Detailed content
🏗️ What OKRs Are and Where They Come From
OKR stands for Objectives and Key Results. It is a goal-setting framework that connects what you want to achieve (Objective) with how you will measure progress (Key Results). Simple on the surface, powerful in practice when executed well.
📖 The Origin: Andy Grove at Intel
Andy Grove, former CEO of Intel, created the system in the 1970s as an evolution of Peter Drucker’s MBO (Management by Objectives). Grove realized that traditional MBO focused too much on the “what” and ignored “how to measure it.” His innovation: separate the Objective (inspirational, qualitative) from the Key Results (measurable, verifiable). The system was described in detail in his book “High Output Management” (1983).
"The key result has to be measurable. But at the end you can look, and without any arguments: did I do that or did I not do it? Yes? No? Simple. No judgments in it." — Andy Grove
🚀 John Doerr and Google
In 1999, venture capitalist John Doerr introduced OKRs to Larry Page and Sergey Brin — when Google had just 20 employees. The system was adopted and became part of the company’s DNA. Doerr codified the method in "Measure What Matters" (2018), making OKRs the most discussed goal-setting framework in the corporate world. Today, companies such as LinkedIn, Twitter, Uber, Airbnb, and Amazon use variations of the system.
Adoption data
More than 65% of Fortune 500 companies use OKRs or an equivalent framework (Gartner, 2023). In Silicon Valley, the rate reaches 80% among Series B+ startups.
⚖️ Objective (qualitative) vs. Key Results (measurable)
The distinction between an Objective and a Key Result is at the heart of the system. Confusing the two is the most common mistake—and what turns OKRs into yet another disguised to-do list. Understanding the difference changes how you write, evaluate, and learn from your goals.
O — Objective
- • Qualitative and inspirational — should energize the team
- • Without a number — is not measurable in itself
- • Answers "where to" — strategic direction
- • Memorable — should fit in one sentence
Example:
“Make our innovation process the fastest in the industry”
KR — Key Result
- • Quantitative and verifiable — yes or no, with no ambiguity
- • Has a number — percentage, value, deadline
- • Answers "how to measure" — evidence of progress
- • Outcome-oriented — outcome, not activity
Example (3 KRs for the objective above):
KR1: Reduce the idea-to-pilot cycle from 90 to 30 days
KR2: Launch 5 validated experiments in the quarter
KR3: Raise the internal NPS for the innovation process from 42 to 70
🔬 The Doerr Test
John Doerr uses a simple test to check whether a KR is valid: "At the end of the period, can I unequivocally verify whether I achieved it or not?" If the answer depends on interpretation, the KR is weak. If the answer is a verifiable number or binary state, the KR is strong.
// Weak KR (not testable):
"Improve the company's innovation culture"
// Strong KR (testable):
"Raise the psychological safety score on eNPS from 55 to 75 points"
💡 Rule of thumb: 2-5 KRs per Objective
Doerr recommends 3 to 5 KRs per Objective. Fewer than 2 likely leaves important dimensions unmeasured. More than 5 indicates that the Objective is too broad or that you’re listing tasks, not outcomes. The sweet spot is 3 KRs covering complementary dimensions: speed, quality, and impact, for example.
🌕 Aspirational OKRs (Moonshots) for Innovation
Google popularized two types of OKRs that need to coexist: Rooftop OKRs (achievable goals, 70-100% confidence) and Moonshot OKRs (aspirational goals, 40-60% confidence). The distinction is critical for innovation: a team that only does rooftops will never go beyond incremental improvements.
Rooftop OKR
Objectives the team believes it can achieve with normal effort. Confidence of 70-100%. Failure is a problem — it indicates poor execution.
- • Incremental innovation
- • Improving existing processes
- • Growth in established metrics
Moonshot OKR
Radically ambitious objectives. Confidence of 40-60%. Reaching 70% is already a success. Failure is a learning data point, not something to be punished.
- • Disruptive innovation
- • New markets or models
- • Long-term bets
🔑 Google's 70% Rule
At Google, hitting 100% of a Moonshot OKR is a sign that the goal was too small. The ideal benchmark is 60–70%. This creates a counterintuitive system: you set goals you probably won’t fully reach, and that’s intentional. The logic: if the team always hits 100%, the OKRs aren’t driving real growth.
0-30%
Failure — review the strategy
60-70%
Ideal for moonshots
100%
Was the objective too small?
⚠️ Pay attention to context
The 70% logic works for innovation teams with a psychologically safe culture and leadership that genuinely accepts failure as a learning opportunity. In punitive cultures, ambitious OKRs lead to sandbagging (topic 6): people set low targets to guarantee 100% and look good. Before adopting moonshots, assess the team’s level of psychological safety.
🔁 Cadence, Check-ins, and Scoring
OKRs without cadence are decoration. What brings the system to life is regular review cadence: define them at the start of the cycle, hold quick weekly check-ins, a mid-cycle review, a retrospective, and score them at the end. Doerr calls this "CFR" — Conversations, Feedback, Recognition.
Definition (start of the cycle)
1–2 weeks of planning
Collaborative writing of OKRs. Teams define their own OKRs aligned with company goals. Leadership validates alignment; it doesn’t dictate content. This autonomy is critical for engagement.
Weekly Check-in (15 min)
Weekly cadence, not monthly
Each KR gets a status: green (on track), yellow (at risk), red (blocked). It’s not a report — it’s a conversation. “What changed? Do you need help with anything?” A long check-in is a sign that the OKRs are poorly written.
