Measure What Matters: How Google, Bono, and the Gates Foundation Rock the World with OKRs

by John Doerr

What is a “cycle”?

A few goal-setting ground rules: Key results should be succinct, specific, and measurable. A mix of outputs and inputs is helpful. Finally, completion of all key results must result in attainment of the objective. If not, it’s not an OKR.*

Share with Gordon

the OKR’s guardian,

What is this role? Does every team OKR need an individual guardian?

Reaching one billion hours was a game of inches; our engineers were hunting for changes that might yield as little as 0.2 percent more watch time. In 2016 alone, they would find around 150 of those tiny advances. We’d need nearly all of them to reach our objective.

Stretch OKRs tend to set powerful forces into motion, and you can never be sure where they’ll lead.

the contemporary alternative to annual reviews, is continuous performance management. It is implemented with an instrument called CFRs, for: Conversations: an authentic, richly textured exchange between manager and contributor, aimed at driving performance Feedback: bidirectional or networked communication among peers to evaluate progress and guide future improvement Recognition: expressions of appreciation to deserving individuals for contributions of all sizes

“If a conversation is limited to whether you achieved the goal or not, you lose context. You need continuous performance management to surface the critical questions: Was the goal harder to achieve than you’d thought when you set it? Was it the right goal in the first place? Is it motivating? Should we double down on the two or three things that really worked for us last quarter, or is it time to consider a pivot? You need to elicit those insights from all over the organization.

When companies replace—or at least augment—the annual review with ongoing conversations and real-time feedback, they’re better able to make improvements throughout the year. Alignment and transparency become everyday imperatives. When employees are struggling, their managers don’t sit and wait for some scheduled day of reckoning. They jump into tough discussions like firefighters, without hesitation.

We’re constantly surrounded by positive reinforcement and feedback, but many of us haven’t been trained to seek it out. Say you give a presentation to your team. After the fact, somebody comes up to you and says, ‘Hey, nice job.’ Most of us would say, ‘Oh great, thanks,’ and move on. But we want to probe a little deeper: ‘Thank you. What one thing did you like about it?’ The idea is to capture more specific feedback in real time.”

Structure for O3s

Not so many years ago, employees made their voices heard by slipping unsigned notes into the office suggestion box. Today, progressive companies have replaced the box with always-on, anonymous feedback tools, from quick-hitting employee surveys to anonymous social networks and even rating apps for meetings and meeting organizers.

Peer-to-peer (or 360-degree) feedback is an added lens for continuous performance management. It can be anonymous or public or somewhere in between. Is the feedback designed to help employees move forward in their careers? (If so, it’s channeled privately to the individuals.) Is it meant to reveal an organization’s problem areas? (Here it goes straight to HR.) It’s all a matter of context and purpose.

As OKRs are combined with 360-degree feedback, the silo will soon be a relic of the past.

a continuous performance management system has three requirements. The first is executive support. The second is clarity on company objectives and how they align with individual priorities—as set out in our “goals and expectations,” which equate to OKRs. The third is an investment in training to equip managers and leaders to be more effective. We’re not shipping people out to courses. We’re steering them to one-hour sessions online, with role-played vignettes: “Do you need to give difficult feedback? Here are the steps.” Corrective feedback is naturally difficult for people. But when done well, it’s also the greatest gift you can give to someone—because it can change people’s mindset and modify their behavior in the most positive, valuable way. We’re creating an environment where people say, “You know what? It’s okay to make a mistake, because that’s how I’m going to grow the most.” That’s a big part of our culture change.

It’s hard to deny the explicit value of OKRs, like how they help tie an organization to the leadership’s true ambitions. But for young companies like Zume, especially, there’s an equally important implicit value that gets overlooked. OKRs are a superb training tool for executives and managers. They teach you how to manage your business within existing limits.

What’s neat about OKRs is that they formalize reflection. At least once each quarter, they make contributors step back into a quiet place and consider how their decisions align with the company. People start thinking in the macro. They become more pointed and precise, because you can’t write a ninety-page OKR dissertation. You have to choose three to five things and exactly how they should be measured. Then when the day comes and someone says, “Okay, you’re a manager,” you’ve already learned how to think like one. And that’s huge.

