Growth OKRs · 10 min read
Growth OKRs That Reward Shots on Goal, Not One Hero Number
I've run growth across a few companies over the last decade, and I've watched more growth teams die from one OKR than any other team I've worked with. Not because they were lazy. They bet the whole quarter on a single number they couldn't actually move on command. Growth isn't a number you will into existence. It's a rate of measured experiments, and the OKR should treat it that way. Here's how I write them so the team gets rewarded for learning, not just for getting lucky.
By Max Bondarenko · Last updated June 2026
Growth is throughput, not a wish
The trap this team falls into is picking one hero metric, signups or revenue or whatever the board is staring at this quarter, and writing it as the objective. Then the team spends twelve weeks praying. If the number moves, nobody knows why. If it doesn't, nobody learned anything. Either way you've burned a quarter and you can't repeat the result.
So here's the rule I hold every growth OKR to. It has to reward shots on goal that you actually measured. I want experiment velocity in the OKR alongside the outcome metric. A growth team that ran twelve clean tests and moved the number 40% of the way is healthier than one that ran two tests, got lucky on one, and hit target. The first team can do it again next quarter. The second can't. If a key result can't be tied to a baseline and a measured change, I'd kill it in the planning meeting.
Acquisition loops
Acquisition isn't a channel budget. It's a loop. A new user does something that brings the next user. If you can't draw the loop on a whiteboard, you're buying traffic, not building growth.
Turn paid acquisition from a leaky bucket into a self-funding loop we understand
KR1Cut blended CAC payback from 14 months to 9 months
KR2Get non-paid loop sources from 18% of new signups up to 35%
KR3Raise experiment velocity from 4 tests a month to 12, each logged with a baseline
The payback KR keeps people honest. It ties spend to a real time horizon instead of a vanity signup count. The velocity KR is the one most teams leave out, and it's the one I fight hardest for. Without it, a team can hit payback once by accident and never reproduce it.
Activation
Activation is where most of the money you spent on acquisition quietly leaks out. It's also the cheapest place to find a win, because the user already showed up.
Get new users to the moment they understand the product, fast, before they cool off
KR1Lift signup-to-activation rate from 30% to 50%
KR2Pull median time-to-first-value from 6 days down to under 2 days
KR3Cut day-one drop-off on the setup flow from 41% to 25%
Pick one activation event and define it before the quarter starts. If the team is still arguing about what 'activated' means in week six, the OKR is dead. I like a paired KR here, the rate going up and the time coming down, because you can game one without the other. Fast and shallow is not activation.
Conversion
Make the path from interested visitor to paying customer feel obvious and short
KR1Raise landing page conversion from 2.2% to 4%
KR2Lift free-to-paid conversion from 4.5% to 7%
KR3Ship and read out conversion experiments, from 3 last quarter to 9
Landing conversion is the seductive one because it moves fast and looks great in a deck, but it's worthless if traffic quality drops underneath it. That's why I anchor it to free-to-paid in the same objective. Together they tell you whether you converted the right people. The experiment-count KR stops the team from hiding behind one big redesign that eats the whole quarter and teaches nothing.
Referral / virality
Referral is the loop everyone wants and almost nobody earns. A viral coefficient below 1 isn't a failure, but you have to know your number cold before you try to move it.
Build a referral loop our best users actually want to use, not a coupon nobody clicks
KR1Move viral coefficient from 0.3 to 0.8
KR2Get active users sending at least one invite from 7% up to 22%
KR3Raise invite-to-signup conversion from 12% to 25%
The coefficient is the headline, but the two KRs under it are the steering wheel. How many people share, and how well a share converts. Move either and the coefficient follows. Teams that only track the coefficient end up staring at a number they have no levers for, which is the cardinal sin of growth OKRs.