Mid-Cycle Review
Adjust or discard
OKRs can be abandoned if the context has changed radically. That’s a feature, not a bug. An OKR that has lost relevance should be dropped without guilt—keeping it out of obligation creates work without value.
Retrospectives and Scoring
Learn, don’t judge
Score each KR from 0.0 to 1.0. Average the Objective scores. Document: What did we learn? What would we change? The score should not be tied to bonuses or performance reviews — that destroys aspirational behavior.
⚠️ OKRs tied to bonuses = disaster
Doerr and Grove agree: tying OKRs to variable pay kills the system. People stop writing ambitious goals and start writing guaranteed ones. The result is systematic sandbagging. OKRs are for learning and alignment, not for evaluating individual performance.
📚 Learning as a Key Result (Not Just Output)
One of the most subtle mistakes in innovation OKRs is measuring only outputs (what was delivered) instead of outcomes (what changed in the world) and learning milestones (what we learned). In innovation, validated learning is often more valuable than the deliverable itself.
🧪 Output vs. Outcome vs. Learning
Output KR (weak for innovation)
"Launch 3 new product features by the end of the quarter"
Measures delivery. Doesn’t measure value. You can deliver 3 features nobody uses.
Outcome KR (better)
"Increase D30 retention from 22% to 35% with the new features"
Measures real impact on user behavior.
Learning KR (ideal for innovation)
“Validate or falsify the hypothesis that users would pay R$29/month for feature X with 100 interviews and 1 pricing test”
Measures learning. Valid regardless of whether the result is positive or negative.
The concept of a Learning KR is especially powerful in the early stages of innovation, when you still don't know whether you're headed in the right direction. A well-written OKR for a discovery phase could be: "Conduct 50 interviews with B2B customers and document the 5 most important unmet Jobs to Be Done". A good or bad interview result is equally valuable.
📊 Types of KRs by innovation phase
Discovery
Learning KRs dominate. Validated hypotheses, interviews conducted, insights documented.
Validation
A mix of learning and outcomes. MVP conversion rate, early retention, prototype NPS.
Scale
Outcomes dominate. Revenue growth, CAC reduction, market expansion.
💡 Tip: document what didn’t work
A learning KR with a negative result (“the hypothesis was falsified”) should be celebrated, not buried. Document what did not work with the same rigor as what did. Organizations that systematically learn from failures innovate faster than those that record only successes.
🚫 Common Mistakes in Innovation OKRs
OKRs are easy to understand and hard to execute well. The three most destructive mistakes—KRs turning into to-do lists, sandbagging, and having too many OKRs—are also the most common. Identifying and correcting each one is what separates teams that improve with OKRs from those that abandon them after two cycles.
✗ Mistakes to Avoid
- ✗ KR is a task list (“hold 3 workshops”)
- ✗ Sandbagging: goals so easy that a 100% result is guaranteed
- ✗ Too many OKRs: more than 5 objectives per team per cycle
- ✗ Purely top-down: leadership imposes OKRs without the team’s participation
- ✗ Set and forget: define them at the start and forget about them until the end
✓ How to Fix It
- ✓ Ask, "Does this measure an outcome or activity?" — a KR should measure an outcome
- ✓ Calibrate with "if we hit this easily, was the objective ambitious enough?"
- ✓ Grove's rule: a maximum of 3-5 objectives, with 3-5 KRs each. Focus and intentional scarcity.
- ✓ 60–70% of OKRs come from the bottom up — the team proposes, leadership aligns
- ✓ Weekly 15-minute check-ins keep them active and relevant
🔎 Anatomy of an Error: KR as a To-Do List
The most common mistake: writing activities as if they were outcomes. The difference is subtle but fundamental. One activity describes what you will do. A result describes what will change in the world because of what you did. You can complete 100% of the activities and still not move the needle at all.
// Activity (wrong):
KR: “Implement an ideas program with a digital suggestion box”
// Result (correct):
KR: “Increase idea submissions from 12 to 80/month and the implementation rate from 2% to 15%”
Quick test: if the KR starts with an action verb (“implement,” “create,” “launch,” “carry out”), it’s probably a task. If it starts with a current metric and a target (“from X to Y”), it’s probably an outcome.
🎯 Sandbagging: The Silent Poison
Sandbagging happens when individuals or teams deliberately set low goals to ensure they’ll be met. In cultures where OKRs are tied to performance and compensation, or where failure is punished, sandbagging naturally emerges as a form of self-protection. The result: the entire company seems to be hitting 100% of its OKRs, but it isn’t growing. The system has become theater.
💡 How to Diagnose Sandbagging
If a team consistently achieves 90-100% of its OKRs quarter after quarter, don’t celebrate — investigate. Ask: “If you had more confidence to take a shot, what goal would you set?” The answer reveals the real OKR.
📋 Too Many OKRs: Everything Is a Priority = Nothing Is a Priority
Grove was explicit: the point of OKRs is to create focus through scarcity. When a team has 10 objectives and 40 KRs, the framework has lost its purpose. Doerr and Grove's practical rule: maximum of 3-5 objectives per level (company, team, individual) with 3-5 KRs each. If you cannot list your OKRs from memory, you have too many OKRs. The real strategic thinking happens when you struggle to choose what to cut.
📋 Module Summary
Next Module:
2.7 - Systematic Creativity: tools for generating innovation repeatably