Introduction to OKRs

Chapter 4. Improve Weekly Status Emails with OKRs

I remember the first time I had to write a status email. I had just been promoted to manager at Yahoo! back in 2000 and was running a small team. I was told to “write a status email covering what your team has done that week, due Friday.” Well, you can easily imagine how I felt. I had to prove my team was getting things done! Not only to justify our existence, but to prove we needed more people. Because, you know, more people, amiright?

So I did what everyone does: I listed every single thing my reports did, and made a truly unreadable report. Then, I began managing managers, and had them send me the same, which I collated into an even longer more horrible report. This I sent to my design manager, Irene Au, and my general manager, Jeff Weiner (who sensibly requested that I put a summary at the beginning).

And so it went, as I moved from job to job, writing long, tedious reports that, at best, were skimmed. At one job, I stopped authoring them. I had my managers send them to my project manager, who collated them, and then sent it to me for review. After checking for anything embarrassing, I forwarded it on to my boss. One week I forgot to read it, and didn’t hear anything about it. It was a waste of everybody’s time.

Then, in 2010, I got to Zynga. Now, say what you want about Zynga, but it was really good at some critical things that make an organization run well. One was the status report. All reports were sent to the entire management team, and I enjoyed reading them. Yes, you heard me right: I enjoyed reading them, even when were 20 of them. Why? Because they had important information laid out in a digestible format. I used them to understand what I needed to do and learn from what was going right. Please note that Zynga, in the early days, grew faster than any company I’ve seen. I suspect the efficiency of communication was a big part of that.

When I left Zynga, I started to consult. I adapted the status mail to suit the various companies I worked with, throwing in some tricks from Agile. Now I have a simple, solid format that works across any organization, big or small.

1. Lead with your team’s OKRs, and how much confidence you have that you are going to hit them this quarter.

   You list OKRs to remind everyone (and sometimes yourself) why you are doing the things you do.

   Your confidence is your guess of how likely you feel you will meet your Key Results, on a scale from 1 to 10. Mark your confidence red when it falls below 3, green as it passes 7. Color makes it scannable, making your boss and teammates happy. Listing confidence helps you and your teammates track progress and make corrections early if needed.

2. List last week’s prioritized tasks and whether they were achieved. If they were not, include a few words to explain why. The goal here is to learn what keeps the organization from accomplishing what it needs to accomplish. See Figure 4-1 for format.

3. Next, list next week’s priorities. List only three P1s (priority #1), and make them meaty accomplishments that encompass multiple steps. For example, “Finalize spec for project xeno” is a good P1. It probably encompasses writing, reviews with multiple groups and sign off. It also gives a heads up to other teams and your boss that you’ll be coming by.

   “Talk to legal” is a bad P1. This priority takes about half hour, has no clear outcome, feels like a subtask and, not only that, you didn’t even tell us what you were talking about!

   You can add a couple of P2s, but they should also be meaty, worthy of being next week’s P2s. You want fewer, bigger items.

4. List any risks or blockers. Just as in an Agile standup, note anything you could use help on that you can’t solve yourself. Do not play the blame game. Your manager does not want to play mom, listening to you and a fellow executive say “it’s his fault.”

   As well, list anything you know of that could keep you from accomplishing what you set out to do—a business partner playing hard-to-schedule or a tricky bit of technology that might take longer than planned to sort out. Bosses do not like to be surprised. Don’t surprise them.

5. Notes. Finally, if you have anything that doesn’t fit in these categories, but that you absolutely want to include, add a note. “Hired that fantastic guy from Amazon that Jim sent over. Thanks, Jim!” is a decent note, as is, “Reminder: team out Friday for offsite to Giants game.” Make them short, timely and useful. Do not use notes for excuses, therapy or novel writing practice (see Figure 4-1).

Figure 4-1. Listing of priorities

This format also fixes another key challenge large organizations face: coordination. To write a status report the old way, I had to have team status in by Thursday night in order to collate, fact-check, and edit everything. But with this system, I know what my priorities are, and I use my reports’ statuses only as a way to ensure that their priorities are mine. I send out my report Friday, as I receive my reports. We stay committed to one another, honest and focused.

