OKR, short for objectives and key results, is a goal-setting framework in which a team states a qualitative, ambitious objective and attaches two to five measurable key results that show whether the objective has been achieved. In product management, OKRs translate product strategy into quarterly outcomes, so that teams decide what to build according to the change they intend to cause rather than the features they intend to ship.
The method emerged at Intel in the 1970s under Andy Grove and spread through Silicon Valley, where it became a standard operating practice in many SaaS companies and startups. Its central idea is that progress is measured by outcomes, not activity.
How objectives and key results fit together
An OKR consists of one objective and a small set of key results that together define what success looks like for a fixed period, usually a quarter. The objective answers where the team wants to go, and the key results answer how the team will know it has arrived.
A good objective is qualitative, inspiring and time-bound, such as improving onboarding so that new customers reach value in their first session. A good key result is specific, measurable and verifiable, such as raising the share of trial accounts that complete setup within 24 hours from 38% to 55%. The initiatives, projects and tasks that the team expects to move the numbers are kept separate from the OKR itself.
The table below summarizes the three layers that practitioners distinguish.
| Element | Purpose | Characteristics |
|---|---|---|
| Objective | States the desired direction | Qualitative, memorable, one sentence |
| Key result | Proves the objective was reached | Numeric, baseline and target, two to five per objective |
| Initiative | Describes the work expected to help | Projects, experiments, features; not scored |
| Confidence score | Signals the forecast during the cycle | Updated weekly, often on a 1 to 10 scale |
Writing effective key results
Effective key results measure outcomes rather than outputs, which is the most important and most frequently violated rule of the framework. Shipping a feature is an output; the change in customer behavior that the feature should produce is an outcome.
Key results should include a clear baseline and target, be within the team’s sphere of influence and be measurable with existing instrumentation. A useful test is whether the key result could be achieved in a way that does not help the objective; if it could, the metric needs a counterbalancing key result, for example pairing conversion growth with a refund-rate ceiling.
The comparison below shows how output-based key results can be rewritten as outcome-based ones.
| Weak key result (output) | Strong key result (outcome) | Why it is better |
|---|---|---|
| Launch the new dashboard | Weekly active admins rise from 2,100 to 3,000 | Measures adoption, not delivery |
| Run 10 customer interviews | Validate one pricing hypothesis with 60% purchase intent | Captures learning, not effort |
| Release in-app checklist | Day-7 activation rate rises from 41% to 50% | Reflects user behavior change |
| Fix 50 bugs | Support tickets per 100 accounts fall from 9 to 6 | Tracks customer-felt quality |
Committed and aspirational OKRs
Organizations usually distinguish committed OKRs, which the team is expected to deliver fully, from aspirational OKRs, also called stretch or moonshot goals, which deliberately aim beyond what seems achievable. Mixing the two without labeling them causes confusion in reviews.
Scoring commonly uses a 0.0 to 1.0 scale. For aspirational OKRs, an average of about 0.6 to 0.7 is often treated as healthy, because consistently scoring 1.0 suggests the goals were not ambitious enough. For committed OKRs, anything below 1.0 calls for an explanation.
| Type | Ambition level | Expected score | Typical use |
|---|---|---|---|
| Committed | Realistic, must happen | 1.0 | Compliance deadlines, contractual launches |
| Aspirational | Ambitious, uncertain | 0.6 to 0.7 | Growth bets, new market entry |
| Learning | Exploratory | Judged on insight gained | Product discovery, early validation |
OKRs compared with other product management tools
OKRs define why and toward what outcome a team works, while other product management tools define what is built and in which order. They complement the product roadmap, prioritization frameworks and the product backlog rather than replacing them.
| Practice | Main question answered | Relationship to OKRs |
|---|---|---|
| OKR | Which outcomes matter this cycle | Sets the target |
| KPI | Is the business healthy over time | Ongoing metric; an OKR may aim to move one |
| Product roadmap | What will be built and roughly when | Should be derived from OKRs |
| RICE scoring | Which initiative offers most value per effort | Ranks initiatives that serve OKRs |
| Product vision | Where the product is heading long term | OKRs are quarterly steps toward it |
The difference between an OKR and a key performance indicator is often misunderstood. A KPI is a continuously monitored health metric, such as churn rate or uptime, while an OKR is a time-boxed commitment to change something, which may include improving a KPI that has drifted.
Running the OKR cycle
The OKR cycle is a recurring loop of setting, tracking and reviewing goals, most often on a quarterly cadence with annual company-level objectives above it. Alignment works in both directions: leadership states company objectives, and teams propose their own OKRs that contribute to them, a practice known as bidirectional alignment.
- Draft team OKRs two to three weeks before the quarter, using company objectives as input.
- Align across teams to expose dependencies and remove conflicting goals.
- Check in weekly with confidence scores and short notes on blockers.
- Score and reflect at the end of the cycle, separating the grade from performance reviews.
- Carry forward unfinished but still relevant objectives deliberately, not by default.
A widely cited recommendation is to keep OKRs separate from compensation. When bonuses depend on scores, teams set safe goals and sandbag targets, which defeats the purpose of ambitious goal setting.
Common mistakes when adopting OKRs
Most failed OKR programs fail because the framework is used as a task list or reporting ritual instead of a focusing tool. The most frequent problems are listed below.
- Too many OKRs: more than three objectives per team dilutes focus.
- Feature lists disguised as key results: delivery milestones replace outcome metrics.
- Set and forget: goals are written in week one and reviewed only at quarter end.
- Cascading only top-down: teams receive targets they did not help shape and do not own.
- No baseline: targets are set without knowing the current value of the metric.
The idea in brief
An OKR is a way for a team to agree on one important goal and a few numbers that prove it was reached. The goal is written in words that people remember. The numbers make it impossible to pretend success when nothing really changed.
Think of planning a fitness season. The goal might be to become a stronger runner. The proof could be running 10 kilometers under 50 minutes and training four times a week. Product teams do the same thing with customers and software every few months.
A practical scenario
A B2B project management startup notices that many trial users never invite colleagues, and its data shows that accounts with three or more members convert at roughly twice the rate of single-user accounts. The product team sets a quarterly objective to make collaboration the natural first step for new workspaces.
Its key results are to raise the share of trials with three or more members by day 7 from 22% to 35%, to increase the trial-to-paid conversion rate from 9% to 11%, and to keep onboarding completion above 70%. Initiatives include an invite prompt during setup, templates that assume a team and a reminder email. At quarter end, team adoption reaches 31% and conversion 10.6%, a score of about 0.7 on an aspirational OKR, and the team keeps the objective for one more quarter with refined initiatives.