Why most OKR rollouts don’t survive year two (and what I do differently)
Most product businesses don’t have an execution problem. They have a clarity problem. The two look identical from the inside, and OKR is what surfaces the difference.
I’ve spent a decade in product leadership, then the last several years coaching 70+ product-led businesses through OKR rollouts. Clients have included Domino’s, MLC Life Insurance, Inteleos, Scalapay, ELMO, Altro APAC, BAE Systems, SEEK, Carsales, Foundr, Greenpeace, Pepperstone, and Findex. Some of those rollouts were spectacular. A few were painful. All of them taught me something I’m going to share in this guide.
I’m going to spill the beans on what actually happens inside an OKR rollout. The good. The bad. The ugly. Where I cite specific Key Result numbers, I’ll keep the client anonymous unless those numbers are already in the public domain. The patterns are real either way.
By the end of this guide you’ll have the five steps that determine whether your first OKR cycle works or quietly dies before the quarter is out, the prompts I use with Claude and our own AI coach Telos, and the exact Weekly Impact format my clients run.
The five steps, if you’re skimming:
- Set your OKR strategy before you write a single goal.
- Identify the one critical problem you’re actually solving this cycle.
- Write a real OKR. Not a task list with metrics on it.
- Track weekly using a Weekly Impact format your team will actually run.
- Review honestly, then reset for the next cycle.
If you want the packaged shape this fits into, we call it The 90-Day OKR Quickstart Blueprint. A quarter is long enough to achieve something real, short enough to stay focused. Three phases. The first 30 days are assess and strategy alignment. Around day 20 we start co-authoring goals in workshops. From day 30 to day 90 you run the cycle, and weekly confidence scoring keeps it honest. The steps below are the playbook. The Blueprint is how we run it for clients.
Each step has a story behind it. Let me walk you through them.
What is OKR? The 30-second definition before I get into the messy stuff
OKR stands for Objectives and Key Results. It’s a goal-setting method.
The Objective is qualitative. It’s your battle cry for the quarter. Where are we going, and why does it matter? I tell teams to write the Objective like a newspaper headline. Punchy. Specific. Direction-forward.
The Key Results are measurable. They tell you, in numbers, whether you’re getting there. Each one is a shift from a current value to a target. “From 42% to 60%.” Not “improve the thing.”
That’s the framework. The hard part is everything that surrounds it, and that’s what this guide is about.
Related reading: What is OKR? A simple guide with examples.
The honest truth about why OKR programmes fail
I’ve seen this pattern often enough to call it.
Around three quarters of enterprise OKR programmes are reduced or quietly discontinued by year two. The framework almost never fails. The conditions around the framework do.
Here are the four ways I’ve watched it die.
Leadership signs up to the idea but doesn’t model it. The CEO wants their teams to “do OKR” without setting their own goals. The cynicism arrives within a month, and the rest of the org reads the signal.
Teams confuse outputs with outcomes. They write “launch the new portal” as a Key Result, hit it on time, and discover the metric they were trying to move didn’t move at all. By cycle two, they’ve lost confidence in the system. By cycle three, they’re back to project plans.
The weekly check-in never starts, or it starts and drifts. By week eight, half the team can’t remember what their goals were. The cadence is the whole game, and the cadence is the first thing people skip when they’re busy.
Nobody owns the practice internally. The external coach goes home, the cadence stops, and within two quarters the system looks like a graveyard of half-updated spreadsheets.
I’ve watched all four play out at well-funded companies with smart leaders. The five steps below are the antidote.
Step 1: Set your OKR strategy before you write a single goal
Most teams open a spreadsheet on day one and start writing goals. That’s the first mistake.
Before any team writes anything, I make them answer five questions. Together they form your OKR strategy. Skip them and you’re running on hope.
What cadence are you running?
The most common setup I run with clients is annual company-level direction broken into quarterly team-level goals. The annual gives you direction. The quarterly gives you action. Teams check in weekly, leadership reviews monthly.
Some teams run quarterly only, with no annual layer. That works too, especially in the first year while you’re still learning to write decent Key Results. Pick one rhythm and hold it for at least two cycles before you change anything. Constant cadence changes confuse everyone.
Who owns the OKR process internally?
If nobody owns it, it dies. Full stop.
The owner sets the calendar, runs the setting workshops, makes sure the weekly check-ins happen, and holds people accountable when they stop updating. The right person is usually a Chief of Staff, a Head of Product, a senior operations leader, or in a smaller business, the CEO directly.
