OKR Framework: Examples, Templates & How to Write Them
OKRs are the most widely used goal-setting framework in modern teams. They were popularized by Google in the late 1990s and are now standard at companies from Spotify to Airbnb. The structure looks simple: one Objective everyone can repeat, plus three to five Key Results that prove it. Writing good ones is harder than the structure suggests.
This guide walks through what an OKR actually is and how to write one that drives change rather than activity. It includes real examples by function (including agencies), how scoring and grading work, and the mistakes that derail most implementations. Use the builder below to draft your own as you read.
OKR Builder + Quality Scorer
Type an Objective and 3 to 5 Key Results. The widget grades each in real time: number? outcome (not activity)? time-bound? bold? Live tags show what passes and what to fix.
Quick Answer: What Is the OKR Framework?
An OKR (Objectives and Key Results) is a goal-setting framework with two parts. The Objective is a memorable, qualitative statement of what the team wants to achieve in a quarter ("become the agency our SaaS clients call first when onboarding gets complex"). The Key Results are three to five measurable outcomes that prove the Objective is being met ("ship a productized onboarding audit, sold to 4 clients by end of Q3"). OKRs run on a quarterly cadence and are designed to drive change rather than monitor steady-state performance.
The framework has been used at scale by Intel, Google, LinkedIn, Airbnb, Spotify, and many smaller teams. It works because it forces a small number of priorities to the surface, ties them to outcomes everyone can verify, and resets every quarter so the conversation stays current.
Origin and Why It Works
OKRs were introduced at Intel in the 1970s by Andy Grove, who called the system iMBOs (Intel Management by Objectives). Grove documented the framework in his 1983 book High Output Management, devoting roughly five pages to the structure that would later spread across Silicon Valley. John Doerr learned the system as a young engineer at Intel, and in 1999 brought it to Google when he pitched it to Larry Page and Sergey Brin.
"A successful MBO system needs only to answer two questions: Where do I want to go? The answer provides the Objective. How will I pace myself to see if I'm getting there? The answer gives us milestones, or Key Results." - Andrew Grove, High Output Management (1983)
Doerr later popularized the framework in his 2018 book Measure What Matters, which catalogued OKR adoption at organizations from the Gates Foundation to Bono's ONE campaign. The reason it works is mechanical, not philosophical: writing one Objective forces ruthless prioritization, and writing measurable Key Results forces honesty about whether the work moved the number.
"OKRs have helped lead us to 10x growth, many times over." - Larry Page, Alphabet CEO
How to Write an OKR
Strong OKRs share four properties. Memorable: the team can repeat the Objective from memory. Measurable: every Key Result has a number. Outcome-oriented: Key Results describe results, not work. Time-bound: every Key Result has a deadline within the quarter. The widget at the top of this article checks each of these in real time as you draft.
The most common failure mode is writing activity instead of outcome. "Run 4 marketing campaigns this quarter" is an activity; the team can run all four and still end the quarter with the same conversion rate they started with. "Lift trial-to-paid conversion from 12% to 18% by September 30" is an outcome; either the number moved or it did not. Watch for verbs like consult, help, analyze, participate, support, and review. They are signals you have written work, not impact.
"It's not a key result unless it has a number." - Marissa Mayer, formerly Google
Mayer's rule is the cleanest test there is. If you cannot append a number to a Key Result, rewrite it. Numbers can be percentages, ratios, dollar values, count of artifacts shipped, or NPS scores. They cannot be feelings, alignment, or "improved." A Key Result that ends in "improve customer satisfaction" is a draft. "Lift customer onboarding NPS from 42 to 60 by end of quarter" is finished.
