OKR Best Practices for High-Performing Teams
Most teams that try OKRs do not see meaningful results in their first year. Not because the framework is broken, but because how you run it matters far more than which framework you pick.
Next quarterly planning is two weeks away. The team opens the spreadsheet from last cycle and stares at it. Fifteen objectives. Most scored between 0.3 and 0.5. Nobody can quite explain why. I have walked into this exact room numerous times, and the conversation always goes the same way: was the methodology wrong, or did we just not follow through?
In almost every case, the answer is the latter. I have come to think of OKRs less as a goal-setting tool and more as a discipline. The framework itself is simple. Andy Grove wrote the original version on the back of a napkin at Intel. What is hard is the practice of applying it consistently for long enough that it changes how a team thinks.
The research backs this up. Quantive's implementation studies suggest that roughly seven out of ten organisations do not see meaningful improvement from OKRs in their first year. And yet the OKR Impact Report 2022 from Mooncamp, looking at companies that had stuck with the framework longer, found that 83% of them reported a clear positive impact. The methodology does not change between those two numbers. What changes is whether the team gives up at the messy stage or pushes through it.
The most striking data point I have come across is from a Sears Holding study covering roughly twenty thousand employees. Teams that applied OKRs in every quarter saw sales per hour grow by 8.5%. Teams that ran them only intermittently grew by about 3%. Same company, same tools, same training. The only variable was discipline.
The Sears Holding study of 20,000 employees. The framework is identical in both groups. Only the consistency of application changes.
Over the years I have settled on five practices that, when applied together, tend to separate teams that get something out of OKRs from teams that quietly drop the framework after two cycles. None of them are clever. All of them are unforgiving in the way that real disciplines are.
Write fewer OKRs than you think you need
The most common failure mode I see is overload. A team of nine walks out of planning with seven objectives, four key results each, and twelve weeks to deliver them. That is twenty-eight commitments. By week three, most of them are quietly forgotten. By the retrospective, the team feels like they failed at everything.
Andy Grove had a rule about this at Intel. If a leader could not articulate the top three priorities for the quarter without a document in front of them, they did not understand their business well enough. I use a slightly cruder version when I work with teams. After they finish drafting their OKRs, I ask each person to list them from memory. If they cannot, there are too many. Goals you cannot remember cannot guide decisions on a Tuesday afternoon.
Most OKR practitioners settle on three to five objectives per team with two to four key results each. That is the upper end. I usually push teams toward the lower end.
Common mistake: Writing OKRs for everything, including routine operations.
"Maintaining 99.9% uptime" is not a strategic objective. It is an SLA.
Tracking such "OKRs" dilutes the focused work that is supposed to be there.
The fix: OKRs must only cover work that requires focused change. Business-as-usual belongs on an operational dashboard.
Key Results must pass the "stranger" test
By the third cycle, most teams have learned the obvious lesson: do not write key results you cannot measure. The trap that catches them after that is more subtle. They write key results that are measurable but meaningless.
The test I use is quick, straight-forward and very easy to do. Imagine a reasonably intelligent person who has no context about your business looking at the key result on the last day of the quarter. Can they say "yes, this was achieved" or "no, it was not" without asking a single follow-up question? No interpretation, no nuance, just yes or no. If they cannot, the key result is not strong enough.
Weak: "Make user onboarding better."
Strong: "Reduce median adoption time from 15 to 5 days, measured as the interval between release and a working customer install."
Weak: "Launch a new reporting feature."
Strong: "Ship a self-service report builder in the Q2 release with 30% adoption within one month of launch."
The other distinction worth getting right is between leading and lagging indicators. Lagging indicators are the outcomes you ultimately care about: revenue, retention, NPS. Leading indicators are the early signals that tell you whether you are on track to hit them: demos booked, qualified pipeline, time-to-onboard. Teams that only track lagging indicators find out too late. Teams that only track leading indicators end up optimising for activity rather than results. The mature practice is to combine the two.
Leading indicators give you an early signal. Lagging indicators confirm the final outcome. Both belong in your key results.
Check in every week, not once a quarter
If you set goals at the start of the quarter and only revisit them at the end, you have built in twelve weeks of silence. By the time the problem becomes visible, it is too late to do anything meaningful about it. The Mooncamp study put this rather precisely.