Name them before you start. Give them air cover. If their role is informal, it gets ignored.
How will you train people?
This is the most underestimated step in the entire process.
I’ve run OKR workshops for senior leadership teams at large organisations and watched experienced executives confidently write task lists with metrics on them and call them Key Results. Every time. Goal-writing is a craft. Nobody is born good at it.
Skip the thirty-minute explainer video. Run a real session where people draft OKRs, share them with the room, and get direct feedback. Then run another session a month later when the goals are live and the cracks have started to show.
Related reading: OKR training and courses.
Will you use a tool?
For your first cycle, use a spreadsheet. Seriously.
The tool debate is one of the most reliable ways teams procrastinate. Pick a shared Google Sheet, run the first cycle, learn what you actually need to track, then evaluate tools based on what you experienced. Most teams that buy a tool first spend three months configuring it to do something they never end up needing.
Once you have one clean cycle under your belt, you’ll know exactly what to look for.
Who are your OKR champions?
Most guides skip this part. It’s one of the most important things I do with clients.
An OKR champion is someone embedded in a team whose job is to keep OKR alive in their part of the business. They’re not the coach. They’re not the owner. They’re the person who nudges colleagues to score their Key Results before the weekly check-in, who flags when a goal is no longer relevant, who notices when the energy has dropped and surfaces it before the cycle stalls.
I saw this most clearly at BAE Systems. Across multiple business units, the teams with active champions had measurably higher engagement, sharper goals, and tighter weekly updates than the teams without them. The gap was bigger than any other variable I tracked, and it compounded each quarter.
Here’s a sticky one I learned the hard way. With one certification client, the internal champion tried to facilitate three teams’ OKRs simultaneously on top of his existing role, with no backup. The process broke. Not because he was bad at it. Because no single person can carry that load. Train more champions than you think you need, and pair them up so they can cover each other.
Champions don’t need formal authority. They need genuine interest in making the system work and a clear mandate from leadership that the time they spend is valued.
Step 2: Identify the one problem that actually matters this cycle
Here’s the question almost nobody asks before writing OKRs. What problem are we solving?
Not “what goals should we set.” Not “what do we want to achieve.” What specific problem, right now, is in the way of where the business needs to go?
This sounds obvious. It isn’t. I’ve watched executive teams spend two hours writing OKRs without ever agreeing on the problem they were solving. The goals looked great on paper. They measured nothing that mattered.
This is also the conversation most exec teams avoid. The one where they have to publicly admit they haven’t agreed what the business is actually trying to do this quarter. Once that conversation happens honestly, the OKRs almost write themselves.
How to find your one critical problem
Start with strategy. What does the business need to be true in twelve months that isn’t true today? Now narrow. What’s the single biggest obstacle between where you are and where you need to be?
That obstacle is your problem.
When I started with Inteleos, the global healthcare credentialing body, they had 31 objectives and 81 Key Results across the organisation. It looked impressive. It was causing analysis paralysis. Leaders couldn’t tell you which goals mattered most. Teams duplicated work without knowing.
Over five years, we refined them down to 15 objectives and 41 Key Results. The reduction wasn’t the win. The clarity was. Their CIO put it as “evidence of really a lot more precision.” Four of their five long-term Impact goals are now tracking to near-complete achievement well ahead of schedule.
Same pattern at smaller companies. A defence-sector client I’m currently coaching had five Key Results across one OKR and was debating whether to drop to four. The right answer wasn’t a fixed number. It was: do each of these measure a different dimension of the problem, with a clear owner? Two clearly did. Three were overlapping. We collapsed those three into two sharper ones. The cycle ran cleaner.
The one-problem test
When you’ve identified your critical problem, run this test on it.
If you made no progress on anything else this quarter but solved this one problem, would it have been a good quarter?
If the answer is no, you haven’t found the real problem yet. Keep digging.
If the answer is yes, you’re ready to write your OKR.
The wishlist trap
The most common failure I see at this stage is what I call the wishlist. The leadership team lists every priority they care about, then tries to turn each one into an OKR.
That’s not strategy. That’s a backlog.
If everything is a priority, nothing is. The point of OKR is to force the choice. Choose one. Park the rest. Come back next cycle.
Related reading: The Executive Prioritisation Matrix.
Step 3: Write a real OKR
Most OKRs I review aren’t OKRs. They’re task lists with metrics taped to them.
