What's a Good OKR?

What’s a Good OKR?

OKRs have two parts:

• an Objective
• and a set of Key Results.

What’s an Objective?

An Objective is the what. It’s the goal that you and your team are trying to accomplish. Once you articulate that Objective, it becomes the rallying point for your team. Then, each Objective has a set of three to five Key Results.

If Objectives are the what, Key Results are the how – as in how we’ll prove we’ve gotten there. A great set of Key Results are the benchmarks of success. They help measure how much progress is being made and they set the finish line.

By definition, an Objective is accomplished when you fulfill all of its Key Results. Let’s get a bit more specific.

OKRs represent what you want to achieve, the outcome you seek. They describe what you want to create, change, or improve. OKRs are not empty inspirational statements. They’re not about doing a little more than we’re already doing today. They lay out significant challenges.

Characteristics of a good Objective

John says a good Objective is: significant, concrete, action-oriented, and inspirational.

Significant: The Objective represents a top priority that would be a meaningful improvement from where you are today.

Concrete: The Objective is clear and specific enough to generate action. People immediately understand how it affects their day-to-day work. For example, a team with the Objective to “Build the world’s fastest browser” would approach its work differently than one with an Objective to “Build the world’s most secure browser.” Fastest browser. Secure browser.

Action-oriented: The action should be implied in the way the Objective is written. Consider how teams might act differently if the Objective is to (imagine if we had a city): “Make our city center more accessible” than if the Objective were to “Increase storefront revenue in our city center.”

Inspiring: Objectives are simple, memorable, and move you to act. For example, “Make the world’s lowest carbon footprint shoe.”

Once you’ve established meaningful, audacious, and inspiring Objectives (the Os), it’s time to add the KRs (the Key Results).

Characteristics of good Key Results

What are good Key Results?

First, they’re specific and time-bound. They say we’re going to get this far in this amount of time. They avoid fuzzy numbers or vague due dates.

Second, they’re aggressive, yet realistic. We want our teams to stretch because that’s what creates meaningful change.

Finally, Key Results are both measurable and verifiable. Anyone on the team can check the status of the Key Results – which also means that everyone is accountable.

OKR example: Healthcare.gov

I first came across OKRs in October 2013, when Healthcare.gov was launched so people could get healthcare through the Affordable Care Act. This groundbreaking legislation vastly expanded insurance coverage. But on opening day, things went terribly wrong. Hundreds of thousands of people who were trying to apply for health insurance were frustrated by this glitchy website.

I was one of six people asked to help. Now, of course, six people alone can’t fix a project that thousands of people working on, with more than a billion dollars behind it. It was complicated, and everybody had a different idea about what we should do.

I remember the night when Mikey Dickerson, my colleague and an ex-Googler, focused our energies into one overarching rallying cry: In the end, what were we all trying to accomplish?

He said, “We have to fix the website for the vast majority of the people. That’s measured by 7 out of 10 people being able to get through; a one-second response time; a 1% error rate; and 99% uptime.”

And he wrote it all on the board.

Those four measures/items were how we’d know – at least in that moment – that we had “fixed the website for the vast majority.” That stayed on the white board, guiding the rescue effort – ultimately providing millions of people access to affordable health care.

He didn’t label these points as an Objective and a set of Key Results. But OKRs had become part of Mikey’s vernacular, and the rest of us benefited from his experience. Add the O and the KRs and it looks like this:

Objective: Fix healthcare dot gov for the vast majority of people.

Key Results:
KR1: 70% of people get through.
KR2: One-second response time.
KR3: 1% error rate.
KR4: 99% uptime.

If an OKR is written properly, the set of Key Results adds up to meeting the Objective. Each Key Result is essential to the OKR’s success.

Now let’s take a look at what makes this such an elegant OKR.

First, the Objective: Fix Healthcare.gov for the vast majority of people.

Is it significant? Absolutely. The site’s broken. It needs to be fixed so people can get healthcare.

Is it concrete? Yes. “Fix the website” is a very tangible statement. We know what’s being asked of us.

Is it action-oriented? Yes. “Fix” is a clear, actionable verb, We’re not building something new, we are FIXING it.

Is it inspiring? Totally. Because the Objective was so concise and simple, it became our North Star that we turned to time and time again. And the “why” is embedded in the statement. We’re not just fixing the site; we’re fixing it for people. They are our “why”.

Now, the Key Results.

Are they specific? Yes. Not much is left to the imagination with “1% error rate.”

Are they time-bound? In this case, the time frame was implied. We knew we were using month-long cycles.

Are they aggressive, yet realistic? Well – let me tell you – taking a site that was crashing daily to a one-second response time was definitely aggressive. But we also knew if we organized our efforts, we could get there.

Are they measurable and verifiable? These Key Results are indeed measurable. But when we first wrote the OKR, no one was collecting all the data needed to track these KRs. So we had to start by collecting it manually. Later on, we built automated dashboards so everyone on the team could see our progress.

Those are the basics. As you write your OKRs cycle after cycle, you strengthen your goal-setting muscle, and get more of the benefits of this goal-setting system.