OKR stands for Objectives and Key Results. It is the goal-setting framework that Intel invented, that Google adopted in its earliest days, and that most serious tech companies now use to align teams around what actually matters. Understanding OKRs is not optional for product managers — it is the language in which goals are set, debated, and graded.
The formula
An Objective is an inspirational direction. It answers the question: where do we want to go? It should be memorable, motivating, and qualitative. A Key Result is a measurable signal that tells you whether you are getting there. It answers: how will we know we arrived? Every OKR follows the same structure: one Objective paired with two to four Key Results.
Two full examples
Objective: Make onboarding the best it has ever been.KR 1: Increase day-7 activation rate from 38% to 60%. KR 2: Reduce average time to first key action from 14 minutes to 6 minutes. KR 3: Achieve an onboarding NPS of 50 or above (baseline: 31).
Objective: Become the go-to product for freelance designers.KR 1: Grow monthly active freelance users from 4,200 to 10,000. KR 2: Achieve a 30-day retention rate of 55% for new freelance signups (up from 34%). KR 3: Launch two features requested by more than 200 freelancers in the feedback backlog.
What makes a good Key Result
A good KR is measurable — it has a number, not a feeling. It has a baseline and a target: you need to know where you are starting to know whether you moved. It measures an outcome, not an output — "launch the new onboarding flow" is a task, not a KR; "increase activation rate by 20 points" is a KR. And it is ambitious enough that hitting 70% of it still represents real progress.
Grading OKRs
OKRs are graded on a 0.0 to 1.0 scale at the end of each cycle (typically a quarter). A score of 1.0 means you hit the target exactly. The counterintuitive rule at Google and many other companies: a score of 0.7 is considered success. If you are consistently scoring 1.0, your targets were not ambitious enough. If you are consistently scoring below 0.4, your targets were unrealistic or your execution broke down — both worth investigating.
Common mistakes
The most common OKR mistakes are writing tasks as Key Results ("complete user research" instead of "increase qualitative insight coverage to 80% of top user segments"), setting too many OKRs (three per team per quarter is a maximum, not a minimum), skipping the baseline so you cannot calculate a grade, and only reviewing OKRs at the end of the quarter instead of weekly or biweekly. OKRs are only useful if the team looks at them regularly.
For a deeper look at how OKRs connect to product strategy, see the NewRoleKit OKR guide.