What is an OKR — and what separates a good one from a bad one.
Short answer
OKR stands for Objectives and Key Results. An Objective is a qualitative, inspiring direction ("become the most trusted tool in the market"). Key Results are 3–5 measurable outcomes that define what achieving the objective looks like ("grow NPS from 28 to 45"). OKRs were developed at Intel by Andy Grove and popularized at Google. They run on a quarterly cycle and are scored 0.0–1.0 — a 0.7 is ideal; 1.0 means the goal was too easy.
OKRs were developed by Andy Grove at Intel in the 1970s, based on Peter Drucker's Management by Objectives framework. John Doerr brought OKRs to Google in 1999 and documented them in his 2018 book Measure What Matters. Today they're used by companies from early-stage startups to organizations like Google, Netflix, and Twitter.
The structure.
- Objective: What we want to achieve. Qualitative, inspiring, time-bound (quarterly or annual). Written in plain language, not corporate jargon. Example: "Make our onboarding process feel effortless for new users."
- Key Results: How we'll know we got there. Measurable outcomes — not activities, not tasks. 3–5 per objective. Example: "Reduce median time to first successful action from 8 minutes to 3 minutes." Another: "Increase 7-day retention from 42% to 60%."
Good OKR vs. bad OKR.
Bad objective: "Improve our product." (Not inspiring, not time-bound, not concrete.) Good objective: "Become the fastest onboarding experience in our category by Q3."
Bad key result: "Launch onboarding redesign." (Activity, not outcome — you could launch it and fail.) Good key result: "Reduce drop-off at Step 2 from 35% to 15%."
The scoring rule. OKRs should be aspirational — aiming for a 0.7 (70% of the ambitious goal) is considered success. A 1.0 means the goal was too conservative. A 0.3 or lower means either the goal was too ambitious or something went systematically wrong worth learning from. Scoring is honest and public within the organization.
OKRs vs. KPIs. KPIs are ongoing health metrics (monthly active users, revenue, churn). OKRs are time-boxed improvement goals. Both exist at the same company — KPIs tell you if the business is healthy; OKRs tell you if you're improving strategically.
What makes OKRs fail. Treating KRs as task lists. Setting too many objectives (more than 5 per team). Not scoring last quarter before setting new ones. Making OKRs non-public — transparency is what makes alignment work. Letting "business as usual" work dominate the OKRs.
OKR planning happens at a whiteboard — company objectives at the top, team OKRs below, dependency lines connecting them. Snap the board with BoardSnap and get a structured summary of the entire OKR tree, with each KR as a trackable action item.
Frequently asked
How many OKRs should a team have?
3–5 objectives per quarter, with 3–5 key results each. Most effective teams run 3 objectives. More than 5 objectives means you haven't made real choices — everything is a priority, which means nothing is.
Should OKRs be tied to compensation?
Most OKR practitioners (including John Doerr and the Intel tradition) recommend against tying OKRs directly to compensation or performance reviews. When OKRs affect pay, teams set conservative targets they're sure to hit. The 0.7 target score requires aspirational goal-setting — which only works when there's no penalty for missing.
What is the difference between company OKRs and team OKRs?
Company OKRs set the strategic direction for the whole organization. Team OKRs are set by each team and should align with at least one company OKR. Alignment doesn't mean duplication — a team's OKR should describe their specific contribution to the company goal, not just copy the company goal with different words.
See it work in ten seconds.
BoardSnap is free on the App Store. Snap a board — get a summary and action plan.