Work should not be a chore list, but a collective push forward toward shared goals. The status email reminds everyone of this fact and helps us avoid slipping into checkbox thinking.

Coordinating organizational efforts is critical to a company’s ability to compete and innovate. Giving up on the status email is a strategic error. It can be a task that wastes key resources, or it can be a way that teams connect and support one another.

Tracking and Evaluating OKRs

Two weeks before the end of each quarter, it’s time to grade your OKRs and plan for the next cycle. After all, you want to hit the ground running on day one of Q2, right?

There are two common systems for managing OKRs: confidence ratings and grading. Each has its benefits and downsides. We’ll begin with confidence ratings, the system I’ve outlined earlier. Confidence ratings are a simple system best used by startups and smaller teams, or teams at the beginning of OKR adoption. When you decide on your objective and three KRs, you set a difficult number you have a 50 percent confidence in achieving. This is typically noted by a 5/10 rating on the status-four square.

In your Monday commitment meeting, everyone reports on whether and how their confidence levels have changed. This is not a science, it’s an art. You do not want your folks wasting time trying to track down every bit of data to give a perfect answer; you just want to make sure efforts are directionally correct. During the first couple of weeks of OKRs, it’s pretty difficult to know whether you are making progress on achieving your Key results. But somewhere in week three or four, it becomes very clear whether you are getting closer or slipping. All of the team leaders (or team members, if a very small company) will begin to adjust the confidence rating as they begin to feel confident.

Then, the confidence rating will start to swing wildly up and down as progress or setbacks show up. Eventually, around two months, the confidence levels settle into the likely outcome. By two weeks from the end of the quarter you can usually call the OKRs. If they were truly tough goals—the kind you only have a 50/50 chance of making—there is no miracle that can occur in the last two weeks to change the results. The sooner you can call the results, the sooner you can make plans for the next quarter and begin your next cycle.

The advantage of this approach is two-fold. First, the team doesn’t forget about OKRs because they need to constantly update the confidence level. Because the confidence level is a gut check, it’s quick and painless, which is key for getting a young company in the habit of tracking success. The second advantage is that this approach prompts key conversations. If confidence drops, other leads can question why it is happening and brainstorm ways to correct the drop. OKRs are set and shared by the team; any team member’s struggle is a danger to the entire company. A leader should feel comfortable bringing a loss of confidence to the leadership team and know that he’ll have help.

At two weeks before the end of the quarter you mark your confidence as 10 or 0. Success is making two of the three key results. This style of grading leads to doubling down on the possible goals and abandoning efforts on goals that are clearly out of range. The benefit of this is to stop people spinning their wheels on the impossible and focus on what can be done. However, the downside is some people will sandbag by setting one impossible goal, one difficult goal, and one easy goal. It’s the job of the manager to keep an eye out for this.

The second approach to OKRs ratings is the grading approach. Google is the most famous for using the grading approach. At the end of the quarter, the team and individual grade their results with data collected. 0.0 means the result was a failure, and 1.0 means the result was a complete success. Most results should land in the 0.6 to 0.7 range. From the Google official site on using OKRs, ReWork:

The sweet spot for OKRs is somewhere in the 60–70 percent range. Scoring lower may mean the organization is not achieving enough of what it could be. Scoring higher may mean the aspirational goals are not being set high enough. With Google’s 0.0–1.0 scale, the expectation is to get an average of 0.6 to 0.7 across all OKRs. For organizations that are new to OKRs, this tolerance for “failure” to hit the uncomfortable goals is itself uncomfortable.

Ben Lamorte is a coach who helps large organizations get started and sustain their OKR projects. He regularly uses a grading approach rather than confidence approach. From his article, “A Brief History of Scoring Objectives and Key Results”:

As an OKRs coach, I find most organizations that implement a scoring system either score the Key Results at the end of the quarter only or at several intervals during the quarter. However, they generally do not define scoring criteria as part of the definition of the Key Result. If you want to use a standardized scoring system, the scoring criteria for each Key Result MUST be defined as part of the creation of the Key Result. In these cases, I would argue that a Key Result is not finalized until the team agrees on the scoring criteria. The conversation about what makes a “.3” or a “.7” is also not very interesting unless we translate the “.3” and the “.7” into English.

