Objectives and key results

Based on Wikipedia: Objectives and key results

In 1999, a venture capitalist walked into Google's offices and taught the two young founders a management technique that would help turn their search engine into one of the most valuable companies in human history. The technique wasn't sophisticated. It wasn't proprietary. It had been developed at Intel decades earlier, and its core insight could fit on an index card.

Set ambitious goals. Define three to five measurable ways to know if you've achieved them. Repeat every quarter.

That's it. That's the system that Larry Page credits with enabling "ten times growth, many times over." It's called Objectives and Key Results, or OKRs, and its simplicity is both its greatest strength and the source of endless confusion about what it actually means to do it right.

The anatomy of an OKR

An Objective is a goal that matters. It should be significant enough that achieving it would genuinely change something about your situation. It should be concrete, meaning you could explain it to someone outside your organization and they'd understand what you're trying to accomplish. And it should be clearly defined, leaving no ambiguity about what success looks like.

But here's what often gets lost: objectives should also be inspirational. They're not tasks to check off. They're destinations worth traveling toward. "Increase revenue by fifteen percent" is measurable, but it doesn't exactly make anyone leap out of bed in the morning. "Become the company that small business owners recommend to their friends" points in the same direction but gives people a reason to care.

Key Results are how you know whether you've reached your objective. They're the evidence. You typically want three to five of them per objective, and they need to be ruthlessly measurable. Not "improve customer satisfaction" but "achieve a Net Promoter Score of forty-five or higher." Not "launch the new product" but "ship version one to ten thousand beta users by March thirty-first."

The critical requirement is that there should be no grey area. When the quarter ends, you should be able to look at each key result and answer with a simple yes or no: did we hit this number or didn't we? As Andy Grove himself put it, "No judgments in it."

The Intel origins

Andy Grove didn't invent goal-setting, of course. Management by Objectives, or MBO, had been around since Peter Drucker popularized it in the 1950s. What Grove did at Intel in the nineteen seventies was strip away the bureaucracy and make the system actually work.

Traditional MBO had become synonymous with annual reviews and paperwork. Managers would set objectives at the beginning of the year, file them away, and pull them out twelve months later to see how things went. It was retrospective rather than operational. Grove's innovation—which he initially called "Intel Management by Objectives" or iMBOs—was to make the cycle quarterly and to insist that objectives connect visibly to measurable outcomes.

He documented this approach in his 1983 book High Output Management, which became a quiet classic in Silicon Valley. For years, the system remained mostly an Intel thing, known to those who'd worked there but not widely adopted elsewhere.

The Google amplification

The turning point came through John Doerr. In 1975, Doerr was a young salesperson at Intel who happened to take an internal course taught by Grove. He learned about iMBOs and tucked the knowledge away.

Two decades later, Doerr had become one of the most influential venture capitalists in technology, a partner at Kleiner Perkins who had backed companies like Amazon and Netscape. When he invested in Google in 1999, he inherited a seat on the board and noticed that the two founders, Larry Page and Sergey Brin, were brilliant technologists but inexperienced managers trying to scale a company that was growing faster than they could keep up with.

Doerr introduced them to OKRs. The framework gave Google a way to stay aligned as it exploded from a few dozen employees to tens of thousands. Instead of cascading commands from the top down, the company could set clear objectives and let teams figure out how to achieve them. Everyone could see what everyone else was working on. It became, in Google's telling, "a management methodology that helps to ensure that the company focuses efforts on the same important issues throughout the organization."

Google's visible success made OKRs famous. LinkedIn adopted them. Then Twitter, Uber, Microsoft, and GitLab. Doerr wrote a book about the framework in 2018 called Measure What Matters, which became a bestseller and brought OKRs to thousands of organizations that had never heard of Intel's internal management courses.

The seventy percent rule

Here's where OKRs diverge sharply from how most people think about goals. In school, seventy percent is a C. In most performance reviews, hitting seventy percent of your targets means you underperformed. In OKRs, seventy percent is exactly where you want to be.

Doerr's recommendation is that organizations should aim for a seventy percent success rate on their key results. Not one hundred percent. The logic is counterintuitive but important: if you're hitting all your targets, your targets aren't ambitious enough.

OKRs are meant to stretch people. They're supposed to be uncomfortable. If a team consistently achieves one hundred percent of their key results, they should be asking themselves whether they're really pushing toward what's possible or just codifying what they were going to do anyway.

This leads to an important distinction between two types of key results. "Aspirational" key results are the stretch goals where seventy percent—or even sixty percent—counts as success. These are the moonshots. "Committed" key results are binary: did you ship the feature or didn't you? Did you meet the deadline or didn't you? These should hit one hundred percent because they represent things the organization has genuinely committed to delivering.

On the standard scoring scale of zero to one, an aspirational key result at zero-point-seven is healthy. A committed key result should hit one-point-zero.

The language of ambition

Words matter in OKRs, and certain words are red flags.

"Help" is a problem. "Help improve the customer experience" is not an objective—it's an excuse to claim partial credit for whatever happens. You can always argue that you helped. The question is whether you actually moved the needle, and that requires specificity about what moving the needle means.

"Consult" has the same issue. So does "support" or "contribute to." These verbs describe activities, not outcomes. They let people stay busy without being accountable for results.

Equally dangerous is writing objectives that describe business as usual. If your objective is "maintain current customer retention rates," you haven't set a goal—you've described the status quo. Objectives should represent a change in state, a movement from where you are to somewhere better. They should be action-oriented, pointing toward something you have to go get rather than something you're already doing.

Leading versus lagging

One of the subtler mistakes in writing key results is measuring the wrong kind of indicator. The distinction between leading and lagging indicators comes from economics but applies directly to goal-setting.

