The first step of an implementation isn't writing an objective. It's writing down why we're doing it

The goal of an implementation is a written answer to two questions: what result we want to achieve thanks to this change over a one-year horizon, and how we will know the change actually helped. Writing it down is not a formality. An answer kept in your head can't be agreed in a group or revisited six months later. Without it OKRs become an end in themselves, and all the later decisions, the structure of the objectives, the length of the cycle and the make-up of the teams, have nothing to refer back to. While you are at it, ask the opposite question: what would mean, at the end of the year, that this implementation had failed completely. Thinking from that side brings out quite different things than the question about success.

Each such goal gets a metric, and this is where the most interesting part of the episode begins. In a large financial company they were trying to get away from a model in which priorities are decided by the manager who shouts loudest, so they measured the ratio of initiatives planned to initiatives delivered in a quarter and the shortening of the time from picking up a task to releasing it to production. In a product company of around a hundred people, staff didn't know what the strategy was, even though it was well written and sat on the company intranet; that was measured with a survey, and understanding of the goals additionally by a drop in the number of requests for decisions sent to the product owner. The logic is simple: whoever understands what the result should be makes decisions on their own. Some goals are simply qualitative, and then a survey on a scale of one to five, at the start and at the end of the quarter, subjective as it is, is better than no measurement.

Across industries the same four needs keep coming back: focus, agreement on direction and breaking down silos, a sense that we are executing the strategy, and the shift from thinking in initiatives to thinking in results. The goal of the implementation is set by the group that genuinely wants and is able to carry the change through, the board, sometimes widened to include directors and leaders. Someone from outside is useful here for one reason, and it isn't knowledge of the method: it is easier to answer an uncomfortable question for someone who isn't inside the company's context and has nothing to lose in it.

It is a signpost, not a declaration for the start

An implementation goal written down and put away in a drawer is useless. It comes back at two specific moments. The first is deciding how many objectives to set for the quarter: if you declared focus and then set yourselves twenty objectives, it simply isn't consistent, and the implementation goal is the only document that will show it. The second is the retrospective after a quarter in which the objectives weren't delivered, because then it suggests what to look for as the cause. It is also allowed to change, and that is normal, not a mistake in the original decision. After one or two quarters the original problem may be solved and the organisation needs to focus on something else. In one implementation that lasted almost seven years, the implementation goals changed many times along with the company, the market and the strategy.

At the workshop this episode comes from, one of the participants wrote down their implementation goal as achieving the objectives, not just writing them down. That sentence says more about our industry than we would like, and it captures well where these implementations most often stall. The metrics proposed then were boring and that is exactly why they were good: how many conclusions we draw at the end of the quarter, how often we look at the objective at all, what percentage of it has been achieved. The problem wasn't the writing but the lack of a habit of working with what had already been written. The same goes for us: OKRs are one of many management-by-objectives tools, and you choose them to fit the problem, not the other way round.