Several years down the line, it’s still an issue: people expect a Key Result to become an Objective when it is assigned to someone else. This approach is promoted in John Doerr’s book Measure what matters and in Rick Klau’s famous OKR video. It’s also how many OKR tools work.

But it doesn’t take a scientist to understand that this approach is flawed. It simply does not work. 

I’ll show you why and will try to tackle this issue once and for all.

First, the purpose of Key Results

Key Results kill 2 birds with 1 stone.

On the one hand, they help to specify what the Objective means. Objectives can be a bit ambiguous, so you need Key Results to quantify the Objective and take away that ambiguity. Imagine your Objective is to become one of the fastest-growing companies in your industry. You could measure this in terms of revenue, number of customers, number of employees, or multiple indicators at once. Whatever you mean by “fastest-growing” the Key Results will make it clear.

On the other hand, Key Results help you measure progress toward the Objective. Key Results are, as Andrew Grove, the father of OKR, says: “meant to pace a person—to put a stopwatch in his own hand so he can gauge his own performance.” In other words: the Key Results define whether the Objective has been achieved.

In short, an Objective needs Key Results to quantify it and measure progress toward it.

Key Results must contain a metric

In order to quantify the Objective and unlock the ability to measure progress, a Key Result must contain a metric (see also The anatomy of a Key Result).

The word measure is encapsulated in the term metric. The word metric is derived from the Greek word métron (μέτρον), which means “measure” or “something used to measure”. In other words: the metric is what makes the Key Result “measurable.”

Virtually everyone, including John Doerr, seems to agree that a Key Result must always contain a metric. Phew!

The metric defines a Key Result’s progress

When a Key Result contains a metric, you don’t get to define progress for the Key Result yourself anymore.

Let’s say you have an Objective to Improve customer happiness. A Key Result could then be to Increase NPS from 30 to 50. In this case, NPS is your metric, and you’d have to look at your current NPS score to see what the correct value is for your Key Result.

There are lots of things that you can do to increase NPS. You could squash bugs in your product, hire extra support agents to bring down your response time, build highly requested features, and so on. But doing these things won’t automatically mean NPS will go up. NPS will only go up if more of your users give you a high NPS score and become promoters.

In other words, only the (current value for your) metric can and will define progress for your Key Result.

What happens if a Key Result becomes an Objective

Everyone agrees that Key Results drive progress for an Objective.

If I assign my Key Result to Increase NPS from 30 to 50 to you and I want it to become an Objective for you, your Objective will be: Increase NPS from 30 to 50

Your Objective now contains a metric (NPS). If you decide to create Key Results for this Objective, these Key Results cannot drive progress for the Objective. When an Objective or Key Result (or any other goal for that matter) contains a metric, it’s the metric that will define its progress.

So we’re stuck: you have an Objective for which you cannot create Key Results anymore. What’s been the point of the Key Result becoming an Objective for you?

A better approach

A much cleaner and simpler approach would have been to assign the Key Result Increase NPS from 30 to 50 to you without it becoming an Objective for you.

By assigning the Key Result to you, I make you responsible for achieving it. This isn’t easy. There are a number of things you could do. The project and tasks that you’ll initiate to achieve this Key Result are what we call Initiatives at Perdoo. Initiatives don’t impact an Objective’s progress.

As you complete your Initiatives, you’ll now be able to see if they move the needle for your Key Result. If they do, awesome — you’re on the right track. If not, you may need to try other Initiatives.

By differentiating between Key Results and Initiatives, and by tracking them in the same place, you’re all of a sudden able to see how the things you’re doing are helping you achieve the (key) results you desire!

Aligned Objectives

Sometimes it’s not just you working on Initiatives for an Objective and one or more of its Key Results. Sometimes there are others working on an OKR that is aligned with the Objective that you’re working on.

Let’s go back to our example. The Objective is to Improve customer happiness, and one of its Key Results it to Increase NPS from 30 to 50. The engineers were aware of the Objective to improve customer happiness and decided to create an Objective for themselves to reduce the number of bugs. They assume that by reducing both the number of open bugs and the number of new incoming bug reports, customer happiness will improve. So they’ve aligned their Objective to Reduce the number of bugs with the Objective to Improve customer happiness.

While the engineers are probably right that reducing the number of bugs will improve customer happiness, it is an assumption. Successfully achieving the Objective to reduce the number of bugs does not guarantee that customer happiness will improve. Only by looking at the Key Results for the Objective Improve customer happiness will we know for sure that it has improved.

In other words, child-Objectives are always assumed to help achieve parent-Objectives. But only by looking at the parent-Objective’s Key Results will you know if your assumption was correct.

Goals That Matter Metrics