You don't implement a tool, you build a habit, so you start with one small group

Implementing OKRs means changing a set of the organisation's habits, not installing a tool, so it follows the same rules as changing personal habits: a little at a time and regularly. Nobody rearranges their whole life on a single Monday, yet companies try exactly that, opening the method up to every team from day one. Hence the recommendation from this episode: one small pilot team that works through a full quarter with it. The reason can be counted. The more teams work in this rhythm, the more meetings there are, the more time goes into keeping an eye on what happens in them, and the more people have to be able to run them. The cost rises with every additional team, and a habit can't be bought wholesale anyway.

The most important item on the list of first steps is not choosing a tool or a meeting format, but choosing the person who will lead it. That person needs two things: real power to get things done in the organisation and responsibility for a piece of the strategy. Willingness and a free hand are not enough on their own, and the most common counter-example is someone appointed because they happened to be available, who doesn't feel the subject and lacks clout. In that set-up the implementation is unlikely to work. Behind this stands a second, harder point: if the board has no time for strategic matters, it means it is so busy with the day-to-day that OKRs are not its biggest problem right now. In one of the set-ups described, the board divided this explicitly: one person is responsible for the current year, another for the horizon beyond a year, and the quarterly objectives plug into the first role.

The biggest trap of first objectives is describing activity instead of results. At the end of the quarter you are then left with a sentence about how hard we tried, not a fact. That is why first objectives should be as simple as possible and decidable by data, 1 or 2 per quarter, no more. Sometimes such a statement looks trivial: in a company so absorbed by the day-to-day that there was no room left for strategic thinking, the key result for the quarter became 12 strategic meetings, one a week. It worked precisely because it could be settled by data rather than by discussion. And one last thing we say plainly in this episode: during the first quarter you will be very tempted to rewrite the objectives or change them completely. Don't, because the consequences of a badly set objective are learning material here. After that the process improves by one small thing per quarter, and that is exactly how we improve it, instead of staging a revolution after every cycle. After a few quarters, what comes out of it works better.