What Is an OKR? Objectives & Key Results Explained
OKR stands for Objectives and Key Results. It's a goal-setting framework used by teams and companies to set an ambitious Objective and track it with two to three measurable Key Results.
- Key Result 1: Raise 30-day activation from 48% to 70%.
- Key Result 2: Cut time-to-first-value from 6 days to under 1.
- Key Result 3: Lift new-customer NPS from 22 to 40.
Objectives, Key Results & Initiatives
Every OKR is built from three parts. Keeping them distinct is what separates a real OKR from a to-do list with ambitions.
Where you're going
A clear, inspiring, qualitative goal that tells you where to go and why it matters. No number in it — it's the direction, not the measure.
How you'll know
A measurable outcome that proves the Objective is being met. Each one moves from a baseline to a target. If you can't put a number on it, it isn't a Key Result.
What you'll do
The projects and tasks you believe will move the Key Results. Initiatives are the work, not the goal — you complete them, but you measure the Key Result.
OKR examples for every team
The framework looks a little different in each function. Here's one worked OKR for five teams — each an Objective plus outcome Key Results that move from a baseline to a target. For ready-made starting points, browse our OKR templates by role and function.
- Grow qualified pipeline coverage from 2.1x to 3.5x
- Lift win rate on sourced deals from 18% to 28%
- Cut average sales cycle from 74 to 50 days
- Raise weekly active use from 34% to 55%
- Increase feature adoption of the new flow from 0 to 40%
- Cut time-to-first-value from 6 days to <1
- Grow organic-sourced SQLs from 40 to 110/mo
- Lift demo-request conversion from 1.9% to 3.5%
- Rank top-3 for 5 priority head terms
- Cut lead-time-to-deploy from 9 days to 2
- Reduce change-failure rate from 18% to 8%
- Bring P1 incident recovery from 4 hrs to <1
- Ship self-serve onboarding for the top 10 use cases
- Lift 90-day activation from 52% to 75%
- Stand up expansion handoffs: 0 to 40 qualified
- A qualitative Objective with no number
- 2–3 Key Results, each baseline → target
- Outcomes that move, not metrics to hold
OKRs vs KPIs
A KPI measures the health of something you're already doing. An OKR is a deliberate push to change something. They work together, but they aren't the same — and confusing them is why so many “OKRs” never change anyone's behaviour.
| OKR | KPI | |
|---|---|---|
| Purpose | Drive a specific change this cycle | Monitor ongoing health |
| Time frame | A cycle — usually a quarter | Always-on, no end date |
| Ambition | A stretch you're not sure you'll hit | A level you aim to hold |
| Doing its job when | It moves | It stays steady |
| Example | Raise activation 48% → 70% | Keep uptime above 99.9% |
Across 20,952 Key Results, more than half were really KPIs or tasks in disguise — metrics to hold steady, or work to complete, dressed up as goals to move. When half your Key Results aren't built to move, the framework can't do the one thing it's for.
OKRs by level: company, team & individual
OKRs work as a stack. Company objectives set direction; team OKRs turn that direction into outcomes they own; individual OKRs — where used — connect a person's work to both. Alignment is the thread that runs top to bottom.
A small number of company Objectives set the priorities for the whole cycle. They're broad, ambitious, and owned by leadership. Everything below should ladder up to one of them — if it doesn't, it's a sign the work isn't connected to strategy.
Each team writes OKRs that advance a company Objective in their own area. This is where most of the real work lives. Team OKRs cascade from company goals — not copied down, but chosen so that hitting them moves the level above.
Individual OKRs are the most contested level — useful for connecting a person's work to the team's outcomes, risky when they turn into a performance-review checklist. Use them to create clarity, not to grade people.
Most OKRs die after the first cycle. Here's why.
