OKRs (Objectives and Key Results)

Definition: A goal-setting framework that pairs a qualitative Objective with a small set of measurable Key Results used to track progress toward it.

How It Works

Objectives

The Objective states an ambitious, memorable, qualitative goal — something that inspires and clarifies direction without itself being a number, e.g. “Make onboarding effortless” rather than “Increase activation to 60%”
Good Objectives are significant, concrete, and time-bound, usually to a quarter, and they should be understandable by anyone in the company without additional explanation
A company typically sets 3-5 top-level Objectives per quarter; teams then set their own Objectives that ladder up to, and support, the company-level ones
An Objective without a compelling “why” behind it tends to feel arbitrary — the strongest Objectives connect obviously to the company’s mission or current strategic priority
A commonly cited discipline for choosing Objectives: if the company achieved everything on this quarter’s list, would leadership genuinely feel meaningfully better positioned than today? If the honest answer is “not really,” the list needs sharper priorities

Key Results

Each Key Result is a specific, numeric, verifiable target that shows whether the objective was achieved — not a task to complete, but an outcome to reach
A well-formed Key Result answers “how will we know we succeeded?” with a number: “Reach 40% week-4 retention,” not “Improve retention” and not “Ship the onboarding redesign”
Most Objectives have 2-4 Key Results; more than that tends to dilute focus and makes it harder to tell what actually matters most
Key Results should be outcomes the team can influence but not fully control through sheer effort alone — if hitting the number is guaranteed just by showing up, it isn’t ambitious enough
Key Results are sometimes split into two flavors: committed Key Results the team fully expects to hit, and aspirational Key Results that stretch further and are expected to land below 100%
Good Key Results are also independent of each other where possible, so that a team can’t hit its numbers by trading progress on one Key Result for regression on another without anyone noticing

Scoring and Cadence

Teams typically set OKRs quarterly and score each Key Result from 0.0 to 1.0 based on how close it came to the target
A healthy target score is 0.6-0.7 completion on average; consistent 1.0 scores across the board are usually read as a sign goals were set too easily rather than a sign of great execution
Scoring is meant to be a diagnostic tool, not a performance review input — conflating the two encourages people to sandbag targets so they’re easier to hit, which defeats the purpose of ambitious goal-setting
Most teams check in on OKR progress weekly or biweekly rather than waiting for the end of the quarter, since early warning signs are far more useful than a surprise at the final review
A Key Result stuck at 0.0 through the first half of the quarter is a signal to intervene, not just to keep tracking — the check-in cadence only pays off if it actually triggers a change in behavior when needed

OKRs vs. KPIs vs. Traditional Goals

	OKRs	KPIs	Traditional goals
Purpose	Drive focused change in a period	Monitor ongoing business health	Define general direction
Ambition level	Intentionally stretch, ~70% expected	Meant to be consistently hit	Varies widely
Time horizon	Usually quarterly	Continuous, ongoing	Often annual or open-ended
Structure	Qualitative Objective plus numeric Key Results	Usually a single tracked number	Often vague or unmeasured
Consequence of missing	Expected sometimes; treated as signal	Usually treated as a problem	Ambiguous

OKRs and KPIs are complementary rather than competing: KPIs, like a North Star Metric or churn rate, are the ongoing vital signs of the business, while OKRs are the deliberate, time-boxed pushes a team makes to move one or more of those vital signs in a given quarter.

Why It Matters

Aligns every team around a shared set of priorities and makes progress visible, which is critical once a startup grows beyond the size where everyone can coordinate informally over lunch
Forces explicit tradeoffs: because a healthy OKR list is short, setting one means consciously not prioritizing something else, which is a discipline many growing teams otherwise lack
Makes it easy for leadership to spot misalignment early — if a team’s Key Results don’t clearly ladder up to a company Objective, that’s a signal the team may be solving the wrong problem
Creates a natural cadence for reflection: quarterly scoring forces an honest look at what worked, what didn’t, and why, rather than letting goals quietly slide without ever being revisited
Separates ambition from operations — Key Results describe outcomes the company wants, while the day-to-day roadmap and tasks describe how the team plans to try to get there, keeping the “why” distinct from the “how”
Makes performance conversations less personal and more objective, since progress is measured against a pre-agreed number rather than a manager’s subjective end-of-quarter impression
Gives investors and boards a lightweight way to track whether a company is executing against its own stated priorities quarter over quarter, beyond just watching top-line metrics