As people get more familiar with the OKR process, it naturally gets more collaborative. In Q3 of 2016, Alex and I wrote the top company OKRs, and the department heads converted some of our key results into their own objectives. We just cascaded them down. In Q4, the two of us still wrote the company objectives, but our team jumped in and KRed them from the top, which was great. They took on a more creative role, and the OKRs got better. Our goals were still stretched, but people felt they were more realistic.

Every organization has people who are more vocal in asserting themselves. If they don’t win their point the first time, they’re comfortable saying it again. But quieter folks may not be heard so well, and their needs can get neglected. The OKR framework gives equal voice and weight to each department. No one needs to suffer in silence—truthfully, no one has that option. Your objectives will get their turn up on the screen, like everyone else’s, for comment and support.

Before rolling OKRs out to our individual contributors, we put in two full quarters at the exec level. We had to establish the culture first. What we’ve found, oddly enough, is that our most active participants are the ones who were initially most skeptical.

OKRs had an effect on me I didn’t expect. When I did my biweekly check-ins, it gave me a couple minutes to think about what I was doing, and how my goals rolled up to what the company needed for the quarter.

those OKR meditations helped me reset my compass: How do I contribute to the scheme of things? Then it’s not just another report or campaign or field event. It connects to something bigger and more meaningful.

the process forced us to clarify who’s in charge of what. When a fly ball is hit between two outfielders, somebody’s got to call for it—or else the ball drops in, or both people dive for it and crash into each other. Early on, our fielders were marketing and product—but who was responsible for Zume’s revenue targets? The two leads had been with us one month apiece. Not only were they new to OKRs, but they were also new to Zume, and Zume was new to itself. When Alex and I saw their confusion, we broke the objective out into new revenue (marketing) and repeat revenue (product), and the department heads drilled down from there. That was an important conversation. It wasn’t linked to an objective, per se, but it was absolutely a by-product of an early-stage OKR process. When something isn’t clearly delineated, it shows up right away. You can’t miss it.

The seven-member leadership team meets over lunch every Monday, and about every other week we discuss our OKRs. You hear people saying things like “Who owns the customer?” or “How would you key-result that goal?” And everyone knows just what they mean.

Since our company has total alignment, the entire team has already agreed to the key result and the dependency it entails. There’s no judgment, just a problem to be solved. And guess what else happens? The two leads will advocate for each other to get more resources

Every two weeks, each person at Zume has a one-hour, one-on-one conversation with whomever they report to. (Julia and I converse with each other.) It’s a sacred time. You cannot be late; you cannot cancel. There’s only one other rule: You don’t talk about work. The agenda is you, the individual, and what you are trying to accomplish personally over the next two to three years, and how you’re breaking that into a two-week plan. I like to start with three questions: What makes you very happy? What saps your energy? How would you describe your dream job?

It may seem paradoxical, but these nonwork, touchpoint one-on-ones are a forum for ongoing performance feedback. In talking about people’s pursuit of personal goals, you end up learning a lot about what moves them forward—or holds them back—in their careers.

Every new employee goes through mission and values training as part of their onboarding. Alex and I are very clear about what we expect from people. And our clarity forces us to be highly accountable, as an organization and as individuals. We have a best-idea-wins culture, and people are free to call out anybody, including the CEO. Alex: Especially the CEO, that’s the best call-out there is. When people challenge us in an open forum, we always stop and make a huge big deal about how impressive it is that the person spoke up. We try to overdo it, to create permission for people to lean in.

when you know your company objectives like you know your last name, it’s very calming. OKRs help me to be that focused, clear-headed leader. No matter how crazy things get, I can always default back to what matters.

OKRs are clear vessels for leaders’ priorities and insights. CFRs help ensure that those priorities and insights get transmitted. But goals cannot be attained in a vacuum. Like sound waves, they require a medium. For OKRs and CFRs, the medium is an organization’s culture, the living expression of its most cherished values and beliefs.