The logic: why these work
Every KR above passes the same test. There's a baseline, a target, and the gap is something the team can attack with measured experiments. Baseline tells you where you stand, target tells you what ambition looks like, and the experiment-count KR proves you didn't get there by luck. A KR that says 'improve activation' is a wish. A KR that says '30% to 50%' is a bet you can hold someone to and learn from whether they win or lose.
On ambition: I calibrate targets so a good quarter lands around 70% of the way. If a growth team hits 100% every quarter, the targets are sandbagged and I'd push them harder. Here's the scar. One quarter early in my career I got greedy and stacked six objectives onto a growth team of four people, every one a stretch. We landed about 60% on each, which sounds fine until you realize that means we half-finished everything and finished nothing. The team was demoralized, I couldn't point to a single clean win in the board deck, and the lesson cost me a full quarter. Now I cap it at three objectives, and I'd rather nail two than blur five.
The weekly check-in for a growth team
Growth lives or dies on cadence. A monthly OKR review is too slow. By the time you notice an experiment stalled, you've lost three weeks you can't get back. I run a 30-minute weekly where the only thing that matters is what shipped, what it taught us, and what's queued next.
Five questions I ask every week
- 01How many experiments actually shipped this week, and how many did we plan to?
- 02Which test moved its baseline, by how much, and do we believe the read?
- 03What's blocking the experiments stuck in the queue, and who unblocks them?
- 04Is any KR target now obviously wrong given what we've learned, up or down?
- 05What's the single highest-leverage test we're not running yet, and why not?
And revise targets early, not at the end. If by week four it's clear a target was set on a bad baseline, change it then and write down why. The point of the OKR is learning velocity, not protecting a number you guessed at in planning. The teams I've run that revised mid-quarter, out loud, learned faster than the ones that white-knuckled a wrong target to the finish line.
A growth OKR template you can steal
Fill in your own baselines from last quarter's actuals. If you don't have a baseline for a line, that's your first job, not your KR. Numbers below are examples to show the shape.
| Objective | Turn [growth stage, e.g. activation] from our biggest leak into a repeatable loop |
|---|---|
| KR1 | Move [core rate] from [baseline, e.g. 30%] to [target, e.g. 50%] |
| KR2 | Pull [time or efficiency metric] from [baseline, e.g. 14 months] to [target, e.g. 9 months] |
| KR3 | Raise [experiment velocity] from [baseline, e.g. 4/mo] to [target, e.g. 12/mo], each logged with a baseline |
| Cadence | 30-min weekly: what shipped, what it taught, what's next |
| Owner | One growth lead owns the objective; each KR has a named driver |
Drop in your real baselines and targets before you commit this in planning.
Questions people actually ask
How many OKRs should a growth team have in a quarter?
Cap it at three objectives, and honestly two is better for a small team. I once ran six on a four-person team and we landed about 60% on every one, which meant we half-built everything and shipped nothing clean. Fewer objectives with real baselines beat a wall of stretch goals every time.
Should experiment velocity be a key result or just an input?
I put it in the OKR as a key result, and I'll fight anyone who says it's just a leading indicator. Growth is throughput. A team that ran twelve measured tests and moved the number partway is healthier than one that ran two and got lucky, because the first team can repeat it next quarter and the second can't.
What's a realistic target for viral coefficient in a growth OKR?
Know your baseline before you set anything. Most teams are sitting around 0.3 and don't know it. Moving from 0.3 to 0.8 in a quarter is ambitious but real if you attack the two levers under it: how many users share and how well a share converts. Chasing a coefficient above 1 in one quarter is usually a fantasy.
How do I write a growth KR that isn't just a vanity metric?
Anchor every seductive metric to a quality metric in the same objective. Landing page conversion is meaningless if traffic quality drops underneath it, so I pair it with free-to-paid. If a KR can move while the business gets worse, it's a vanity metric and I'd kill it in planning.
Run these without the spreadsheet
Okiar is free during beta. Voice check-ins, AI projections, team health — live in minutes.
Start free →OKR examples for other teams