In 2017 I rolled out OKRs to a team for the first time. We ran the offsite, wrote the objectives, filled a spreadsheet with key results, and felt organised. By week six the spreadsheet was stale. By quarter-end nobody could tell me which goals we'd hit, because nobody had looked since kickoff. The writing was never the problem; keeping them alive was.
That's the honest reputation problem. Most failed rollouts trace back to one habit: the framework gets treated as a planning exercise — done at the start of a quarter and filed away — instead of an operating rhythm run every week. The writing is the easy part.
The median return on OKR software across 330 organisations was $25 of value for every $1 spent — from completion lift, time saved on status reporting, and the reduced cost of strategy quietly drifting off track.
The rest of this guide is about closing that gap: how to write OKRs that can be executed, how to run the cycle that keeps them alive, and what the data says separates the teams that hit their goals from the ones staring at a stale spreadsheet in week six.
A short history of OKRs
The framework has a clear lineage, and it explains why OKRs are shaped the way they are — a qualitative goal paired with measurable outcomes.
- 1954Peter Drucker — Management by ObjectivesMBO introduced the idea that teams should agree on clear objectives and measure results against them, rather than manage by activity.
- 1970sAndy Grove — OKRs at IntelGrove reshaped MBO into the form we'd recognise today — an Objective paired with a few Key Results, run on a short cadence. He turned a planning idea into an operating one.
- 1999John Doerr — OKRs to GoogleDoerr, who learned the method under Grove, brought OKRs to a young Google. The company adopted them company-wide — the case study that made OKRs famous.
- TodayEverywhere — and mostly misusedOKRs are now standard in growing companies, but the method spread faster than the discipline behind it. Which is the rest of this guide.
How to write OKRs that hold up
A good OKR follows a shape. Once you've seen the formula for each part, writing them stops being a blank-page problem.
I will [Objective] as measured by [Key Results] — a qualitative, memorable goal, proven by the measures beneath it. No number in the Objective itself; if it could be a permanent mission statement, it's too broad.Move [metric] from [baseline] to [target] by [end of cycle] — always a baseline and a target. Two to three per Objective. No baseline and it's a task; no target and it's a wish.Weak vs strong: the same Key Result, rewritten
The most common writing mistake is measuring effort instead of outcome. Here's a weak Key Result and the same intent rewritten to actually move something:
It's an initiative, not a result. You either ship it or you don't — it can't tell you whether onboarding got better. This is the output trap in one line.
It's an outcome with a baseline and target. Launching the flow is now an initiative beneath it — the work you do to move the number, not the number itself.
Half of all Key Results have no named owner at all — and Key Results with a single named owner are completed 26% more often than those owned by a group or nobody. Ownership is the difference between a goal someone drives and one everyone assumes someone else is watching.
The OKR format, laid out
Stripped to its structure, every OKR fits the same simple table. This is the format to copy: one Objective, two to three Key Results with a baseline and target each, and the initiatives underneath.
| Element | What it is | Rule of thumb |
|---|---|---|
| Objective | The qualitative goal — where you're going | One per team, no number |
| Key Result 1 | A measurable outcome, baseline → target | Owned by one person |
| Key Result 2 | A second independent measure | Moves, not held steady |
| Key Result 3 | A third, if needed — stop at three | More KRs dilute focus |
| Initiatives | The projects that move the KRs | Completed, not measured |
A useful check: read the Objective and the three Key Results out loud. If the Objective inspires and each Key Result could be argued at quarter-end with a single number, the format is right. If you're explaining what a Key Result “really means,” it isn't measurable enough yet.
| Objective | Make our onboarding the reason customers stay. |
| Key Result 1 | Raise 30-day activation from 48% to 70% |
| Key Result 2 | Cut time-to-first-value from 6 days to under 1 |
| Key Result 3 | Lift new-customer NPS from 22 to 40 |
| Initiatives | In-app onboarding checklist · guided first-run · lifecycle email sequence |
The blank format above, as a ready-to-fill Excel or Google Sheets template — plus example OKRs by function.