By aligning teams to work toward a handful of common objectives, then uniting them through lightweight, goal-oriented communications, OKRs and CFRs create transparency and accountability, the tent poles for sustained high performance. Healthy culture and structured goal setting are interdependent. They’re natural partners in the quest for operating excellence.

In Project Aristotle, an internal Google study of 180 teams, standout performance correlated to affirmative responses to these five questions: Structure and clarity: Are goals, roles, and execution plans on our team clear? Psychological safety: Can we take risks on this team without feeling insecure or embarrassed? Meaning of work: Are we working on something that is personally important for each of us? Dependability: Can we count on each other to do high-quality work on time? Impact of work: Do we fundamentally believe that the work we’re doing matters?

In the high-stakes arena of culture change, OKRs lend us purpose and clarity as we plunge into the new. CFRs supply the energy we need for the journey. Where people have authentic conversations and get constructive feedback and recognition for superior accomplishment, enthusiasm becomes infectious. The same goes for stretch thinking and a commitment to daily improvement. The companies that treat their people as valued partners are the ones with the best customer service. They have the best products and strongest sales growth. They’re the ones who are going to win.

A rulebook can tell me what I can or can’t do. I need culture to tell me what I should do.”

Lumeris seemed a natural fit for Andy Grove’s goal-setting system. But as Andrew Cole, the former head of HR, will tell you, adaptation was anything but simple. If cultural barriers go unaddressed, as Andrew says, “The antibodies will be set loose and the body will reject the donor organ of OKRs.”

Few of our people understood the business rationale behind OKRs.

We were missing explicit buy-in from executive leadership.

nobody held anybody accountable for getting t...

No one had addressed a basic question everyone at Lumeris was asking: “What’s in this for me?” Though the OKR program was sincerely intended to improve goal setting and collaborative communication, people didn’t trust it. Unless we changed the environment, it couldn’t possibly succeed.

HR can be a potent vehicle for operating excellence. It’s also the place where culture change is crystallized—at the end of the day, culture is about the people you recruit and the values they bring to bear. While Lumeris had its share of A and B players in middle management, there were also C players and below who’d been hired with erroneous criteria and vague interviews. There is no tool, OKRs included, that will work with the wrong instruction manual.

Once senior management and frontline employees were fully on board, we tackled the tougher nut: strengthening middle management. That’s typically a three-year process, from start to steady state. When it’s complete, your new culture is assured.

Goal setting is more art than science. We weren’t just teaching people how to refine an objective or a measurable key result. We had a cultural agenda, as well. Why is transparency important? Why would you want people across other departments to know your goals? And why does what we’re doing matter? What is true accountability? What’s the difference between accountability with respect (for others’ failings) and accountability with vulnerability (for our own)? How can OKRs help managers “get work done through others”? (That’s a big factor for scalability in a growing company.) How do we engage other teams to adopt our objective as a priority and help assure that we reach it? When is it time to stretch a team’s workload—or to ease off on the throttle? When do you shift an objective to a different team member, or rewrite a goal to make it clearer, or remove it completely? In building contributors’ confidence, timing is everything.

“OKRs make you focus on working on the business, instead of just working in the business,”

Objectives are the “Whats.” They: express goals and intents; are aggressive yet realistic; must be tangible, objective, and unambiguous; should be obvious to a rational observer whether an objective has been achieved. The successful achievement of an objective must provide clear value for Google.

Key Results are the “Hows.” They: express measurable milestones which, if achieved, will advance objective(s) in a useful manner to their constituents; must describe outcomes, not activities. If your KRs include words like “consult,” “help,” “analyze,” or “participate,” they describe activities. Instead, describe the end-user impact of these activities: “publish average and tail latency measurements from six Colossus cells by March 7,” rather than “assess Colossus latency”; must include evidence of completion. This evidence must be available, credible, and easily discoverable. Examples of evidence include change lists, links to docs, notes, and published metrics reports.

Cross-team OKRs should include all the groups who must materially participate in the OKR, and OKRs committing to each group’s contribution should appear explicitly in each such group’s OKRs. For example, if Ads Development and Ads SRE and Network Deployment must deliver to support a new ads service, then all three teams should have OKRs describing their commitment to deliver their part of the project.