Common Pitfalls

Setting too many Objectives: more than 3-5 company-level Objectives per quarter dilutes focus and turns OKRs into a restatement of the entire roadmap rather than a prioritization tool
Writing Key Results that are really tasks: “Ship the new dashboard” is an output, not an outcome — a better Key Result asks what the dashboard should achieve, e.g. “Reduce support tickets about reporting by 30%”
Sandbagging targets to guarantee a good score: if scoring feeds into performance reviews or bonuses, teams learn to set OKRs they know they can hit, which defeats the framework’s purpose of encouraging ambitious stretch goals
Setting OKRs and never looking at them again until the quarter ends: OKRs need a regular check-in cadence, commonly weekly or biweekly, to actually influence day-to-day decisions rather than becoming a forgotten document
Cascading OKRs too rigidly top-down: when every team’s OKRs are dictated rather than negotiated, teams lose ownership and the framework becomes a compliance exercise instead of a genuine alignment tool
Confusing 100% completion with success: hitting every Key Result at 1.0 every quarter is far more often a sign of unambitious goal-setting than of excellent execution
Letting the framework outgrow its usefulness at the wrong size: very early-stage startups with just a few people often don’t need formal OKRs at all, since the overhead of the process can exceed its benefit before there’s enough organizational complexity to justify it

Writing Good OKRs

Start from a small number of company priorities, not a list of everything the company could plausibly do this quarter
Make each Objective inspiring enough that someone unfamiliar with the details would understand why it matters within a sentence or two
Pressure-test every Key Result with the question “if we hit this number, would we actually believe the Objective was achieved?” — if the answer is no, the Key Result is measuring the wrong thing
Prefer outcome metrics, like retention, revenue, or activation rate, over output metrics, like features shipped or meetings held, wherever possible
Keep the total list short enough to fit on one page and be recited from memory by the people responsible for it
Separate the OKR-setting conversation from the performance-review conversation explicitly, so people aren’t incentivized to lowball their own targets
Write Key Results so a neutral third party could look at the data at quarter’s end and score them without needing the team to explain or interpret the number

Common OKR-Writing Mistakes and Fixes

Weak Key Result	Why it’s weak	Stronger version
“Improve customer satisfaction”	Not measurable, no target	“Raise NPS from 30 to 50”
“Launch new pricing page”	An output, not an outcome	“Increase free-to-paid conversion from 8% to 12%”
“Work on reducing churn”	Vague verb, no number	“Reduce monthly churn from 5% to 3%”
“Have more meetings with customers”	Measures activity, not impact	“Complete 20 customer discovery interviews and identify 3 validated pain points”

OKRs in Practice at a Startup

Early-stage companies often run a lighter version of the framework: a handful of company OKRs set by the founders each quarter, with individual teams choosing their own supporting Key Results in a working session rather than through a heavyweight top-down cascade.

As the company scales past a few dozen people, OKRs typically become more formal — tracked in dedicated software, reviewed in a recurring all-hands, and explicitly tied to the company’s North Star Metric so every team’s targets visibly connect back to the single number leadership cares most about
Some companies run OKRs at two levels simultaneously: annual OKRs that set the year’s broad direction, and quarterly OKRs nested underneath that translate the annual goals into near-term, actionable targets
A recurring mid-quarter check-in, distinct from the final scoring session, gives teams a chance to flag a Key Result that’s clearly off track early enough to still course-correct
Retrospectives at the end of each quarter — discussing not just the score but why a Key Result landed where it did — are often more valuable than the score itself, since they surface planning mistakes worth fixing next quarter

Committed vs. Aspirational Key Results

	Committed Key Results	Aspirational Key Results
Expected score	Close to 1.0	Often 0.6-0.7
Purpose	Predictable operational targets	Stretch goals that push the ceiling
Example	“Close the books within 5 business days each month”	“Grow monthly active users from 10,000 to 30,000”
Risk if missed	Usually signals a real operational problem	Often expected, and treated as a learning opportunity

Mixing the two without labeling them clearly is a common source of confusion — a team that misses an aspirational Key Result at 0.6 can look like it underperformed, when in context that score might represent a genuinely strong quarter.

A Typical Quarterly OKR Cadence

Week	Activity
Week 1	Company and team Objectives finalized; Key Results set and communicated
Weeks 2-5	Regular execution; first biweekly check-in surfaces early risks
Weeks 6-9	Mid-quarter check-in; teams reprioritize or adjust tactics if a Key Result is off track
Weeks 10-12	Final execution push; teams gather data needed to score accurately
Week 13	Scoring, retrospective, and the next quarter’s OKRs are drafted using what was learned

This cadence keeps OKRs from becoming a “set it and forget it” exercise — the check-ins in the middle of the quarter are often where the framework earns its keep.

