You can't write good goals in a company that hasn't decided where it is going

For the first few years we did exactly what we were asked to do. We went into a company, sat down with the teams and set objectives for the quarter. The objectives came out correct. They had measurable key results, they had owners, they had deadlines. And then they lost. Not to the competition, but inside the company: the fight for people's time, for money, for the board's attention. There was always something more important.

It took us a while to understand why. An objective is a choice, and a choice can only be justified when you know where the company is heading. If you don't know that, every objective is equally good and equally easy to put off. Lack of clarity at the top doesn't stay at the top. It multiplies on the way down, and at the level of execution it turns into objectives nobody believes in.

For us, by the way, this was a path upwards, not downwards. We started with OKR training, then moved on to implementations, and every new attempt showed that the problem lay one level above the one we happened to be working on. At some point we stopped taking on assignments framed as help with writing objectives, because we knew they would end in disappointment on both sides.

That is why today, before anyone writes their first quarterly objective with us, we take two steps back. And that is the whole content of this text: what the path looks like from an empty room to a company that looks at its numbers every week and knows what to do with them.

The first step isn't strategy. It's agreeing on where we are

Strategic analysis has a bad reputation, partly deserved, because too many companies stop their work there. SWOT, PESTEL, competitor analysis: none of this is a strategy. It is a tool for one specific thing, namely making sure everyone in the room sees the same picture of the current situation. Without that you can't make a decision, because everyone is deciding from a different map.

On top of that come the mission, vision and values, the part boards are most eager to skip as ceremony. Yet this is the first filter. A company that knows why it exists and what it won't do even for good money has already ruled out half of the ideas before anyone starts doing the sums.

Only then come the decisions. The main challenges, the policies, the guiding principles that address them, and measurable strategic goals with a measurement formula and a person responsible. All of it fits on one page, and that isn't coquetry. A strategy that doesn't fit on one page is usually not a strategy but a collection of analyses nobody will go back to.

This document deliberately contains no OKRs yet. We draw a clear line there. We write quarterly objectives only once the decisions and policies are agreed, because otherwise the objectives become the first opportunity to smuggle in whatever couldn't be agreed earlier.

It is also worth settling one question of language straight away, because later it saves weeks of misunderstanding: what this particular company means by an objective and what it means by a key result. We see large organisations in which the objective is an unmeasurable signpost and all the work happens under the key results. That can be run, but you need to know it is already a different system from the one we are talking about, and that a moment comes when such a split starts to get in the way.

What to do about it this quarter

If you have next year's planning ahead of you, start with a retrospective of the year that is ending. Not with ideas for the future, but with the conclusions of all the working groups, the current status of the goals, and a separate retrospective at board level. This is the only moment when the company has complete information about itself, and it would be a shame to waste it on a brainstorm.

Then count your resources before you approve the objectives. Skills, technologies, capital, access to partners. And separately count how much capacity the day-to-day business takes, because that is what eats the quarter. In practice the deciding factors are thoroughly unromantic: recruiting a specialist takes three months and waiting for funding another three, so an ambitious objective simply moves further out, and it is better to know that in January than in March.

Don't run two separate worlds either. Quarterly objectives and ongoing projects are ultimately one list of priorities, not two. A company that has OKRs alongside its normal work will always choose the normal work, because it is urgent, while the objectives are important. Urgent beats important every day of the week.

Finally, set the rhythm. We look at quarterly objectives every week and annual ones every month, and the meeting starts with the values of the key results, not with a task list. It looks like a detail, but it is the difference between a team that asks itself whether it is heading in the right direction and a team that ticks off projects and finds out in the last week that it wasn't.

For every key result we track four values: the baseline, the target, the current value and the confidence of delivery. The fourth is most often filled in as a formality, and then it is useless. Its point is that someone enters low confidence and the meeting starts with the question of what, in that case, we change in the way we work. That is the whole difference between a spreadsheet and a conversation.

What you can't see from the inside

There is one more thing we talk about in this episode, which isn't a method and can't be written into the OnePager. The greatest value someone from outside brings is not knowing OKRs. It is being present at team meetings, week after week. Nobody inside the organisation sees this, because when the boss is sitting in the meeting, the meeting looks different.

I write about this with some embarrassment, because for years we sold the method while the real value lay elsewhere. Today we tell clients plainly at the start of the work: we will come, sit with your teams and tell you things you won't hear from your own people. That can be uncomfortable, and it is exactly the part worth paying for.

Sometimes the role stays for longer, because there is nobody in the company to fill it. In organisations without a chief operating officer we step into that gap and either fill it or help the company develop such a role of its own. The COO role is also the least sharply defined of all the roles on a board, because it usually came about to make up for the particular weaknesses of a particular CEO, not from a textbook.

We also monitor at several levels at once, and the last of them is the least obvious. We look at quarterly objectives every week and annual ones every month, after each quarter we run retrospectives, and separately we ask whether the method itself is helping this company deliver its strategy at all. That question sounds awkward coming from someone who makes a living from implementing OKRs, and that is why it has to be asked out loud. If the company already has another way of working with goals and it works well, implementing the method for the method's sake is a waste of the client's money.

And one last remark, so that this doesn't sound like a promise of a miracle. The numbers in the key results only really start to move after about two months. The first quarter usually doesn't look spectacular. It looks like a company that, for the first time in a long while, knows what it isn't doing.