Goals That Cascade: The OKR Method

Every goal-setting system in modern business descends from an argument about what a goal is missing. Peter Drucker introduced management by objectives in The Practice of Management in 1954: goals set jointly, reviewed against results. In the 1970s, Andy Grove — running Intel — found MBO too slow and too private for a company moving at semiconductor speed, and rebuilt it. His version, which Intel called iMBOs, kept the objective and added the element the name now carries: key results — a small set of measurable benchmarks that state, in advance, how everyone will know whether the objective was reached. Grove documented the system in High Output Management in 1983. One of the young Intel employees who sat in Grove’s course on the method, in 1975, was John Doerr.

The trip to Google, as its principals tell it

Doerr became a venture capitalist at Kleiner Perkins, and the standard history of OKRs — told principally by Doerr himself, most fully in his 2018 book Measure What Matters — runs like this. In the fall of 1999, having just led his firm’s investment in a young search company, he presented Grove’s system to its founders, Larry Page and Sergey Brin, and to a company of roughly forty people. Google adopted it, kept it as it grew, and made it famous. That provenance matters for an honest reason: the method’s history is largely a story its beneficiaries tell about themselves, and this article repeats no adoption statistics beyond it. What can be examined directly is not the legend but the grammar — and the grammar is genuinely well designed, which is why the method survived the telling.

The grammar has two parts and one cadence. An objective states what is to be achieved: qualitative, memorable, and few enough to fit on an index card — direction you can repeat in an elevator. Each objective carries key results: typically a handful of measurable outcomes, each with a number and a date, chosen so that hitting them leaves no argument about whether the objective happened. The cadence is the cascade: company objectives are set first, in the open; teams write theirs to serve them; individuals write theirs to serve the team’s — with the crucial discipline that goals are public by default and graded on a short cycle, usually a quarter, rather than discovered at year’s end.

What the method cannot do

OKRs transmit strategy; they do not create it. A company whose leadership has not decided what matters will produce beautifully formatted objectives that point in twelve directions, and the cascade will faithfully amplify the confusion to every desk. The method also degrades in two predictable ways. When key results are tied mechanically to pay, people learn to sandbag — to write results they are certain to hit — and the system quietly converts from an ambition instrument into a negotiation. And when every team writes key results that depend on other teams’ unlisted work, the quarter ends in a pile of green scores and an unshipped product. Neither failure is a flaw in the grammar. Both are reminders that the grammar is the easy part: the hard part is a leadership willing to name three priorities, in public, and let the fourth one go.

For a single team, the stripped-down version is a fair test before any company-wide rollout: one objective for the quarter, three key results with real numbers, posted where the team works, and a fifteen-minute grading conversation at the quarter’s end that asks only two questions — what did the numbers say, and what did we learn about how we set them. If that small loop cannot be run honestly, no software and no consultant will save the larger one. If it can, you have built the thing Grove actually designed: not a form, but a rhythm by which a group of people agrees, out loud and on a schedule, about what matters — and then checks.

← All Insights

© 2026 Cognosco Learning. All rights reserved.