A lagging indicator tells you what already happened. Revenue is a lagging indicator. Customer churn is a lagging indicator. By the time you see the number, the events that created it are in the past. You can learn from them, but you can't change them.

A leading indicator tells you what's likely to happen. If you're trying to improve customer retention, the number of support tickets resolved within twenty-four hours is a leading indicator. If you're trying to grow revenue, the number of qualified sales conversations happening this week is a leading indicator. These metrics give you early warning when something is going wrong, while there's still time to course-correct.

The recommendation is to bias key results toward leading indicators wherever possible. It's more useful to know you're heading toward a cliff while you can still turn the wheel than to learn about it after you've gone over the edge.

The level question

When OKRs first gained popularity outside Intel and Google, the standard advice was to set them at three levels: organizational, team, and individual. The company would define its objectives. Teams would define objectives that supported the company's objectives. Individuals would define objectives that supported their team's objectives. Everything would cascade neatly from top to bottom.

This approach came partly from a widely-watched Google Ventures video from 2014, where Rick Klau explained how Google used OKRs at all three levels. It seemed like the official playbook.

Then something interesting happened. In 2017, Klau posted a clarification: "Skip individual OKRs altogether. Especially for younger, smaller companies. They're redundant. Focus on company- and team-level OKRs."

What changed? Practice revealed problems with individual-level OKRs. They tended to look suspiciously like task lists—do this project, finish that analysis, attend these meetings. They conflated goal-setting with performance reviews, creating confusion about whether OKRs were supposed to measure what you achieved or determine your annual raise.

Worse, the cascading structure could recreate exactly what OKRs were designed to avoid: a top-down waterfall where every goal was predetermined by the level above it, leaving no room for teams to identify opportunities that leadership hadn't anticipated. The whole point was to align around outcomes while giving people autonomy over how to achieve them. Individual OKRs often just added bureaucracy.

What OKRs are not

Understanding any framework requires knowing what it isn't.

OKRs are not Key Performance Indicators, or KPIs, though the terms are often confused. KPIs are ongoing metrics you track to monitor the health of the business—things like revenue growth, customer acquisition cost, or employee satisfaction scores. They persist from quarter to quarter. OKRs are time-bound goals you're actively trying to achieve. You set them, you pursue them, and when the quarter ends, you score them and set new ones.

OKRs are not project plans. They don't tell you what to do. They tell you where you're trying to get to. How you get there is a separate question that might involve multiple projects, experiments, and pivots along the way.

OKRs are not performance review criteria, though many organizations make the mistake of treating them that way. When key results become tied directly to compensation or career advancement, people start gaming the system by setting easily achievable goals. The seventy percent target becomes impossible to maintain because no one wants to volunteer for objectives they might miss.

And OKRs are not magic. They're a communication and alignment tool. They help organizations stay focused on what matters and create shared language for discussing progress. But they don't replace strategy, or execution, or all the hard work of actually building something valuable. Badly-chosen OKRs will efficiently focus an organization on the wrong things.

The broader landscape

OKRs exist within a family of goal-setting and measurement frameworks, each with its own emphasis.

The Balanced Scorecard, developed in the early 1990s, takes a broader view by asking organizations to track goals across four perspectives: financial, customer, internal process, and learning and growth. It's more comprehensive than OKRs but also more complex.

SMART criteria—Specific, Measurable, Achievable, Relevant, Time-bound—provide a checklist for evaluating whether a goal is well-formed. OKRs naturally incorporate most of these criteria, though the "Achievable" part sits in tension with the stretch-goal philosophy.

The Goal Question Metric approach, or GQM, comes from software engineering and works backward from goals to questions to metrics. It's particularly useful when you're not sure what to measure.

OGSM—Objectives, Goals, Strategies, and Measures—adds strategic choices to the framework, asking not just what you want to achieve but how you plan to achieve it.

None of these frameworks is definitively better than the others. They're tools, and different tools suit different situations. OKRs have become dominant in technology companies partly because of Google's influence and partly because their simplicity matches the fast iteration cycles of software development. A manufacturing company or a government agency might find other frameworks more appropriate.

The enduring appeal

What makes OKRs persist, decades after Andy Grove first sketched them out at Intel, is that they solve a real problem that every growing organization faces: how do you keep hundreds or thousands of people working toward the same goals without turning into a bureaucracy?

Traditional management approaches answer this question with hierarchy and control. Senior leaders decide what to do. Middle managers break it down into tasks. Workers execute the tasks. Information flows down; results flow up.

This works, to a point. But it's slow, and it wastes the intelligence of everyone below the top of the org chart. The people closest to customers often have the best insights about what would actually help them. The engineers building the product often see opportunities that executives never would. Command-and-control structures struggle to capture this distributed knowledge.

OKRs offer a different answer. Define where you're trying to go. Define how you'll know if you get there. Then trust people to figure out the path.

It's not a complete management system. It doesn't tell you how to hire, or how to give feedback, or how to make decisions when people disagree. But as a tool for alignment—for making sure that everyone's efforts add up to something coherent—it has proven remarkably durable.

Larry Page's endorsement from the foreword of Doerr's book captures both the power and the humility of the approach:

OKRs have helped lead us to ten times growth, many times over. They've helped make our crazily bold mission of 'organizing the world's information' perhaps even achievable. They've kept me and the rest of the company on time and on track when it mattered the most.

Not "OKRs made us successful." Not "OKRs are the secret." Just: they helped. They kept us on track. When it mattered.

That's probably the right way to think about any management framework. Not as a solution, but as a help. Not as a guarantee, but as a tool that might keep you on track when it matters most.