Coffee with OKRs · 4 min read
Goal completion rates say nothing about an implementation. Measure strategy delivery and whether people see value in it
The first implementation metric mentioned in this episode was: twelve meetings in a quarter. It said nothing about value. What it did build was the habit of a weekly hour in which you talk about direction, not the day-to-day, and everything that came later grew out of that habit.
From the episode · in Polish
Four people, four scales of organisation, one question
The conversation has four participants, and that is its greatest strength: the same questions get answers from organisations of completely different scale, from a startup in a phase of exhausting growth to a company with fifty-four teams. Specific measurement tools keep coming back: a survey run separately for the board, management and the rest of the organisation, questions about the quality of synchronisation between teams, and a weekly engagement survey drawing questions at random from a pool of around forty. The part with the admission that at the start we were measuring completely the wrong things is roughly halfway through. from 00:27:20
The recording is being edited. Leave your address and you'll get the link when it's ready.The question of how to measure an implementation sounds like an auditor's question, but it is a question about money. An implementation costs training, people's time and a change in the way of managing, so before a pilot is extended to further teams, someone has to be able to say whether it pays off. You begin with the goals of the implementation itself, because they determine what is to be measured. If we turned to goals because several teams were doing the same thing in parallel, then synchronisation is what needs checking, not satisfaction with the tool.
At the start, though, people measure what is easy. How many teams have created objectives at all, whether they are updated regularly, whether any conversation takes place around them. In this episode we say plainly that we ourselves started with things like that, very process-oriented, and that they weren't the right things: we were measuring how far the tool had been rolled out, not the effect we had introduced it for. That hygiene isn't worthless, it just serves a different function from assessment. It is a warning signal, and it gives most in comparison: set the teams where the process is alive against those where it has died out, and look for the difference, instead of averaging the result across the whole company.
Instead, you measure two things. The first is delivery of the strategy: if the strategy is moving forward, the management method is probably helping, and that is the ultimate metric, not the state of the tool. The second is the value people feel, and it is worth asking the board, management and the rest of the organisation separately, because the answers differ, and that difference is part of the information. Feelings alone, without hard data, can be misleading. An answer that it doesn't help usually means the method has been misunderstood and is being misused, not that it is bad. That is why you look at the trend over successive quarters, not at the level in a single measurement. The first results won't be good anyway.
Takeaways
Four things that measuring an implementation is really about
- 01
Start with the goals of the implementation itself
Without them, measurement is pieced together from whatever happens to be countable, meaning the state of the tool. Weave the question about the method into the quarterly summary instead of building a separate oversight process; monitoring the implementation shouldn't grow into a goal that itself has to be delivered.
- 02
Mechanics are a signal, not an assessment
The number of teams with filled-in objectives and the regularity of updates tell you where the process has died out, and that is all. Value appears only when you compare live teams with dead ones.
- 03
Progress without context carries no information
The textbook 70 per cent completion, which we started out using to assess ambitious objectives, over time stopped meaning anything. A team showing 300 per cent progress and a team lowering its ambitions to come out at a hundred are telling you about the culture of accountability, not about the method.
- 04
This is continuous measurement, not an audit after a year
The first signals show after a few weeks, you can talk about success when a large part of the organisation works with the method, and the real proof is the second year, in which the organisation still wants to use it and does so better.
The whole topic in one place
Why don't OKRs work in my company? An implementation audit
Here we write about how to tell that an implementation is heading in the right direction. The opposite question, where to look for the cause when measurement shows it has stalled, we have taken apart in the text on auditing an implementation. We have also gathered there all the episodes in which this topic came back.
Author
Lena Zhukova
founder; strategy workshops, quarterly sessions
Founder of OKR Poland. For twenty years she has worked with boards of companies from 30 to a few thousand people; she wrote dozens of strategies with them and helped implement them in more than 40 organisations.
Next texts by e-mail
The OKR Poland newsletter
We write when we have something to say. To start: eight hours of recording on preparing a company for strategy workshops, plus the OnePager.