Commitments are OKRs that we agree will be achieved, and we will be willing to adjust schedules and resources to ensure that they are delivered. The expected score for a committed OKR is 1.0; a score of less than 1.0 requires explanation for the miss, as it shows errors in planning and/or execution. By contrast, aspirational OKRs express how we’d like the world to look, even though we have no clear idea how to get there and/or the resources necessary to deliver the OKR. Aspirational OKRs have an expected average score of 0.7, with high variance.

TRAP #1: Failing to differentiate between committed and aspirational OKRs. Marking a committed OKR as aspirational increases the chance of failure. Teams may not take it seriously and may not change their other priorities to focus on delivering the OKR. On the other hand, marking an aspirational OKR as committed creates defensiveness in teams who cannot find a way to deliver the OKR, and it invites priority inversion as committed OKRs are de-staffed to focus on the aspirational OKR. TRAP #2: Business-as-usual OKRs. OKRs are often written principally based on what the team believes it can achieve without changing anything they’re currently doing, as opposed to what the team or its customers really want. TRAP #3: Timid aspirational OKRs. Aspirational OKRs very often start from the current state and effectively ask, “What could we do if we had extra staff and got a bit lucky?” An alternative and better approach is to start with, “What could my [or my customers’] world look like in several years if we were freed from most constraints?” By definition, you’re not going to know how to achieve this state when the OKR is first formulated—that is why it is an aspirational OKR. But without understanding and articulating the desired end state, you guarantee that you are not going to be able to achieve it. The litmus test: If you ask your customers what they really want, does your aspirational objective meet or exceed their request? TRAP #4: Sandbagging. A team’s committed OKRs should ...

TRAP #5: Low Value Objectives (aka the “Who cares?” OKR). OKRs must promise clear business value—otherwise, there’s no reason to expend resources doing them. Low Value Objectives (LVOs) are those for which, even if the Objective is completed with a 1.0, no one will notice or care. A classic (and seductive) LVO example: “Increase task CPU utilization by 3 percent.” This objective by itself does not help users or Google directly. However, the (presumably related) goal, “Decrease quantity of cores required to serve peak queries by 3 percent with no change to quality/latency/ . . . and return resulting excess cores to the free pool” has clear economic value. That’s a superior objective. Here is a litmus test: Could the OKR get a 1.0 under reasonable circumstances without providing direct end-user or economic benefit? If so, then reword the OKR to focus on the tangible benefit. A classic example: “Launch X,” with no criteria for success. Better: “Double fleet-wide Y by launching X to 90+ percent of borg cells.” TRAP #6: Insufficient KRs for committed Os. OKRs are divided into the desired outcome (the objective) and the measurable steps required to achieve that outcome (the key results). It is critical that KRs are written such that scoring 1.0 on all key results generates a 1.0 score for the objective. A common error is writing key results that are necessary but not sufficient to collectively complete the objective. The error is tempting because it allows a team to avoid the di...

Reading, Interpreting, and Acting on OKRs For committed OKRs Teams are expected to rearrange their other priorities to ensure an on-schedule 1.0 delivery. Teams who cannot credibly promise to deliver a 1.0 on a committed OKR must escalate promptly. This is a key point: Escalating in this (common) situation is not only OK, it is required. Whether the issue arose because of disagreement about the OKR, disagreement about its priority, or inability to allocate enough time/people/resources, escalation is good. It allows the team’s management to develop options and resolve conflicts. The corollary is that every new OKR is likely to involve some amount of escalation, since it requires a change to existing priorities and commitments. An OKR that requires no changes to any group’s activities is a business-as-usual OKR, and those are unlikely to be new—although they may not have previously been written down. A committed OKR that fails to achieve a 1.0 by its due date requires a postmortem. This is not intended to punish teams. It is intended to understand what occurred in the planning and/or execution of the OKR, so that teams may improve their ability to reliably hit 1.0 on committed OKRs. Examples of classes of committed OKRs are ensuring that a service meets its SLA (service level agreement) for the quarter; or delivering a defined feature or improvement to an infrastructure system by a set date; or manufacturing and delivering a quantity of servers at a cost point.