Measure What Matters

My employer, the OEF, went through making our OKRs for this quarter a few weeks ago. It was the second time I’d been at an organization that uses OKRs; we also had them at Wikimedia. The process was also very different, and it made me realize that I only knew OKRs through folk wisdom and cult practices. So, I decided to start reading up on OKRs, and the main book for doing that is Measure What Matters.

MWM was written by John Doerr, legendary venture capital investor from Kleiner Perkins. Doerr learned the OKR system while working at Intel for Andy Grove in the 1960s to 1980s. When he began investing, he took the OKR system to the tech startups he was working with, helping them focus on what was important to their success.

If you’re unfamiliar with the term, here’s a crash course: OKR stands for “Objectives and Key Results”. They’re a framework for defining business goals (objectives) and ways to measure progress towards those goals (key results). So, if the Objective is to move a mobile app company into the Canadian market, some related key results would be 100,000 downloads in the Canadian app store, 100 reviews in the Canadian app store, and 3 stories in Canadian national press.

Usually, companies set OKRs for a quarter (3 months), with a goal of about 2-3 company-wide objectives and about 3 key results per objective. Teams set their own OKRs, and individuals too. Typically the objectives for teams are related to the objectives or KRs of the company. So the product team might have an objective to localize the mobile app for the Canadian market, which “cascades” from the company-wide objective.

I like the framework. It’s got a real feeling of being the simplest thing that could possibly work, in the spirit of Ward Cunningham’s quote for the design principles for wikis. I’m also reminded of Open Space technology for conferences, which focuses on a few key process points and then lets people structure their own methods around it. But, like most

Doerr’s book is half a memoir, half a guidebook. He writes a lot about his own personal life as he moved from Harvard to Silicon Valley, and his work at Intel. Most of the case studies are for companies Doerr invested in, Google being the most famous. There’s an entire chapter dedicated to his friend and mentor Bill Campbell, which only tangentially mentions OKRs at all. But he does get into the theory and practice of OKRs for companies, which makes it worthwhile.

One thing that struck me was how many of the objectives Doerr talked about in the book were about obtaining near-monopolies in the technology sector. A big case study is about YouTube getting to 1 billion hours of video watch time per day, which of course requires being the dominant video platform on the Internet. Andy Grove at Intel led Operation Crush, an effort to push the superior Motorola 68xxx processor out of the market by getting companies to commit to designs that used Intel’s chips instead.

I actually “read” the audiobook, which Doerr reads, accompanied by the actual entrepreneurs from the case studies, who read their own segments. It works pretty well, especially when it comes to Bono’s turn to talk about using OKRs at 1. This was probably the most inspirational and relevant section for my work, since it was about delivering value for people who need it via a non-profit, and including them in the process. Bono probably makes too many references to how “rock and roll” OKRs are, which, like, OK, but otherwise it’s a pretty good segment.

There were a few main things that I learned about OKRs from the book.

  • KRs are measured by degree of belief, that is, how sure you are that they’ll be achieved, on a range from 0 to 10. (Basically, it’s a percentage, multiplied by 10 and rounded off). So, if you get to 25,000 downloads in Canada halfway through the quarter, the KR isn’t at 25% complete, but might be at 8 or 9 confidence level — because you’re almost certain to get to 100K by the end of quarter. Once you get to the end of quarter, you’ll know better!
  • People colour-code confidence levels; 0-3 is red, 4 to 6 is yellow, 7 and above is red.
  • OKRs don’t just cascade down from the top; they are also negotiated bottom up (individual -> team -> company) as well as horizontally (say, product department and tech department). This was a big problem when I was at WMF; our OKRs came lumbering out of the executive team around the beginning of the quarter, and everyone waited as smaller and smaller sub-teams decided on their OKRs and then passed them down.
  • OKRs belong to teams and individuals. This was another thing that was weird at WMF; the team lead’s individual OKRs were identical to the team’s OKRs. It felt weirdly feudal.
  • The big difference between the O and the KR are qualitative versus quantitative. O is what you want to do, and KR is how you know you did it. If there are numbers in an objective, it’s actually a KR, and you should figure out what the objective really is!
  • There’s a difference between committed objectives (things that you’re supposed to do) versus aspirational objectives (things that would change your business). So, committed objectives might be around existing products and markets, or around your organizational structure and processes. Aspirational OKRs are about big change. Doerr says that you should usually get to green on your committed KRs or know why, but aspirational KRs are more likely to be red or yellow.
  • The biggest way people fail at OKRs is by setting goals at the beginning of the quarter, forgetting them entirely, and then maybe coming back to the at the end of the quarter. OKRs should be a part of status meetings, update emails, and any other review and reporting process.

Overall, the book does the trick; it’s a great introduction to OKRs with details on how to use them effectively. It’s got some archaic stories from Silicon Valley of yore, too, which can be entertaining or mortifying, depending on your outlook. If you’re using OKRs, I recommend it.

One thought on “Measure What Matters

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s