Measure What Matters: how to write OKRs that work

Who should use it, and when not to

Use OKRs when a team works hard but pulls in different directions, when priorities change every week, or when nobody can say what success looks like this quarter. They help most in teams of five or more, and in companies where several teams must line up.

They help less in a team of two that already talks every day, or for routine work with fixed targets, such as a monthly close. They don't replace a strategy. If you don't yet know what to focus on, decide that first, for example with the strategy kernel.

The core ideas

Objectives and key results. The objective is the direction ("Make onboarding painless for new clinics"). Key results are outcomes, not tasks ("Median setup time from 9 days to 3"). A key result you can tick off by doing work, like "launch the new wizard", is a task in disguise.

The superpowers. The book frames the benefits of OKRs as four superpowers: focusing and committing to priorities, alignment across teams, tracking progress, and stretching for more than seems possible. The What Matters site splits commitment out as its own letter and summarizes them as F.A.C.T.S.

Committed and aspirational. Committed OKRs must be met in full. Aspirational ones aim high, and reaching most of the way counts as a good result. Say which is which when you write them.

Conversations, feedback and recognition. The book pairs OKRs with frequent check-ins instead of once-a-year reviews. What Matters calls this CFRs: conversations, feedback and recognition.

A worked mini example

This is an illustration. A 12-person team sells booking software to physio clinics. Last quarter, new clinics took a median of 9 days to set up, and 1 in 5 cancelled in the first month.

The wizard, the articles and the calls are still on the plan. They are now the work the team bets on to move the key results, not the goals themselves.

How to write OKRs

1. Start from the strategy. Write the one or two things that matter most this quarter, and why.
2. Write 1 to 3 objectives. Plain, memorable, and ambitious. No numbers in the objective.
3. Add 3 to 5 key results to each. Each is an outcome with a baseline, a target and an owner.
4. Test every key result. If you could finish it by doing a task, rewrite it as the result that task should cause.
5. Mark each OKR committed or aspirational. Agree what score counts as success for each kind.
6. Share and align. Publish the OKRs, and check that team OKRs support the company's, without copying them down line by line.
7. Check in weekly, grade at the end. Track progress, talk about blockers, and grade honestly. Keep the grade out of pay decisions.

The OKR draft skill runs these steps with your goals and numbers, and flags key results that are really tasks.

Common mistakes

• Too many OKRs. Ten objectives means no focus.
• Tasks as key results. Shipping a feature is not the same as solving the problem.
• No baseline. "Increase retention" means nothing without a starting number.
• Tying OKRs to bonuses. People then set targets they know they'll hit.
• Setting and forgetting. OKRs read only at the start and end of a quarter change nothing in between.

For how Andy Grove ran the method at Intel, see Andy Grove's OKRs and managerial leverage and High Output Management.

Where it comes from

John Doerr, a longtime partner at Kleiner Perkins and an early investor in Google, published Measure What Matters in 2018. He learned the method while working at Intel under Andy Grove, who built it on Peter Drucker's management by objectives and described it in High Output Management (1983). Doerr introduced OKRs to Google's founders in 1999. The book combines his account with stories from companies and organizations that used OKRs. Its companion site, What Matters, publishes articles and examples.

This page is an independent summary in our own words. It is not affiliated with or endorsed by John Doerr or his publishers. The book's title is used only to name it. More explainers are in the business books index.

Questions

What is an OKR?

An objective is what you want to achieve, in words people remember. Key results are three to five measurable outcomes that show whether you got there. Together they set a team's priorities for a period, usually a quarter.

What is the difference between committed and aspirational OKRs?

John Doerr's What Matters site describes committed OKRs as ones the team is expected to fully meet, and aspirational OKRs as stretch goals where falling short is normal. Mixing them up leads to either timid goals or missed promises.

Should OKRs be tied to bonuses?

The What Matters site describes OKRs as separate from compensation, a practice it traces back to Andy Grove at Intel. Tying them to pay pushes people to set targets they know they can hit.