OKR Examples by Function
Examples make the structure concrete. The table below shows one Objective and three Key Results per function, written to the rules above (numeric, time-bound, outcome-focused). Use them as starting points, not copy-paste templates; the right Objective for your team depends on what specifically needs to change this quarter.
| Function | Sample Objective + Key Results |
|---|---|
| Marketing | Become the highest-converting acquisition channel by end of Q3 Lift trial-to-paid conversion from 12% to 18% Publish 8 SEO-optimized comparison articles ranking on page 1 Hit 1,200 monthly product-qualified signups by September |
| Sales | Open the mid-market segment as a reliable growth lane Close 12 mid-market deals (50-200 FTE) by end of quarter Average deal size from $8K to $14K ARR Pipeline coverage of 4x quota by week 8 |
| Product | Make onboarding the reason customers stay past day 30 Day-30 retention from 58% to 72% by end of quarter Time-to-first-value under 7 minutes for 80% of new users Ship the redesigned welcome flow to 100% of new accounts by July 15 |
| Engineering | Cut friction in the critical-path release cycle Reduce average PR-to-prod time from 3 days to under 24 hours Bug rate per shipped feature from 4 to 1.5 Deploy automated rollback for 90% of production services by August |
| Agency | Become the agency our SaaS clients call first when onboarding gets complex Ship a productized onboarding audit, sold to 4 existing clients by end of Q3 Publish 3 client onboarding case studies on the agency blog by September 30 Average client onboarding NPS of 60 across the quarter |
| HR / People | Build a hiring engine that does not stall as we scale Time-to-hire under 30 days for 90% of roles Offer-acceptance rate above 80% across the quarter Ship a structured-interview rubric for all 5 priority roles by August 31 |
The agency row deserves a closer look because most public OKR examples skip this audience. Service businesses face a unique challenge: a lot of the metric you would track (client retention, expansion revenue, NPS) depends on client decisions you do not fully control. The cleanest pattern is to write Objectives about capabilities the agency builds (productized service launches, case-study output, internal tools) and let the client-facing outcomes follow as Key Results. The agency controls whether it ships the audit; the client controls whether they buy. Both still belong in the same OKR.
How OKRs Cascade Across the Company
At company scale, OKRs need to align across levels without becoming top-down theatre. Company-level OKRs set the strategic direction. Team OKRs translate that direction into specific outcomes the team can credibly commit to. Individual OKRs (when used at all) translate the team's outcomes into personal contributions.
The principle Google's playbook returns to is roughly 50/50. About half of any team's OKRs should come from the company OKRs above them. The other half should come from the team itself, based on what it sees on the ground. Pure top-down cascade kills the framework. When team OKRs are just restated company OKRs, no one owns them and the quarter ends with everyone pointing at someone else.
The cleanest cascade has the team OKR addressing the how of the company OKR, not the what. Example: a company OKR is "lift gross margin by 4 points by Q4." The product team's supporting OKR is not "lift gross margin by 4 points." It is "ship the new pricing tier to 80% of accounts by Q4 and reduce support cost per ticket by 30%." The product team owns levers that move the company number; the company OKR sets the direction.
For most teams under 50 people, two levels (company plus team) is plenty. Adding individual OKRs on top tends to produce paperwork without changing behavior. Reserve the individual layer for organizations large enough that team OKRs are not specific enough to drive a single person's quarter.
Committed vs Aspirational OKRs
Not every OKR carries the same weight. Google's published playbook draws an explicit line between two types. Committed OKRs are commitments the team must hit at 1.0 (100%). They cover work that is non-negotiable for the quarter, like a dated launch or a regulatory deadline. Aspirational OKRs are stretch goals where 0.6 to 0.7 (60-70%) is success. They cover bold targets the team is reaching for, where the discipline of trying creates progress even if the full number is not hit.
The discipline that makes this work is tagging the type when you write the Objective, not at the end of the quarter. Marking an aspirational OKR as committed creates panic at 70% and hides what is actually working. Marking a committed OKR as aspirational invites the team to miss it. Be explicit about which kind each OKR is and the team's behavior follows.
How to Score and Review OKRs
OKRs are scored on a 0.0 to 1.0 scale, where each Key Result gets a decimal grade based on how much of the target was achieved. The Objective's score is typically the average of its Key Result scores. Google's published guidance says a healthy team should average around 0.7 across its aspirational OKRs over the year. Consistently scoring 1.0 is a sign of sandbagging (the goals were not bold enough); consistently scoring under 0.4 is a sign of overcommitment.