“Launch the new onboarding flow” is not a Key Result. It’s a deliverable. “Complete the customer research” is not a Key Result. It’s a project milestone. “Ship the dashboard” is not a Key Result. It’s a Tuesday.
A real Key Result measures a change. A shift in behaviour. A shift in a number. A shift in an outcome that matters to a customer or the business. It does not measure whether you delivered something. It measures whether what you delivered worked.
The anatomy of a strong Objective
Your Objective should be qualitative, clear, and inspiring enough that your team actually wants to achieve it. Write it like a newspaper headline.
Weak: “Improve customer experience.” Better: “Make it easy for new customers to hit real value in their first week.”
The second has direction. You can tell what you’re trying to change. The first could mean almost anything.
The anatomy of a strong Key Result
Every Key Result needs three things. A metric. A starting point. A target. The “from X to Y” format is non-negotiable for me.
Weak: “Increase conversion rate.” Strong: “Increase checkout completion rate from 42% to 58%.”
Use leading indicators where you can. Revenue is a lagging indicator. Activation, repeat usage, qualified leads, time-to-first-value: those are leading. Track what, if it moves, will pull the business outcome along behind it.
Each Key Result needs one accountable owner. Not a committee. Not “the team.” One name.
The good: a FinTech company
When I first sat with a FinTech company, they brought what they thought were goals. They were really KPI lists with ambition attached. Monthly deposits. Active client counts. Trading volume. All lagging. All output-shaped. None of them told the team what to do differently.
We rebuilt their OKR around the customer behaviour they were trying to change. Over the next year they hit all three of their revised Key Results: deposits, active clients, and trading volume all moved meaningfully against their starting points. The following quarter they raised their own targets because they’d built confidence in the system.
The goals didn’t change dramatically in subject. The shift was in how they thought about what they were measuring, and why.
The good: a healthcare services client
A diagnostic imaging client was buried in call centre volume. Their OKR for one quarter was built around a single problem. Too many patients calling in for things they should be able to self-serve.
Key Result: Reduce general inquiry and appointment-change call volume from 31% to 28% of total calls.
They hit a 3% reduction. And along the way, the weekly check-in surfaced an opportunity nobody had planned for. Putting a cancellation form inside the hold queue. Usage jumped from around 30 forms per day to a peak of 106. Not because of a strategy meeting. Because the weekly check-in gave the team space to notice what was working and act on it.
The bad: when an OKR is really a project
For balance, here’s one I keep seeing. From a recent US engagement, the team wrote three OKRs. The first was a genuine outcome (reducing support volume). The second was an Auth0 migration, treated as an OKR even though it was a fixed-scope project. The third was a vendor rollout, treated the same way.
By week six, the project-shaped OKRs were dragging the cycle off course. The team kept reporting on tasks completed, not outcomes shifted. Mid-cycle, we formally reverted one of them back to project tracking and let the OKR sit on the genuine outcome work. The team was relieved. They’d felt the friction for weeks. Nobody had wanted to name it.
The lesson: not all important work belongs in an OKR. If the work is fixed-scope, fixed-outcome delivery, run it as a project. Save OKRs for the change you’re actually trying to drive.
Related reading: How to make effective Objectives and Key Results.
The ugly: director-level resistance
One of the hardest things to manage in an OKR rollout is the layer of the org chart where authority has historically come down rather than out. Directors who’ve spent twenty years being told what to do, then translating it into tasks for their teams, often experience OKR setting as exposing. They have to articulate their own priorities. They have to commit publicly to a number. They have to admit they don’t know how to.
I’ve worked with a US certification body where the director layer openly resisted setting their own goals. The phrase that came up was “I’m not sure how to ask.” The fear underneath it was simpler. If I write a number, and I miss it, what happens? They’d been taught their whole career that visible failure was career-limiting. OKR was asking them to volunteer for that exposure.
The fix isn’t a workshop. It’s leadership behaviour. The CEO or COO has to set their own OKR first, miss the target publicly, talk about what they learned, and not lose their job. After three quarters of that, the directors start writing real goals. Before then, they’ll hedge, water down their Key Results, and look for ways to make the target the thing they were going to do anyway.
Watch what happens to the goal-setting energy in the room when the most senior person says “I’m going to set mine first, and I’m going to make it bigger than I’m comfortable with.” The whole conversation changes.
