Measure What Matters Summary and key ideas

Measure What Matters, by John Doerr.

Goals can bring clarity, but they can also distort behavior. Measure What Matters presents OKRs as a way to make goals useful for collective execution without turning them into orders. The system joins a clear destination to observable evidence, then keeps both open to discussion. Its premise is practical: ideas, credentials, and ambition matter only when people convert them into results.

OKR is a compact grammar. The objective says what should be achieved: it should be significant, concrete, action-oriented, and ideally inspirational. Key results say how progress will be demonstrated. They are specific, time-bound, aggressive yet realistic, measurable, and verifiable; a key result is not complete without a number. An objective may remain relevant for a year or longer, while key results are usually assessed at the end of a quarter or another defined period and can change as work becomes clearer. This moves attention from activity to output, especially in knowledge work, where visible busyness is not meaningful contribution.

The intellectual ancestry begins with Peter Drucker’s management by objectives and self-control. Earlier scientific management associated systematic measurement with authoritarian command. Drucker argued for a results-driven but humanistic organization, built on trust, respect, consultation, data, dialogue, and shared direction. People who help choose a course of action, the book says, are more likely to carry it through. A cited meta-analysis gives this logic a measurable contrast: productivity gains were much larger when commitment was high than when it was low. MBOs supplied the participatory foundation, but their later failures exposed what the system still needed.

In many organizations, MBOs became centrally imposed, slow to update, hidden in silos, reduced to contextless performance indicators, and tied to salary or bonuses. Those choices weakened ownership and discouraged risk-taking. Andy Grove adapted the idea at Intel by bringing manufacturing discipline to administrative, professional, and managerial work. The difficult question was what meaningful output looked like when the worker was paid to think. Grove’s answer emphasized the relationship between what would be accomplished and how accomplishment would be verified.

Intel’s version used quarterly or monthly cycles, public goals, transparent information, and room for bottom-up or sideways contributions. It encouraged aggressive aspirations while keeping most goal measurement separate from compensation. Doerr’s first Intel OKR made the method concrete: demonstrate that the 8080 microprocessor outperformed the Motorola 6800 through five benchmarks, a demonstration, sales-training material, and calls on three customers. Publishing that goal made his priorities visible, helped him decline unrelated requests, and gave colleagues a shared basis for coordination.

Doerr later brought the system to Google in 1999. The young company had an enormous mission and ambitious founders, but little management experience. Written, measurable goals offered a way to make choices, track progress, use timely information, and stop projects that were no longer worth pursuing. The book presents Google’s later success as multi-causal: OKRs contributed, but leadership, technical resources, and an open, team-oriented culture mattered too.

The case for goals is not a case for blind obedience. The book takes seriously the warning that poorly managed targets can narrow attention, encourage unethical behavior or excessive risk, reduce cooperation, and demotivate people. It points to destructive target pursuit in examples such as the Ford Pinto fuel-tank failures, Sears repair gouging, inflated Enron sales, and the 1996 Mount Everest disaster. At the same time, it cites Edwin Locke’s research tradition: hard goals tend to outperform easy ones, and specific hard goals tend to outperform vague ones. The author reports more than a thousand supporting studies and productivity gains in most field experiments, while presenting those figures as cited evidence rather than a guarantee.

That tension explains the safeguards. Goals should be few enough to force choice: Grove recommends three to five objectives and generally no more than five key results for each. They should be negotiated rather than simply assigned. Teams and individuals are expected to create about half of their goals with managers, because exclusively top-down targets can erode motivation. OKRs are a cooperative social contract: an organization may set an objective, but the key results remain open to collective agreement. They can also be revised or discarded when circumstances make them irrelevant. Aspirational goals may stretch beyond immediate reach and tolerate failure, while operational commitments still require full delivery.

Finally, OKRs should measure learning and progress, not serve as a weapon, a legal record, or a direct formula for compensation. Separating them from bonuses reduces the temptation to sandbag targets and preserves room for honest ambition. The method demands repeated practice; Intel initially misunderstood it, and organizations may need several cycles to build mature goal-setting habits. The result is a living system: ambitious enough to move work forward, transparent enough to invite challenge, and humane enough to leave judgment in the hands of the people doing the work.