OKRs in matrix organizations: How to successfully implement OKRs in different organizational structures
Objectives and key results can be integrated into any form of organization but they are not always equally leveraged to their full potential. The right organizational structure is essential to the successful rollout of OKRs.
We are repeatedly asked which team or organizational structure is required for OKRs. The answer is simple: "none in particular." The good thing about the OKR method is that there are relatively few prerequisites. In principle, OKRs can work with all organizational structures that exist within a company because the process aligns itself with the organizational structure at hand. Whether this is always idealis a very different question. It often results in a new team structure accompanying the rollout of OKRs. A new goal or resource-allocation system often raises many queries that have been lingering for quite some time but have not yet been answered; either because no one has felt responsible for doing so yet or because (in many cases) the answers are not so simple.
It is worth reviewing organizational structures in order to make clear which resources are available and on which issues they should be focused. In many cases, what you focus your content on will depend entirely on the possible distribution of resources.
The basic principle behind OKRs is to ensure that all efforts are aligned in the same direction. The convergence of focus happens at a higher level, which then provides the following levels with a clear orientation to utilize their resources in the most meaningful way. This system is dependent on all sub-levels knowing exactly what resources they have available to them and exactly what goals they should contribute to. This may sound obvious but, unfortunately, it often isn't.
This problem can have several causes. The most obvious issue is that for many employees it’s not clear which tasks they are responsible for in their roles and which reporting structures apply to them (and how). It can become quite a big challenge when several stakeholder groups are placing different demands on the same individuals or groups. This becomes an overwhelming problem for employees if the sum of the demands exceeds what is possible based on their existing resources. This leaves employees burdened with the task of prioritizing an array of tasks which each respective stakeholder group would likely classify as “important.” The overwhelmed employee usually ends up making a random decision on what he or she can deliver. This frustrating phenomenon is most often observed in matrix organizations.
"The matrix has always been a pain - OKRs can't change that."
In matrix organizations, several managers are delegating to the same employee, showering him or her with goals and tasks, not bothering to coordinate management needs amongst peers first to ensure a manageable set of expectations lands on the employee. The employee is then typically left with the challenge to prioritize him/herself without clearly defined authorities that would enable him or her to decide between conflicting requests. Formulating OKRs in this scenario that include the interests and goals of allstakeholders does usually not improve the likelihood that the goals will be achieved. And it does certainly not lead to employee satisfaction! On the contrary, we believe this is one of the main drivers of stress and psychological burden on employees. It is irresponsible to leave employees alone with these problems!
In regard to developing OKR sets, we repeatedly speak of an audacious yet realistic assessment of what could be accomplished in a quarter. This overall picture should be present in every OKR set. In order to be able to come up with a realistic assessment, it’s extremely helpful to get a good estimate of the current and required efforts from each area within the company. For this reason, we think it’s essential that all branches converge together at a central point in order to identify and resolve tensions.
This central focus should not be confused with rigid, cold, and hierarchical leadership. Instead, it is a means to identify bottlenecks early on - namely during planning - and resolving them through mindful focus. In concrete terms, this means that an OKR set should be clearly derived from another set on a higher level and notbe composed of numerous different ones. Why? For two reasons; A) resource bottlenecks could not otherwise be identified and B) the task of prioritization would otherwise be left to the employee. Employees simply cannot identify the larger correlations beyond their scope of responsibility, which increases the risk of them making sub-par decisions based on insufficient information.
Does these mean we are suggesting to surrender all the advantages of a matrix organization? No, not at all. The central idea of having both multi-disciplinary and organizational perspectives is very valuable. This can result in savings, increaed quality levels, or defining and adhering to standards. All this is very positive but it can come at a price. Inevitably, this is an investment decision that should be made conscientiously. Maintaining certain standards comes at the price of speed and thereby potential execution power; for example, the completion of new features. On the other hand, if you disregard focusing on quality and/or standards for too long, you risk accumulating a massive (ie. technological) legacy. Experience has taught us that there is no “one-size-fits-all” solution to this dilemma.
For example, a distribution according to fixed percentages of resources among different stakeholder groups does not result in a logical goal structure. These are decisions that need to be made on an individual and strategic basis. In one case, speed that comes at the expense of quality could bring a clear advantage; another time, such a decision could amount to massive future costs.
Since decisions need to be taken on a case-by-case basis, we have come up with a solution that prioritizes the various perspectives against each other during the OKR definiton. In a matrix organization, the OKRs are always aligned according to the content perspective, not the disciplinary point of view. Goals are therefore derived from the overlying OKR set according to the content axis, not the disciplinary point of view. As an example, a UX Designer who is part of a product team does derives the OKRs from the product team’s OKR set instead of the Head of UX’s OKRs.
Functional goals – which are often personal goals in a line-organization or focus on personal development – need to be introduced via the owner of the overlying OKR set according to the content axis, in order to maintain focus and clarity of resources.
So the teams know that– in addition to their topics – there are also structural, individual, or subject-specific topics that need to be taken into account. The disciplinary organization, therefore, shifts to a back seat and takes on a much more consultative role than before. The demands are collected and discussed at the intersections of the content teams and then taken into account by the team in the definition of their goals.
In today’s growing world of agile working, there are a multitude of organizational structures in addition to the classic matrix, which must deal with similar challenges. So-called cross-functional teams in combination with classic line-organizations, present an organization with extreme challenges, especially during change processes. The Spotify model with its squads or Holacracy structures, have similar challenges that can be solved with the same logic. In the end, each OKR set needs a clear point of reference within the organization. Other stakeholders take on a consultative role and consolidate their topics for the leaders or "links" of cross-functional teams or circles to bring them into the OKR planning process.