Using AI in OKR setting: the diagnostic prompts and our coach Telos
I now bake AI into our OKR setting process at almost every step. Not to replace thinking. To pressure-test it.
The single most useful thing AI does in this space is catch the most common mistake. People confuse delivery with impact.
The diagnostic prompt that saves a workshop
Drop your draft OKR into Claude, ChatGPT, or your tool of choice with this prompt:
“Is each of these Key Results measuring an output or an outcome? If a Key Result is measuring an output, suggest an outcome-based alternative in the format ‘from X to Y’ using a leading indicator where possible.”
The AI will catch task-shaped Key Results immediately. It won’t get every nuance right. But it’ll surface 80% of the common errors in 30 seconds. That’s faster than a coach and cheap enough to run before every setting session.
Meet Telos: the AI OKR coach we built
We’ve built a more specialised version of this called Telos. Telos is our AI OKR coach. Personality: a seasoned coach who’s made every mistake in the OKR book and learned exactly what it takes to create an incredible OKR. Empathetic to the learning process. Won’t let bad OKRs slide.
What makes Telos useful is how it structures the conversation. It opens by asking whether you’re trying to maintain current performance (which is a health metric, not an OKR) or drive meaningful change (which is). It separates the two. Then it asks whether your goal is transformational, material, or minor. Most teams overstate the first category and underrate the second. Material improvements compound. They’re not consolation prizes.
Telos calibrates your Key Results against ten quality standards. Leading indicators over lagging. Shifts from X to Y. One OKR per team by default. Human language, not corporate language. No task-shaped Key Results. Real metrics the team can actually track. No reworded Objectives masquerading as Key Results. No milestone goals. Two to four Key Results, not ten. One accountable owner per KR.
You can use Telos as a thinking partner before your goal-setting workshop, as a diagnostic check after, or as a mid-cycle calibrator when the energy on a Key Result starts to drift. It works alongside a coach. It doesn’t replace one. The good OKR conversations still happen in the room. Telos just makes sure you arrive at the room having already cleared the obvious mistakes.
What AI is not for in OKR setting
Be careful here.
AI won’t tell you what problem to solve. It doesn’t know your strategy, your customers, or the political dynamics in your leadership team. If you ask it to write your OKR from scratch, you’ll get goals that sound plausible and mean nothing.
AI also won’t replace the cultural shift. The biggest reason rollouts fail is not bad goal-writing. It’s leadership not modelling the practice. No prompt fixes that.
Use AI for pressure-testing, calibration, and the boring work of cleaning up language. Keep humans in charge of the thinking.
Related reading: How to use AI effectively in your OKR setting.
Step 4: Track weekly with the Weekly Impact format
Setting the OKR is 10% of the value. The other 90% sits in the weekly check-in you probably don’t run.
Most teams set goals, file them in a tool, and check on them at the end of the quarter. By then it’s too late to act. OKR is built for course-correction, not performance review. You need to see the signal early enough to do something about it.
The Weekly Impact format
Our Weekly Impact check-in runs against three questions per Key Result.
What’s the current score, expressed as a confidence level on a 0 to 1 scale or as a traffic light?
What happened this week that moved the score, up or down?
What’s the plan for next week, and what do you need from the rest of the team to get there?
That’s the whole format. It doesn’t need to be long. Fifteen minutes for a focused team. Thirty if you have multiple Key Results or real blockers to talk through.
The discipline is the scoring. Every Key Result gets scored every week, even if nothing has changed. The act of scoring forces you to confront reality rather than assume things are fine. “Same as last week” is a valid answer once. Three weeks in a row, it’s a signal.
Why confidence scores beat raw metrics
This is one technique that consistently makes a difference for my clients.
The confidence score is the OKR owner’s rating of how likely they are to hit the target by the end of the quarter. Not where they are now. Where they think they’ll land.
A team can be at 50% of their target in week six and be completely on track. They can also be at 80% and know the next four weeks are going to be brutal. The raw metric doesn’t tell you which is which. The confidence score does.
At Altro APAC, Tim Attard’s leadership team described the combination of weekly check-ins and confidence scores as “game changing” for how they tracked goals. Their team effectiveness scores moved up 67% over a single cycle. The OKR itself wasn’t dramatically different. The cadence was.
What champions do during the cycle
The champion network earns its keep here.
