Building the Company
How Google, Bono, and the Gates Foundation Rock the World with OKRs
A practical case for OKRs: a simple goal system that turns a founder's priorities into numbers the whole team can grade.
The big idea
Doerr argues that most organizations fail at execution, not ideas, and that a lightweight discipline called Objectives and Key Results fixes much of that. An objective states what matters; a few measurable key results define what done looks like by a date. Kept few in number, made public across the company, checked weekly and graded honestly at the end of each cycle, OKRs create focus and alignment. Paired with frequent conversations and feedback instead of annual reviews, they keep people stretching without gaming the score.
Key takeaways
Each OKR pairs one objective, which should be significant and concrete, with roughly three to five key results that are specific, time-bound and checkable. A key result is either hit or missed; there is no room to argue about it. If you cannot tell at the end of the quarter whether it happened, it is an activity or a hope, not a key result.
Doerr credits Andy Grove at Intel with the discipline of focus: a short list of objectives each cycle, chosen as much for what they leave out as what they include. When a company lists ten priorities, it effectively has none. The hard part of goal setting is the cutting, and the founder has to own it.
OKRs work best when everyone, including the CEO, can see everyone else's. Rather than cascading goals rigidly down the org chart, Doerr recommends that a large share be proposed by teams themselves and linked to company objectives. Transparency exposes duplicated work and hidden dependencies early, and lets people connect their own work to the company's direction.
OKRs are not set-and-forget. Teams check progress during the cycle, update status, and change or drop key results that no longer make sense. At the end, each key result gets a score from 0 to 1.0 and a short written reflection on what was learned. The value is in honest scoring, not precise scoring.
Committed OKRs, such as revenue, hiring or delivery promises, are expected to land in full, and a miss needs an explanation. Aspirational OKRs are deliberately ambitious, and at Google a score around 0.7 on them counts as success. Mixing the two without labels either demoralizes people or teaches them to set goals they know they can beat.
If bonuses ride directly on OKR scores, people quietly lower their targets. Doerr recommends decoupling the two and pairing OKRs with CFRs: ongoing conversations between manager and report, two-way feedback, and recognition. He describes companies such as Adobe replacing annual reviews with regular check-ins anchored to goals.
Doerr introduced OKRs to Google when it was still a small startup, shortly after his firm invested. For founders the lesson is that investors back execution. A monthly update that shows committed results hit, stretch results partly hit, and why, builds more confidence than adjectives about momentum.
The framework
Pick three to five objectives for the cycle; write down what you are deliberately not doing.
Give each objective three to five measurable, dated results that are clearly hit or missed.
Publish every OKR company-wide; let teams propose many of their own and link them up.
Review progress in short check-ins; flag what is off track and adjust or drop stale results.
Score each key result 0 to 1.0, write what you learned, and feed it into the next cycle.
“Ideas are easy. Execution is everything.”
— John Doerr
Apply it this week
Read next
A practical manual for the stage after product-market fit, when the company starts growing faster than the people running it.
A candid operator's guide to leading when the company might fail: layoffs, firings, bad news, and decisions without good options.
How to choose the one number that matters at your stage, measure it honestly, and turn it into evidence of traction.
This page is an original summary and commentary prepared by XRAYCLOUD.AI with AI assistance. It is not affiliated with or endorsed by the author or publisher, and it is no substitute for the book — read it for the full argument.