Measure What Matters

by John Doerr

In 1968, the year Intel opened shop, a psychology professor at the University of Maryland cast a theory that surely influenced Andy Grove. First, said Edwin Locke, “hard goals” drive performance more effectively than easy goals. Second, specific hard goals “produce a higher level of output” than vaguely worded ones. In the intervening half century, more than a thousand studies have confirmed Locke’s discovery as “one of the most tested, and proven,

again—in print this time, and in enough depth to do the subject justice. This book—with its companion website, whatmatters.com—is my chance to bring a long-held passion to you, my reader. I hope

Through the Andy Grove era, OKRs were Intel’s lifeblood. They stood front and center at weekly one-on-ones, biweekly staff meetings, monthly and quarterly divisional reviews. That was how Intel managed tens of thousands of people to etch a million lines of silicon or copper to within a millionth of a meter in accuracy.

Less is more. “A few extremely well-chosen objectives,” Grove wrote, “impart a clear message about what we say ‘yes’ to and what we say ‘no’ to.” A limit of three to five OKRs per cycle leads companies, teams, and individuals to choose what matters most. In general, each objective should be tied to five or fewer key results.

ENGINEERING DEPARTMENT OBJECTIVE (Q2 1980) Deliver 500 8MHz 8086 parts to CGW by May 30. KEY RESULTS 1. Develop final art to photo plot by April 5. 2. Deliver Rev 2.3 masks to fab on April 9. 3. Test tapes completed by May 15. 4. Fab red tag start no later than May 1.

What is most important for the next three (or six, or twelve) months? Successful organizations focus on the handful of initiatives that can make a real difference, deferring less urgent ones.

Bill Campbell, the Intuit CEO who later coached the Google executive team: “When you’re the CEO or the founder of a company … you’ve got to say ‘This is what we’re doing,’ and then you have to model it. Because if you don’t model it, no one’s going to do it.”

There is no religion to this protocol, no one-size-fits-all. An engineering team might opt for six-week OKR cycles to stay in sync with development sprints. A monthly cycle could do the trick for an early-stage company still finding its product-market fit. The best OKR cadence is the one that fits the context and culture of your business.

A few goal-setting ground rules: Key results should be succinct, specific, and measurable. A mix of outputs and inputs is helpful.

OKR.

one thing an [OKR] system should provide par excellence is focus. This can only happen if we keep the number of objectives small …. Each time you make a commitment, you forfeit your chance to commit to something else. This, of course, is an inevitable, inescapable consequence of allocating any finite resource. People who plan have to have the guts, honesty, and discipline to drop projects as well as to initiate them, to shake their heads “no” as well as to smile “yes.” … We must realize—and act on the realization—that if we try to focus on everything, we focus on nothing. Above all, top-line

OBJECTIVE Continue to build a world-class team. KEY RESULTS 1. Recruit 10 engineers [0.8]. 2. Hire commercial sales leader [1.0]. 3. One hundred percent of candidates feel they had a well-organized, professional experience even if Nuna does not extend an offer [0.5].

organization.

So we switched it up. We began pinning our key results to deadlines instead of revenue or projected users. (Example: “Launch MFP Premium by 5/1/15.”) After a feature launched and some real data came back, we’d be in a stronger position to assess its impact and potential. Then our next round of OKRs could be more realistically keyed (or stretched) to projected outputs.

Early on, I found it challenging to separate my individual goals from the department’s OKRs. As IT’s leader, I thought they should logically coincide. But it wasn’t a good optic. Most of our top-level objectives endured from quarter to quarter, typically for eighteen months. Down the line, teams and individuals would modify their own OKRs as the environment shifted and we kept making progress. And they were asking, quite reasonably, “What is the CIO doing if his goals never change?” I got the message. Now I have my own objectives, and I ladder up to our top-level OKRs like everyone

In God we trust; all others must bring data. —W. Edwards Deming

Salesforce, JIRA, and Zendesk. With three or four clicks, users can navigate a digital dashboard to create, track, edit, and score their OKRs. These platforms deliver transformative OKR values:

OKRs are adaptable by nature. They’re meant to be guardrails, not chains or blinders. As we track and audit our OKRs, we have four options at any point in the cycle: • Continue: If a green zone (“on track”) goal isn’t broken, don’t fix it. • Update: Modify a yellow zone (“needs attention”) key result or objective to respond to changes in the workflow or external environment. What could be done differently to get the goal on track? Does it need a revised time line? Do we back-burner other initiatives to free up resources for this one? • Start: Launch a new OKR mid-cycle, whenever the need arises. • Stop: When a red zone (“at risk”) goal has outlived its usefulness, the best solution may be to drop it.

Google uses a scale of 0 to 1.0: • 0.7 to 1.0 = green.fn2 (We delivered.) • 0.4 to 0.6 = yellow. (We made progress, but fell short of completion.) • 0.0 to 0.3 = red. (We failed to make real progress.)