Champions aren’t there to chase people. They’re there to read the room. When a team stops updating their OKRs, it almost always means one of three things. The goal is no longer relevant. The person reporting doesn’t feel safe reporting a low score. Or the process has slid into a checkbox exercise.
A useful signal: if a confidence score hasn’t moved in three consecutive weeks, something is wrong. Either the team isn’t tracking properly, or they’ve quietly given up. Either way, it needs a conversation. Champions spot this early and flag it before it becomes a problem.
The good: the mid-cycle pivot you only get with weekly tracking
A client in the industrial sector had set an OKR around a new onboarding process. By week three of the quarter, the weekly check-in surfaced that the new flow was generating more support tickets, not fewer. Their confidence score dropped from 0.7 to 0.4 over two weeks.
Because they caught it early, they had time to redesign. They rebuilt the most confusing step, retrained the front-line team, and shipped the change in week five. By week ten, the confidence score was back to 0.8 and the metric was tracking again.
If they’d been on a quarterly review cadence, they would have lost the entire cycle.
The bad: the cadence that didn’t run
The flip side. A client in financial services launched OKR with a big workshop, set ambitious goals, and then the executive sponsor stopped showing up to the weekly check-in by week four. Within three weeks of that, attendance had halved. By week eight, the meeting had been “moved” three times and not actually run twice. The quarter ended with most Key Results unscored. The team’s verdict on OKR was that “it didn’t really work for us.”
The framework hadn’t failed. The sponsor had. The lesson, every time: the cadence is not an “if we have time” meeting. It is the operating rhythm of the system. Skip it and you don’t have OKR. You have a spreadsheet.
Related reading: The OKR Confidence Scoring template.
Step 5: Review honestly at the end and reset properly
The end-of-quarter review is the most undervalued part of the OKR process. Most teams do a quick score, declare victory or failure, and move straight to setting the next cycle’s goals. That isn’t a review. That’s a formality.
A real review answers four questions.
What did we achieve?
Score each Key Result honestly. 70 to 80% is healthy. 100% consistently means you weren’t aiming high enough. Below 30% means something was structurally wrong, either with the goal or with the execution.
I see this misread often. Teams that hit 80% feel like they failed. They didn’t. OKR is designed to be ambitious. The 70 to 80% you actually hit is almost always more than you would have achieved without the goal at all. That is the win.
What did we learn?
This is the question most teams skip. What do you now know that you didn’t know at the start of the quarter?
One client, a certification body, discovered three weeks into a cycle that their biggest support problem wasn’t what they thought. The login and password issues they were trying to fix were partly masking a deeper problem in their registration process. Because they were tracking weekly, they caught the signal and reshaped the work. They came close to their support-volume moonshot for the quarter. At the end-of-quarter review, the learning shaped their next cycle’s goal. They raised their own moonshot meaningfully higher because they now understood the problem well enough to be bolder.
The first cycle’s achievement mattered. The second cycle’s learning mattered more.
What should we carry forward?
Not every Key Result that underperforms should be retired. Some of them reflect a problem that’s still live. The question is whether the goal was right, or just the approach.
If the problem is still real and you didn’t solve it, keep the goal and revise the approach. If the goal turned out to be the wrong goal, retire it cleanly and write down why. The discipline of writing down why prevents the same mistake from creeping back in three quarters later.
What do we need to improve in the process itself?
This is the meta-question. Most teams skip it.
Did the weekly check-in actually run every week? Did the champions stay active? Did the Objective still feel relevant in week ten? Were the Key Results measuring what we expected them to measure? Did anybody hide bad news?
OKR is a learning system. Every cycle should make you better at running the next one. If your process is identical at the end of cycle four as it was at the start of cycle one, something is off.
The focus dividend: Domino’s
A story I keep coming back to. When I started working with Domino’s, they were running a lot of goals across a lot of countries. After we worked through their setting and review process, they simplified to a small number of crucial quarterly priorities. One quarter after that, they reported 33% global sales growth across 12 countries. That isn’t the OKR doing the heavy lifting on its own. It’s focus, combined with a tight review-and-reset rhythm, combined with leadership behaviour that modelled the practice.
Matt Kershaw, their Global Head of People Development, described the engagement as “an absolute standout.” The actual mechanic was unglamorous. Pick fewer goals. Track them weekly. Review honestly. Reset.
Related reading: The Mid-Cycle Review Checklist.
The ugly stories I see every quarter
I promised the good, the bad, and the ugly. Here are the patterns I see most often that destroy programmes. Anonymised but real. If any of these sound familiar, you’re not alone.
