A popular goal-setting framework, Objectives and key results (or OKRs) are an effective method for planning and measuring success on a team level. They fall short, however, when companies attempt to apply them to individual contributors. Setting individual OKRs generally leads to goals that are either not true indications of meaningful progress or that are easily gameable. Instead, individual contributors should be assessed based on the extent to which their work contributes to team goals that add real value to the company and its customers.
Objectives and key results, or OKRs, have become one of the most popular frameworks for teams looking to plan and measure the success of their work. With this system, leaders at each level of the organization start by defining high-level, qualitative, inspirational goals, called “objectives.” They then define who will be the user or the consumer of their team’s work, and determine what behavioral changes they would expect to see in those consumers that could be used to quantify whether the team is achieving their high-level goals. These quantifiable outcomes are called “key results,” and are used to measure how successful teams are with respect to their objectives.
This approach keeps planning and progress-tracking focused on the impact the work is having, rather than micromanaging the specific work that teams are doing on a daily basis. Because of this, it is an effective mechanism for aligning top-down strategy with bottom-up, team-level commitments to intermediate goals in support of that strategy. The strength of OKRs explicitly lies in de-emphasizing specific tasks, and instead emphasizing the value that those tasks deliver.
Where OKRs often fall short, however, is when they’re applied at the individual contributor level. Asking employees to set their own individual objectives and key results generally leads to one of two results:
- They create binary goals that are easy to measure but don’t help determine whether they’ve grown or improved in a meaningful way.
- They choose targets they know they can hit, rather than taking a risk on more ambitious goals.
Let’s start with the first failure mode. Earlier this year, I worked with an online gaming company to help them transition to OKRs. As part of the project, the company had asked everyone — including individual contributors — to develop personal OKRs. Here’s what one software engineer came up with:
Objective: Improve my coding skills and achieve a mid-level software developer rating by the end of Q2 2021.
Key Result: Take three courses on the latest programming languages.
Key Result: Read 10 books on becoming a great software engineer.
Key Result: Achieve my certification as a DevOps professional.
Here’s another example from a client of mine, for an individual contributor in marketing:
Objective: Make a greater impact on our online advertising campaigns by the end of Q2 2021.
Key Result: Launch 30% more campaigns.
Key Result: Hire a higher-caliber design agency.
Key Result: Improve reporting accuracy to the business units by 50%.
While these may seem like valid goals, the problem is that none of them actually measure whether these employees have become better programmers or marketers. For key results to be effective, they need to measure a change in the behavior of the target audience of your work. This is what makes them an objective measure of success, and not simply a documentation that energy has been spent.
In both of the cases described above, the supposed key results aren’t actually measures of behavioral change at all. They describe outputs — work that the engineer and the marketer will complete in the hopes that it will help them to achieve their objectives — but not indications that value has actually been added. Simply executing these initiatives doesn’t help assess, objectively, whether there has been any improvement in the employee’s skills or their ability to add value to their companies or customers. They are simply activities they hope to do by the end of Q2 (activities which may or may not have any real impact on anyone).
Next, let’s look at the second failure mode. In the case of the engineer, if she changes her key results to be more aligned with her actual work output, they become easily gameable:
Key Result: Reduce the number of bugs I ship to production by 50%
While reducing bugs certainly should be a goal for every software engineer, if employers tie performance reviews and compensation to achieving key results like these, their employees will simply choose less risky work to do (a mistake that many organizations continue to make, despite experts’ advice to the contrary). After all, taking fewer risks is likely to reduce innovation, ultimately limiting the success of the team and the organization as a whole.
Similarly, the marketer might attempt to rework his key results as follows:
Key Result: Ensure all marketing campaigns pass legal approval on the first try.
Again, a key result like this will lead the marketer to take fewer risks, in order to ensure that all of their campaigns get approved on the first attempt, thus reducing creativity and eliminating the potential for significant performance improvements.
To address these two failure modes, the OKRs would have to be reworked to identify goals that are relevant not just to an individual’s work, but to the overall product or initiative that the individual is working on. This is why OKRs are so powerful — and so difficult to implement. The measure of success is not what the individual does (the output), but how those who interact with the individual’s work change their behavior (the outcome). And goals that are scaled up to this level by definition cannot be achieved individually. They must involve multiple team members.
For example, a more meaningful goal for the engineer might be to improve the accuracy of search engine results by 25% — but that’s not something she can achieve on her own, since it would involve not only the coding of the search algorithm, but also the design of the search results page, the personalization of results to specific users, the analysis of large quantities of data, and many other components. Truly impact-focused goals almost always require a coordinated team effort, and cannot be achieved by any one individual alone.
One interesting caveat to all this is that in our personal lives, the OKR framework can actually work well on an individual level. This can apply to personal goals, family life, career development, or any other element of your personal life. For example, you may set a personal objective for yourself to build a greater reputation in your industry. The key results in this case could be how others perceive your reputation, or more concretely, what actions that perception motivates them to take — actions such as subscribing to your newsletter, connecting with you on LinkedIn, or offering you a speaking slot at a conference. The key here is that the consumer of your efforts (i.e., the person whose behavior you’re trying to change) is a third party — not yourself.
But in the workplace, OKRs are a team-based goal setting methodology. A shared objective and quantifiable metrics can help a team to coordinate their activities, align with stakeholders, and act with more than just their own immediate goals in mind. Within this framework, success is measured not by what any one individual does, but rather, by the impact of the team as a whole on the users of the products and services they’re building. As such, instead of attempting to define OKRs at an individual level, it’s far more effective to take a team-level view in which performance reviews and compensation are tied not to individual goals and metrics, but to the extent to which individual contributors support their team’s objectives and key results.