Successful companies show 28% higher communication intensity and run OKR check-ins more frequently than the rest. OKR Impact Report 2022, Mooncamp (n=40 companies)
The mechanic I recommend is a fifteen-minute weekly check-in per team. Not a status meeting. Not a long review. Just three questions, asked in order.
What is your current confidence level? A simple traffic light: green if on track, yellow if at risk, red if off track. "It's complicated" and "It depends" are not valid colours.
What changed since last week? Not what was done. What changed. A metric, a risk, an assumption. If nothing changed, the team is either coasting or not paying attention.
What single action would accelerate progress this week? One thing. This is what keeps the check-in from collapsing into a to-do list.
John Doerr calls the complement to OKRs the CFR framework: Conversations, Feedback, Recognition. OKRs tell you what to do. CFRs tell you how to talk about it. A dashboard without honest conversation is just measurement. It is not management.
The pattern across our client work mirrors the published research. Teams that check in weekly hit their key results at more than three times the rate of teams that only review quarterly.
Cascade alignment, not control
The classic top-down approach goes like this. The CEO sets goals. The VPs translate them. The directors break them down. By the time it reaches the individual contributor, the goal has been through four layers of interpretation and barely resembles what the CEO originally meant. It is corporate telephone, with the same quality of output.
I see three failures in pure top-down OKRs. They strip teams of autonomy, which kills the engagement OKRs are supposed to create. They produce alignment that looks neat on a slide and falls apart on contact with reality. And they delay the start of every cycle by two or three weeks while everyone waits for the cascade.
The pattern that works in mature organisations is what I call the 60/40 rule. Roughly sixty percent of OKRs originate bottom-up, from the teams who know their work best. The other forty percent come top-down, representing the strategic priorities that genuinely need coordination across the organisation. The goal is that every team can answer one question without hesitation: how does our work contribute to where the company is going?
Structured autonomy. The teams own most of the goal-setting, while leadership reserves the right to set strategic direction.
Separate OKRs from compensation
This one I have also seen very often - bonus or sales commision tied to goals achievement. Linking OKR results directly to bonuses or promotions fundamentally undermines the purpose of the framework. Breaking this pattern of working with OKRs causes the most resistance.
The problem is what I think of as the sandbagging spiral. In Q1, the team sets an ambitious goal, hits 80% of it, and watches their bonus get cut. In Q2, they set a conservative goal, hit 105% of it, and collect a full bonus. The company has just trained its best people to aim low. Within a few cycles, every goal is sandbagged. Growth slows. Leadership wonders why their ambitious framework is producing pedestrian results.
Google understood this from day one. Their internal standard is that an OKR score of 0.6 to 0.7 is ideal. It signals that the goal was ambitious enough. A consistent 1.0 means the team set the bar too low. But this philosophy only works if scores are decoupled from compensation. The moment a 0.7 costs someone their bonus, the entire system collapses into theatre.
What works instead is evaluating people on a richer picture:
Quality of contribution. What real impact did the work have, regardless of the numerical score? A team at 0.6 on an ambitious goal may have contributed more than a team at 1.0 on a trivial one.
Peer recognition. Systematic feedback from colleagues catches contributions that managers miss.
Growth and learning. New skills, challenges taken on, deliberate work outside the comfort zone.
CFR data over time. The conversations, feedback, and recognition accumulated across a quarter give a much richer picture than any single score.
The result is in the discipline, not the tool
Each of these five practices - writing fewer OKRs, key results that pass the "stranger" test, weekly check-ins, two-way alignment, scores separated from compensation - is useful on its own. The real power only shows up when they reinforce each other.
The arc I see with most clients is the same. The first quarter is messy. Half the OKRs end up reworded mid-cycle and the team feels uncertain. The second quarter is noticeably better. By the third quarter, something quiet happens: people start looking forward to planning. The goals become the operational rhythm of the team, not an administrative ritual imposed from above.
The framework is simple. The practices are concrete. The hard part, as always, is the discipline to apply them long enough that they stop feeling like practices and start feeling like the way your team works.
Ivaylo Ivanov works with CoTransition's clients on goal management, agile leadership, and operating cadence. He has coached teams through OKR adoption in fintech, product, and engineering organisations across Bulgaria and the wider region.