The leader who treats OKR as something the team does
I see this pattern at almost every kickoff. The CEO or COO commissions an OKR rollout, attends the workshop, and then never sets their own goals. The leadership team follows their cue. Within two quarters, the only people running OKR are middle managers who can’t get traction with their executives.
The fix: leaders set their own OKR first, in public, and own a number. If you can’t get your CEO to do this, don’t launch.
The dashboard nobody trusts
A regulated-industry client I worked with spent six months building a dashboard tied to their OKRs. The numbers on the dashboard never matched the numbers people reported in the weekly check-in, because the data definitions kept being adjusted by people who wanted different stories told. By month four, leadership had stopped trusting any of the figures.
The fix: agree the metric definitions before you agree the targets. If you can’t define how it’s measured, you can’t write a Key Result against it yet.
The team that hides bad news
I once worked with a team that stopped reporting amber or red status because the executive sponsor had publicly criticised someone for reporting a poor confidence score. Within a month, every team was reporting green. Within a quarter, the entire programme was unrecoverable.
The fix: never tie OKR scores to performance reviews or compensation. Reward honest scoring. Praise the team that flags an amber three weeks early. Make it expensive to hide.
The internal champion who burns out
I see this almost everywhere. The internal champion is the most committed person on the team. They run the cadence. They chase the updates. They wave the flag. And then their workload becomes unsustainable, especially if they’re trying to support more than two teams.
The fix: train more champions than you think you need. Pair them. Rotate the load. Treat champion work as a real part of someone’s role, not a tax on their evenings.
The cycle that gets hijacked by a board solution
A pattern at organisations with active boards. The board, well-meaning, picks a solution they’ve heard about elsewhere (“we should be doing AI avatars”) and tells leadership to set an OKR around delivering it. The OKR ends up measuring the delivery of a predetermined solution rather than the problem it was supposed to solve.
The fix: when a board hands you a solution, treat it as a hypothesis. Define the problem first, then write the OKR around the problem, then test whether the proposed solution moves the metric. If your job feels political, run the test small and let the data win the argument.
A note on directing AI inside your OKR process
I want to bring this back to AI for a moment, because I see teams misuse it often.
There are three good places to use AI in your OKR process.
First, at the front end, as a diagnostic on draft goals. The “is this an output or an outcome” prompt above is the quickest possible quality check. Run it before every workshop.
Second, at the metric-options stage. When a team is stuck on how to measure a problem, AI is excellent at generating five or six possible angles. You won’t use them verbatim. You’ll use them as starting points for a sharper conversation.
Third, at the summary stage. After each weekly check-in, AI can turn the meeting notes into a concise status update for leadership that takes 30 seconds instead of 20 minutes.
The places to be careful are anywhere you’re tempted to outsource thinking. Strategy. Problem definition. The choice of which problem to solve. The judgement on what “ambitious enough” means in your context. None of that belongs to the AI. All of that belongs to you and your team.
Telos is built around this distinction. It coaches you on the craft. It doesn’t make the strategic choice for you. If a teammate brings you a perfectly polished Telos-generated OKR and tells you it’s done, that’s a signal they haven’t done the harder work yet.
Getting started: the minimum viable setup
If you want to run your first cycle without overengineering it, here’s the minimum.
Before the cycle: agree your cadence (quarterly is fine), name an owner, identify two or three champions per team, run a two-hour OKR setting workshop. The output is one OKR per team, written to the “from X to Y” standard, with one accountable owner per Key Result.
During the cycle: run a fifteen-minute Weekly Impact check-in every week without exception. Score each Key Result. Note what changed. Flag anything amber or red. Make the meeting non-skippable.
At the end of the cycle: run a one-hour review against the four questions in Step 5. Capture the learning in writing. Then reset.
That’s the system. Everything else is refinement on top.
Frequently asked questions about getting started with OKR
How long does it take to set up an OKR programme?
The first cycle takes the most upfront work. Allow two to three weeks from the decision to launch to the first weekly check-in. That includes strategy alignment, leadership training, the goal-setting workshop, and getting your tracking system in place. Subsequent cycles run in a week or less.
At ELMO, we had 160 people on OKRs within six weeks of starting. At Scalapay, CEO Simone Mancini described it this way: “within a couple of weeks, everybody had meaningful OKRs.” Speed of adoption depends on leadership commitment more than anything else.