The cadence that works in practice has three layers. Weekly: a 15-minute check-in on each Key Result with the owner. Mid-quarter: a calibration meeting where the team reallocates effort to OKRs at risk and rescopes anything stalled. End of quarter: a grading session where each Key Result gets a 0.0 to 1.0 score and the team agrees what to learn from misses.
The mid-quarter calibration is the step most teams skip. It is also the step that delivers most of the framework's value. By week 6 or 7, you usually know which OKRs are tracking and which are not. Acting on that information instead of waiting for the end of the quarter is the difference between OKRs as a goal-setting ritual and OKRs as an operating system.
Common Mistakes
The patterns below show up across teams that adopt OKRs and lose the value within two quarters. Most of them come from the same root cause: treating OKRs as a planning artifact rather than a live operating system.
- Writing activities, not outcomes "Run 4 marketing campaigns" is an activity. "Hit 1,200 product-qualified signups by September" is an outcome. Activity-language Key Results turn the OKR into a to-do list and remove the accountability for whether the work actually moves the number.
- Setting too many OKRs Teams that adopt 6 to 8 OKRs end up sprinting nowhere. The whole point of the framework is focus. Cap each team at 2 to 4 OKRs per quarter, with 3 to 5 Key Results per Objective. More than that, and the team is back to a wish list.
- Confusing committed and aspirational Committed OKRs are commitments the team must hit; aspirational OKRs are stretch goals where 70% completion is success. Tagging an aspirational OKR as committed creates panic at 70% and hides what is working. Set the type when you write the Objective, not at end of quarter.
- Tying OKRs to performance reviews Google's playbook is explicit on this: OKRs are a strategic tool, not a performance evaluation. The moment compensation rides on OKR completion, teams sandbag the targets and the framework loses its bite. Keep OKR scoring and HR reviews in separate systems.
- Setting it in January, ignoring it until October OKRs that get written at the start of the quarter and reviewed only at the end are wallpaper. The cadence that delivers results is weekly check-ins, mid-quarter calibration, and an end-of-quarter grading session. The midpoint review is where most teams skip steps and lose the year.
- Disconnecting OKRs from the daily workflow An OKR that lives in a slide deck or quarterly memo drifts away from the work. The teams that get value pull the Objective and Key Results into the same workspace as their tasks and chat, so weekly check-ins happen against the metric, not against a separate dashboard the team forgets to open.
The biggest of these is the activity-vs-outcome trap. If you take one rule from this guide, take that one. Every Key Result must describe an outcome with a number, not a piece of work the team plans to do. The widget at the top flags activity verbs in real time. That live feedback is the fastest way to learn the muscle, especially when the team is new to OKRs.
What We Recommend
At Rock we run OKRs on a four-rhythm cadence in the same workspace where the work happens. Each team space holds a pinned OKR note (one Objective, three to four Key Results, plus type and owner per result). The KR owners post a one-line update in chat each Monday. A 30-minute mid-quarter calibration sits on the calendar by default in week 6. End of quarter, the team scores each KR, writes a short reflection, and the new quarter's Objectives go up.
The reason for keeping OKRs in the same workspace as tasks and chat is the failure mode otherwise. The OKR lives in a slide deck, the work lives in a different tool, and the two drift apart by week 4. Pair this with the broader strategy stack and the OKR is the operational layer underneath, whether you are running a 5-person small business or a 50-person team. SWOT covers situation. Strategic Choice Cascade covers integrated choice. PESTEL covers macro context. Porter's Five Forces covers industry structure. OKRs translate those choices into the specific outcomes the team commits to this quarter. For ongoing performance metrics that sit alongside OKRs, the OKR vs KPI guide covers when to use each and how they hand off; the KPI framework covers the discipline of what counts as a KPI, and the vanity metrics deep dive covers what to cut. For function-specific applications, see marketing KPIs, sales KPIs, and agency KPIs.
Set OKRs alongside the tasks that move them. Rock combines chat, tasks, and notes in one workspace. One flat price, unlimited users. Get started for free.