Intel followed a similar formula. You may recall the OKR for Operation Crush, the company’s push to reclaim the microprocessor market. Here are Andy Grove’s actual marching orders from Q2 1980, as endorsed by his executive team (with end-of-quarter grades in brackets): CORPORATE OBJECTIVE Establish the 8086 as the highest-performance 16-bit microprocessor family, as measured by: KEY RESULTS (Q2 1980) 1. Develop and publish five benchmarks showing superior 8086 family performance [0.6]. 2. Repackage the entire 8086 family of products [1.0]. 3. Get the 8MHz part into production [0]. 4. Sample the arithmetic coprocessor no later than June 15 [0.9]. And here is how these scores were determined: • We completed three of five benchmarks for an 0.6, a borderline green. • We did indeed repackage the 8086 family, under a new product line called iAPX. So that’s a perfect 1.0. • Production of the 8MHz part, set for early May, was a fiasco.fn3 Because of problems with the polysilicon, the target had to be pushed to October. That’s a zero. • As for the arithmetic coprocessor, the goal was to ship 500 parts by June 15. We wound up shipping 470—which computes to 0.9, another green.

Altogether, we averaged 62.5 percent (or a raw score of 0.625) on our KRs for this objective, a respectable grade. The Intel board judged it b...

things.

“self-actualization”:

In Wired, Steven Levy elaborated: The way Page sees it, a ten percent improvement means that you’re doing the same thing as everybody else. You probably won’t fail spectacularly, but you are guaranteed not to succeed wildly.

Google to be “uncomfortably excited.” He wanted us to have “a healthy disregard for the impossible.”

As Microsoft CEO Satya Nadella has pointed out: In a world where computing power is nearly limitless, “the true scarce commodity is increasingly human attention.” When users spend more of their valuable time watching YouTube videos, they must perforce be happier with those videos. It’s a virtuous circle: More satisfied viewership (watch time) begets more advertising, which incentivizes more content creators, which draws more viewership.

That transformational system, 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

Table 15.1: Annual Performance Management Versus Continuous Performance Management Annual Performance Management Continuous Performance Management Annual feedback Continuous feedback Tied to compensation Decoupled from compensation Directing/autocratic Coaching/democratic Outcome focused Process focused Weakness based Strength based Prone to bias Fact driven

“The first is a set of monthly one-on-one conversations between employees and their managers about how things are going. “The second is a quarterly review of progress against our OKRs. We sit down and say, ‘What did you set out to accomplish this quarter? What were you able to do—and what weren’t you able to do? Why or why not? What can we change?’ “Third, we have a semiannual professional development conversation.

Employees talk about their career trajectory—where they’ve been, where they are, where they want to go. And how their managers and the organization can support their new direction. “The fourth bit is ongoing, self-driven insight. We’re constantly surrounded by positive

The second is an ongoing, forward-looking dialogue between leaders and contributors. It centers on five questions: • What are you working on? • How are you doing; how are your OKRs coming along? • Is there anything impeding your work?

What do you need from me to be (more) successful? • How do you need to grow to achieve your career goals? Now, I’m not proposing that performance reviews and goals

BetterWorks’ experience with hundreds of enterprises, five critical areas have emerged of conversation between manager and contributor: Goal setting and reflection, where the employee’s OKR plan is set for the coming cycle. The discussion focuses on how best to align individual objectives and key results with organizational priorities. Ongoing progress updates, the brief and data-driven check-ins on the employee’s real-time progress, with problem solving as needed.fn3 Two-way coaching, to help contributors reach their potential and managers do a better job. Career growth, to develop skills, identify growth opportunities, and expand employees’ vision of their future at the company. Lightweight performance reviews, a feedback mechanism to gather inputs and summarize what the employee has accomplished since the last meeting, in the context of the organization’s needs. (As noted earlier, this conversation is held apart from an employee’s annual compensation/bonus review.)

Here’s the precious opportunity for people to say to their leaders, What do you need from me to be successful? And now let me tell you what I need from you. Not so many years ago, employees made their voices heard by

‘[H]igh-recognition’ companies have 31 percent lower voluntary turnover than companies with poor recognition cultures.” Here are some ways to implement it:

recipe done in time for everyone who’s waiting on it. My deadline’s built into a key result. I can see the bigger picture more clearly. It’s definitely a team-building

like to start with three questions: What makes you very happy? What saps your energy? How would you

The point is that a strong and positive corporate culture is absolutely essential.” As an engineer, Grove equated culture with efficiency, a manual for quicker, more reliable decisions. When a company is culturally coherent, the way forward is understood: Someone adhering to the values of a corporate culture—an intelligent corporate citizen—will behave in consistent fashion under similar conditions, which means that managers don’t have to suffer the inefficiencies engendered by formal rules, procedures, and regulations …. [M]anagement has to develop and nurture the common set of values, objectives, and methods essential to the existence of trust. How do we do that? One way is by articulation, by spelling [them] out …. The other even more important way is by example. As an executive, Grove role-modeled

values:

In Project Aristotle, an internal Google study of 180 teams, standout performance correlated to affirmative responses to these five questions: 1. Structure and clarity: Are goals, roles, and execution plans on our team clear? 2. Psychological safety: Can we take risks on this team without feeling insecure or embarrassed? 3. Meaning of work: Are we working on something that is personally important for each of us? 4. Dependability: Can we count on each other to do high-quality work on time? 5. Impact of work: Do we fundamentally believe that the work we’re doing matters?