Do we need an OKR tool to get started?
No. A shared spreadsheet is fine for the first one to two cycles. Once you have a clean cycle under your belt and you understand what you need to track, evaluate tools based on your actual requirements. Most teams over-invest in tooling before they understand the process.
Should OKR be tied to performance reviews or compensation?
No. This is one of the most reliable ways to kill an OKR programme. When goals are linked to pay, people set safe goals and hit them easily. Ambition disappears. Bad news gets hidden. Keep OKR separate from performance management. It is a learning and alignment tool.
How many OKRs should a team have?
One to start. Seriously. One Objective and two to four Key Results. Add more only once the team has run at least two cycles cleanly and can genuinely manage the complexity.
What is the difference between OKR and KPI?
A KPI is a health metric. It tells you whether the business is operating normally. An OKR is a change target. It tells you what you’re trying to shift this quarter. Most businesses need both. OKR organises the change. KPIs tell you the rest of the business is holding steady while you make it.
How do I get leadership buy-in for OKR?
This is the question underneath every failed rollout. The short answer: don’t launch without it. If your exec team isn’t willing to set their own OKR, own a number, and show up to the weekly check-in, the rest of the organisation will read the signal and disengage.
The longer answer: start with the problem leadership already cares about. Don’t pitch OKR as a framework. Pitch it as the answer to a problem they’re losing sleep over. Then show them how the system works through one cycle of evidence. One real cycle is more convincing than any deck.
Can AI help us set better OKRs?
Yes, in a specific way. AI is genuinely useful for pressure-testing whether a Key Result measures an output or an outcome, generating alternative metric options when a team is stuck, drafting and reshaping an Objective for clarity, and turning weekly check-in notes into a clean status update. AI won’t tell you what problem to solve, and it won’t fix a leadership team that won’t model the practice. Our own coach Telos is built to handle the calibration so humans can focus on the strategic call.
What is a moonshot target and when should I set one?
A moonshot is an aspirational target you wouldn’t realistically expect to hit fully. You set it because aiming low locks you into incremental thinking. 70% of a moonshot is usually more than 100% of a safe goal. Set them on goals where the upside justifies the risk. Don’t set them on operational metrics where you genuinely need to hit 100% (uptime, compliance, safety). One client raised their support-volume moonshot mid-engagement once they understood the problem well enough to be bolder.
Related reading: OKR vs KPI: what’s the difference?.
A note on expectations
OKR isn’t a transformation programme. It’s a practice.
Your first cycle will be messy. Goals will be too broad. Key Results will slide back toward outputs. The weekly check-in will feel awkward. The champion network will be uneven. Half your team will quietly doubt the whole thing in week four. That’s normal.
The value compounds. By cycle three, most of the teams I coach find they’re making sharper decisions faster, surfacing problems earlier, and having more honest conversations about what is and isn’t working. That is the return.
One of my longest-running clients, Juan Sanchez at Inteleos, put it this way after several years of coaching: “Do not wait to get a coach because this journey is difficult. It’s totally worth it. It’s like, oh, I’m going to go do Everest and I’m not going to have anybody to help me up the mountain.”
Start messy. Get better. That’s OKR.
Want to go deeper?
If you want to run a properly structured OKR launch with your leadership team, that’s what The 90-Day OKR Quickstart Blueprint covers. Goal-setting workshops, champion training, weekly confidence scoring, Telos as a coaching tool, and hands-on coaching through the first cycle. I’ve helped global franchise networks, financial services firms, healthcare organisations, defence primes, fintech scale-ups, and government agencies through this exact journey. The pattern is the same. The practice compounds.
If you’re a founder or executive team in a product-led business who’s busy but not focused, let’s talk.
Related reading: The OKR Quickstart podcast.
About the author
Tim Newbold is an OKR Coach, Product Strategy Coach, Keynote Speaker, and Founder of OKR Quickstart. He helps product-led businesses turn strategy into measurable results, including 33% quarterly sales growth at Domino’s, 88% work visibility gains at MLC Life Insurance, and transformations into outcome-focused product organisations at companies like Titan FX and Findex. Clients include SEEK, Carsales, Scalapay, BAE Systems, Inteleos, and Foundr. Based in Melbourne, Australia. Known for making strategy executable and building the operating conditions that let great teams do their best work. Speaks internationally on strategy execution, OKRs, and product-led transformation. hello@okrquickstart.com