OKRs and Performance Reviews

Most OKR practitioners recommend keeping OKR scores explicitly separate from formal performance evaluations, even though the two inevitably inform each other informally
The reasoning is behavioral: the moment a low OKR score can hurt someone’s review or bonus, people begin unconsciously writing safer, easier-to-hit Key Results, which erodes the ambition the framework depends on
A better pattern many companies use is to evaluate performance based on how someone worked toward their OKRs — judgment, collaboration, problem-solving — rather than on the raw score itself
Managers can still use OKRs as a useful input to performance conversations; the key distinction is treating the score as one data point among several, not as an automatic formula for a rating
Some companies address the tension directly by publishing this separation as an explicit policy — “OKR scores are never an input to compensation” — so the incentive to sandbag targets is removed by written rule rather than left to trust alone

Company-Level vs. Team-Level OKRs

Company-level OKRs are set by founders or the executive team and represent the handful of things that matter most to the business as a whole this quarter
Team-level OKRs are set by individual functions — product, sales, engineering, marketing — and should each ladder up to at least one company-level Objective, even if the connection isn’t always a perfectly clean one-to-one mapping
Individual OKRs, used by some companies, break team OKRs down further to a single contributor level; many companies deliberately skip this layer, since it tends to reintroduce the task-tracking behavior OKRs are meant to avoid
The healthiest cascades leave room for teams to propose their own Key Results rather than receiving fully dictated ones, since teams closest to the work usually understand the achievable range better than leadership does
A useful sanity check at the end of OKR-setting season: read every team’s OKRs side by side and confirm they don’t quietly conflict or duplicate effort on the same underlying goal
When two teams’ Key Results pull in opposite directions — one optimizing for speed of shipping, another for stability — surfacing that tension during OKR-setting is far cheaper than discovering it mid-quarter through friction between the teams

Signs an OKR Program Is Working

People can recite the company’s top Objectives from memory without checking a document
Teams reference OKRs unprompted when deciding what to prioritize in weekly planning, not just when asked to report on them
Scores vary meaningfully across teams and quarters instead of clustering suspiciously close to 1.0 every time
Retrospective conversations produce real changes to how the next quarter’s OKRs are written, not just repeated targets with updated numbers
Leadership uses OKRs to say no to good ideas that don’t ladder up to a current priority, rather than only using them to justify ideas already underway

Signs an OKR Program Has Stalled

OKRs are written once each quarter and never referenced again until the next round of goal-setting begins
Every Key Result scores between 0.9 and 1.0, suggesting targets are being set conservatively rather than ambitiously
Teams can’t explain how their Key Results connect to a company-level Objective when asked directly
The same Key Results reappear quarter after quarter with only the numbers changed, suggesting the process has become rote rather than a genuine strategic exercise
More time gets spent formatting the OKR document than discussing whether the goals themselves are the right ones
New hires are handed the OKR document during onboarding but nobody can explain, off the top of their head, why those particular Objectives were chosen over the alternatives

Example

A startup sets the company Objective “Delight early customers” with Key Results including “Reach 40% week-4 retention” (up from 25%), “Achieve NPS above 50” (up from 30), and “Resolve 90% of support tickets within 24 hours.” The product team’s own quarterly OKRs then ladder up to that same Objective: their Key Result is “Ship and validate a redesigned onboarding flow that lifts week-1 activation to 65%,” a metric they believe feeds directly into the company’s week-4 retention target.

Midway through the quarter, a biweekly check-in reveals that week-1 activation has only reached 52% against a target trajectory of 58% at that point, so the team pulls forward a planned experiment — simplifying the account setup step — rather than waiting until the end of the quarter to react. At quarter’s end, the company lands at 36% week-4 retention, a 0.85 score against the stretch target, and an NPS of 48, a 0.9 score — a strong quarter by OKR standards even though neither number technically hit 100%, because the targets were deliberately set to be ambitious rather than safely achievable, and the mid-quarter check-in gave the team enough runway to adjust course before it was too late.

At the quarterly retrospective, the team also notices that the support-ticket Key Result landed at only 0.5, well below its committed target, because a product launch mid-quarter briefly overwhelmed the support queue. Rather than treating that as a failure to be glossed over, the retrospective becomes the basis for next quarter’s Objective — “Build a support system that scales with launches” — turning a missed number into next quarter’s clearest priority instead of a one-off excuse.

Referenced by

Dig deeper

Spot an error in this note?