November 24, 2020

OKRs: The Good, The Bad, and the Ugly

Kurt Bittner

View profile

OKRs are a powerful tool for setting and measuring goals, but like all power tools, they can become dangerous in unskilled hands. When used in informed ways, they can improve focus and lift outcomes, but when mishandled, they can increase waste and inflict real damage upon the organizations that adopt them.

The basic idea behind  OKRs is simple: it’s a good idea to identify goals that you want to achieve and the measures that will indicate whether you’ve achieved those goals. No one would argue that this is a bad thing, at least if the goals are focused on conditions that the organization would like to achieve. Where OKRs can go “bad” is when they focus on merely producing things, and things can get really “ugly” when they focus on merely doing things.

Expressing Objectives in terms of Outcomes produces better results

Scrum.org has developed an improvement framework, Evidence-Based Management, that nicely complements OKRs. In it, we identify three types of measures:

  • Outcomes: changes in the satisfaction or experience of some stakeholder, such as customers or users of a product, but also employees and investors.
  • Outputs: things that the organization produces, such as products or services, but also things like reports and presentations.
  • Activities: things that people in the organization do, such as developing products or services, but also things like attending meetings and developing plans.

Achieving outcomes is always the real goal of producing outputs or performing activities; the outputs and activities may be a means to an end, but they are never the end themselves.  Expressing objectives in terms of outputs is bad because it can lead to micromanaging people to produce something that doesn’t help the organization move closer to the objective. Expressing objectives in terms of activities is even worse: it literally micromanages people by telling them what to do.

Bottom-up intelligence leads to better outcomes

Micromanaging people is counterproductive and demoralizing. It presumes that people lack the competence to make decisions about the best way to achieve a goal. In nearly all cases, the people closest to the work are better able to make decisions about how to achieve a goal because they have the most information about available alternatives; this is known as bottom-up intelligence.  When organizations free their people to contribute using all of their knowledge, experience, and passion, they achieve better results.

Ask “Why?” to uncover the real objective

Objectives expressed in terms of outputs or activities can be improved by using a technique similar to the 5 Whys approach: When presented with an objective, ask “Why is that important?” If the answer is “because it will lead to something else”, or some variant thereof, then the thing it will lead to is the real objective. You may have to apply this more than once to get to the real goal.

Consider the following example:

The weak objective dictates an activity, creating and making available virtual training. The real objective is on the right: students don’t really care whether they obtain training virtually or in person, but they would like to be able to learn whenever they need to, regardless of where they are located or the course schedule. 

Another example we often encounter is from organizations who, when asked why they want to be more agile, say that they want to “go faster”, meaning that they want to release faster, or get products to market faster.  This objective fails the “why?” test; it is merely the means to some other goal, such as to be more responsive to customers, or to competitor actions.

In fact, merely delivering faster is not enough to satisfy the goal of being more responsive; in addition to delivering faster, an organization must also “listen better”, to better understand customer needs and to measure the change in customer outcomes resulting from delivering faster. Without this, organizations who deliver faster may be simply creating waste at a faster rate.

Use Key Results as indicators of progress, not success criteria

Key Results are indicators that tell people whether they are making progress toward a goal. Or at least that’s the theory.  The reality is that they are assumptions based on a presumed causal relationship between the Key Result and the Objective, that if the organization achieves the Key Result, it will be making progress toward the Objective.

Achieving Key Result is generally regarded as positive and failing to achieve the result is generally regarded as negative. This is flawed; achieving the Key Result might have no impact on achieving the Objective. Failing to achieve a Key Result might not be a bad thing if achieving it is unrelated to achieving the Objective; in fact, it might be a good thing if the organization learns something about the things that will help it to achieve the Objective.

It’s better to regard Key Results as hypotheses about things that the organization can do to move toward the Objective. Framing Key Results in this way makes it easier to discuss what the organization learned from the results, to inspect these results and adapt based on what the organization learns.

Unfortunately, too often, Key Results are used to gauge performance, to reward people when they achieve the Key Result, and to punish them when they don’t. Weaponizing Key Results makes it almost impossible to learn anything, and it produces counterproductive behaviors like gaming measures and shifting blame. Using Key Results for reward and punishment also can be a form of micromanagement that undermines creativity and initiative.

Conclusion

OKRs are a powerful tool for setting and measuring goals. Done correctly they enable better alignment between the work people do, the things they produce, and the results they achieve. Freed from micromanagement, people are able to fully engage their creativity in solving complicated problems, but only when their goals are expressed in terms of desired outcomes, not activities or outputs. 

Making the real goals transparent by expressing them in terms of outcomes helps organizations to better inspect and adapt those goals, in addition to the key results that the organization achieves. Doing so helps the organization improve not only the means by which it pursues goals, but also the goals themselves.