What are OKRs? Objectives and Key Results explained

The one-sentence definition

An OKR pairs an inspiring, qualitative Objective (where you want to get to) with two to four Key Results (how you will know you have got there). Together, they replace vague ambitions with a clear, measurable commitment.

The framework was created at Intel in the 1970s by Andy Grove, popularised at Google when John Doerr introduced it in 1999, and is now used by some of the world's highest-performing organisations - from Spotify and Airbnb to NHS Trusts and FTSE 100 boardrooms.

The OKR formula

Every OKR follows the same simple sentence structure, sometimes called the "OKR formula":

The Objective answers "what do we want to achieve?" It should be inspiring, directional, and qualitative - something a team can rally around. It does not contain numbers.

The Key Results answer "how will we know we've achieved it?" They must be specific, measurable, and time-bound. If you cannot score a Key Result with a number between 0 and 1, it is not a Key Result - it is a task.

A worked example

Abstract frameworks are easier to understand with a concrete example. Here is an OKR for a B2B SaaS company's customer success team:

Notice how the Objective is ambitious and human - you would want to work on it. The Key Results leave no room for interpretation: either NPS is 65 or it is not. This combination of inspiration and rigour is what makes OKRs uniquely effective.

Why OKRs work - and why most goal-setting frameworks don't

Traditional goal-setting tends to fail in one of two ways. Either goals are set so conservatively that hitting them requires no meaningful change, or they are set so vaguely that no-one is clear when they have been achieved.

OKRs solve both problems simultaneously. The Objective stretches the team - OKR practitioners typically aim to achieve 60–70% of a Key Result, treating anything higher as a sign the target was not ambitious enough. The Key Results create an unambiguous definition of success that removes political debate at review time.

A second critical advantage is alignment. When every team's OKRs are visible across the organisation, it becomes immediately obvious where effort is duplicated, where dependencies exist, and where teams are pulling in different directions. This transparency is particularly powerful in complex, matrixed organisations where strategy is easy to lose between the board and the front line.

OKRs vs traditional goal-setting

Understanding what OKRs replace is as important as understanding what they add. The most common comparison is with KPIs - Key Performance Indicators.

KPIs measure the ongoing health of business as usual. Revenue per customer, churn rate, system uptime - these are dials that tell you whether your existing operation is performing as expected. They look backward at what has happened.

OKRs set ambitious targets for change and measure progress toward them. They look forward at what you are trying to make happen. The two are complementary, not competing: your KPIs tell you the business is healthy enough to invest in OKR-led transformation.

OKRs also differ from annual strategic plans. A plan specifies how you will achieve something - the tasks, activities, and workstreams. An OKR specifies what you will achieve and by when, leaving the team free to discover the best path. This distinction matters enormously in fast-moving environments where the right approach is not knowable in advance.

The four most common OKR mistakes

Most OKR implementations stumble for the same predictable reasons. Knowing them in advance is the fastest way to avoid them.

1. Writing tasks as Key Results

"Launch the new onboarding flow" is not a Key Result - it is a task. A Key Result describes the change in the world that the task is meant to produce: "Increase 30-day activation rate from 34% to 60%." The distinction feels pedantic until you realise that launching the feature and actually improving activation are two very different things. Teams that write tasks as Key Results end up shipping things nobody uses.

2. Too many OKRs

The word "objectives" is plural, but that does not mean "as many as possible." Most teams should have three to five objectives, each with two to four Key Results. Beyond that, the focus that makes OKRs work evaporates. If everything is a priority, nothing is.

3. Using OKRs to measure BAU work

OKRs are for what needs to change, not what needs to continue. If your team's work would happen regardless of the OKR, it belongs in an operational dashboard - not in the OKR. Reserve the framework for the things that represent genuine strategic progress.

4. Setting and forgetting

OKRs require a regular cadence of check-ins, scoring, and honest reflection. A quarterly OKR that is only reviewed at the quarterly retrospective is not an OKR - it is a forgotten aspiration. The check-in cadence is where the real work happens: surfacing blockers early, reassigning effort, and keeping teams accountable in a psychologically safe way.

The OKR review cadence

A well-run OKR programme operates on two timescales simultaneously: a quarterly cycle for setting and resetting, and a weekly or fortnightly rhythm for check-ins.

At each weekly check-in, team members update their Key Result scores (typically 0.0 to 1.0), flag any blockers, and note what confidence they have in hitting the target. This is not a status meeting - it should take fifteen minutes and create a visible record of progress.

At the quarterly retrospective, the team scores final attainment, reflects on what worked and what did not, and feeds those learnings into the next quarter's OKR cycle. Most teams score between 0.6 and 0.7 on average - if scores are consistently 1.0, targets are not ambitious enough.

At the annual level, OKRs should connect upward to the organisation's multi-year strategy. This vertical alignment - from company-level strategic objectives down to team and individual OKRs - is what separates a mature OKR programme from a collection of disconnected team goals.

Getting started with OKRs in your organisation

The fastest way to get started is to run a single quarter with a small, willing team before scaling. Pick a team of 8–15 people, run a two-hour OKR workshop to draft objectives and key results, and commit to a fortnightly check-in. After 12 weeks you will have enough data and experience to know how to roll OKRs out more broadly.

If you are managing OKRs at scale - across departments, business units, or geographies - a dedicated strategy execution platform makes a material difference. The manual overhead of spreadsheet-based OKRs becomes unmanageable above around 20 teams, and the lack of real-time visibility means blockers surface too late to act on.