Explain what you learnt
• Explain what changes next quarter
• Explain what remains strategically true

Gallup's guidance on communicating employee survey results is useful here as a more general management principle: when organisations are transparent about findings and next steps, it creates a culture of trust. It also stresses being timely and transparent, explaining what happens next, and closing the loop with updates on progress. The same logic applies cleanly to OKRs.

There is a particular kind of trust that forms when leaders can say, plainly: we aimed here, we got this far, here is what we learnt, and here is how we are adjusting. That kind of trust is much stronger than the brittle trust built on pretending everything is on course.

“Communicate missed OKRs well. That is where a more durable kind of trust is built.”

Review quarterly, update intelligently

OKRs should not be embalmed for a year.

McKinsey found employees are more motivated when goals are updated throughout the year to reflect evolving team and company priorities. Google's OKR guidance is explicitly built around the upcoming quarter. That combination is sensible: set a clear quarterly frame, then adjust intelligently as evidence changes.

This matters even more now because AI is increasing the speed at which context can change. The answer is not to rewrite the destination every month. It is to keep the destination steady and let the route adjust as new research, market movement or operational learning demands it.

So the discipline should look like this:

• Set OKRs quarterly
• Inspect progress regularly
• Update tactics continuously
• Update OKRs only when the underlying priority genuinely changes
• And explain the change clearly when it does

That is how you get responsiveness without chaos.

A simple test for whether your OKRs are any good

A good OKR set should pass a few very unromantic tests.

• Can a new joiner understand what matters this quarter by reading them?
• Can a team member use them to decide what not to do?
• Do the key results measure impact rather than effort?
• Is the set small enough to create focus?
• Can leadership explain the connection between these OKRs and the wider strategy?
• And if the quarter ends badly, will the organisation still trust the leadership team more because of how clearly the result was communicated?

If the answer to those questions is no, the problem is probably not the framework.

It is the quality of the thinking behind it.

Because good OKRs are not paperwork.

They are the first serious translation of vision and strategy into something people can act on together.