I’ve arrived on the following guidelines that my clients are finding very useful (see Figure 4-2).

Figure 4-2. Guidelines for grading OKRs

Here’s an example showing the power of defining scoring criteria upfront for a Key Result.

Key Result: Launch new product ABC with 10 active users by end of Q3.

• Grade 0.3 = Prototype tested by three internal users

• Grade 0.7 = Prototype tested and approved with launch date in Q4

• Grade 1.0 = Product launched with 10 active users

This forces a conversation about what is aspirational versus realistic. The engineering team might come back and say that even the 0.3 score is going to be difficult. Having these conversations before finalizing the Key Result ensures that everyone’s on the same page from the start.

As well as precision, Google sets high value on transparency. All OKRs, individual and team, are posted on the intranet, and team progress is shared throughout. Again, from ReWork:

Publicly grade organizational OKRs. At Google, organizational OKRs are typically shared and graded annually and quarterly. At the start of the year, there is a company-wide meeting where the grades for the prior OKRs are shared and the new OKRs are shared both for the year and for the upcoming quarter. Then the company meets quarterly to review grades and set new OKRs. At these company meetings, the owner for each OKR (usually the leader from the relevant team) explains the grade and any adjustments for the upcoming quarter.

And ReWork: warns against the danger of set and forget:

Check in throughout the quarter. Prior to assigning a final grade, it can be helpful to have a mid-quarter check-in for all levels of OKRs to give both individuals and teams a sense of where they are. An end of quarter check-in can be used to prepare ahead of the final grading.

This is also done differently across teams. Some do a midpoint check, like a midterm grade; others check in monthly. Google has always embraced the approach of “hire smart people, give them a goal, and then leave them alone to accomplish it.” As it’s grown, OKRs are implemented unevenly, but OKRs continue to allow that philosophy to live on.

Ben Lamorte also shares a simple technique to keep OKR progress visible: progress posters. Several of his clients have set up posters in the hallway that are updated regularly with progress. Not only does this make OKRs more transparent and visible across teams, it can be effective for communicating scores on Key Results and really creating more accountability. It just doesn’t look good if your team hasn’t updated any scores when you’re already a month into the quarter. Most of these posters include placeholders to update scores at four to eight planned check-ins throughout the quarter. Certainly OKR posters are not for all organizations, but they can be quite effective in some cases.

No matter whether you use confidence check-ins or formal grading (or a combination of both), there is one last piece of advice from ReWork: that is important to keep in mind:

OKRs are not synonymous with performance evaluation. This means OKRs are not a comprehensive means to evaluate an individual (or an organization). Rather, they can be used as a summary of what an individual has worked on in the last period of time and can show contributions and impact to the larger organizational OKRs.

Use the accomplishments of each person to determine bonuses and raises. If you use the status report system described in this report, it should be easy for individuals to review their work and write up a short summary of their accomplishments. This report can guide your performance review conversations. Some things shouldn’t be automated, and the most important part of being a manager is having real conversations about what an employee has contributed—and what he hasn’t.

If you rely on OKR results to guide your decisions, you will encourage sandbagging and punish your biggest dreamers. Reward what people do, not how good they are at working the system.

A Short Note on OKR Software

When you set a resolution, what is the first thing you do? Want to lose weight? Buy an expensive treadmill. Want to start running? Buy fancy shoes. Plan to diet? Buy the best scale on the market. Or maybe you just buy 15 diet books. Sadly, adopting OKRs is treated the same way. People buy software, and hope it’ll do the hard work of setting and managing your goals.

There are a ton of tools for OKRs out there, and many are quite good. But buying a tool is the last step you want to take, not the first. The right way to adopt OKRs is to adopt them in a lightweight fashion, then experiment with different approaches until you find the system that gets you results.

Begin with these tools first:

• A whiteboard, to write ideas of what your objectives will be

• Post-it notes, to brainstorm good KRs

• PowerPoint/Keynote, to track confidence and efforts against the objectives

• Email, to send out statuses

• Excel, if you decide you want to do formal grading (Google offers a tool for grading on its Rework website

The first quarter that you feel you have truly mastered OKRs, go shopping.