OKR Process
A Technique For Planning and Ensuring Resource Alignment
The Problem
Every company struggles with resource alignment. Even in small startups, as people focus on different areas of the business, competing priorities will naturally arise. Sales will focus on meeting this quarter’s targets. Engineering will focus on building “the next big thing,” or perhaps stabilizing existing products, or maybe shipping the features they were supposed to ship last quarter. Marketing will focus on developing and growing interest around existing products, or perhaps “the next big thing” some of the developers are off building, or maybe both. As a company grows there will be competition for the same resources, even within a department. Each group or role within a team will have its own goals.
Is it possible to get sales, marketing, engineering, product management, and quality assurance into alignment and working together? For small companies, alignment comes naturally — everyone sits together (or talks as a group). There is usually a single person, hopefully the CEO, who owns and evangelizes the vision.
A single decision maker can achieve alignment, but at the cost of contention. As the company grows, the decision maker will be stretched thinner and thinner. This contention will lead to slow decisions and inefficiency. It also leads to friction amongst team members.
Is it possible to empower teams and individuals to make good choices without needing constant approval? Is there a technique that enables alignment to scale to larger groups of people?
What if we apply agile methodologies to “speed up” the process? The misalignment and contention will probably be further exacerbated. Changing priorities between sprints at the team level could negatively impact the overall business strategy. Getting the correct stakeholders in the room will be a constant struggle, and meetings will have dozens of stakeholders trying to keep tabs on the current state and direction of work. How do we ensure teams can be nimble and agile, yet ensure they are still headed towards the correct destination? Agile is a great tactical management tool, especially at the team level. However, without higher level coordination, agile will just add fuel to the fire.
There is a solution that turns out to be excellent for solving these problems — Objectives and Key Results (OKRs). OKRs provide us with a framework for high level planning and resource alignment. They give us the structure to negotiate and communicate priorities. They force conflicts to the surface so they can be resolved quickly and early in the process, rather than when it is too late to address them. Because there is a clear “destination,” teams are empowered to focus on the highest value work, so agile project management is more effective.
OKRs — What are they?
An Objective defines a top level “business” outcome. Each Objective has a set of measurement criteria, Key Results, used to evaluate progress. OKRs are typically hierarchical, starting from the company level, then departments, and perhaps teams and even individuals. The hierarchical structure helps ensure alignment both across and within departments/teams.
Writing good OKRs is hard; OKRs require thought and iteration. A well written OKR, however, will align resources and empower contributors to make good decisions that support the business’ needs. These decisions will be possible with less direct input from leadership — enabling your organization to better scale.
OKRs were made popular in Andy Grove’s High Output Management, where they were referred to as Management By Objective. They were later famously employed with great success by Google. Other prominent tech companies, such as LinkedIn, Oracle, and Zynga also leverage OKRs as a tool for aligning their organizations.
OKRs — What are they not?
OKRs are not a task list. A task list is a separate project management tool. Task lists are valuable, but they are tactical in nature; they specify an implementation plan. OKRs are strategic: an action plan to realize a vision. Agile methodologies, such as Scrum or Kanban, work well in conjunction with OKRs. The OKRs provide the high-level strategic vision (and measurement criteria!) while the “tactical” project management method manages the day-to-day work needed to reach that vision.
OKRs are not a substitute for employee performance evaluations, even when developing individual-level OKRs. Employee evaluations are a reflective evaluation tool. Individual OKRs are useful to align an employee’s activities and efforts with business needs.
OKRs are not set in stone. OKRs communicate business priorities and must be dynamic — just as businesses must be. If the OKRs are updated mid-cycle, changes should be actively and clearly communicated to stakeholders. Changes should not be communicated only via updating a document or dashboard since others may be depending on those outcomes. Note that if the planning process is working well, mid-cycle OKR changes will be infrequent.
What does a good Objective look like?
The first component of an OKR is a clear objective. Articulate the ideal state of the world. Be aggressive — slightly past what feels realistically attainable. The objective must be clear and unambiguous. Setting an aggressive objective is meant to drive and motivate people to achieve. Think of the objective as an outcome rather than a deliverable, solution, or feature.
A good objective is clear in its intent. It encapsulates intention without specifying how. Again, OKRs are strategic. Leave tactics to the implementors. This gives the implementor the ability to achieve the desired outcome via the best means available. Let us use an example of improving our account creation process.
Objective: Increase daily user signups by 50%.
This objective is clear, concise, and unambiguous. It does not specify the implementation details — only the desired outcome. Remember, the objective is about the desired outcome, not a specific “task.” For instance, “add Facebook login integration” would not be a good objective.
What do good Key Results look like?
Key Results serve as grading criteria and will be used to evaluate progress toward reaching the objectives. Because they are grading criteria, they should be unambiguous enough for you to periodically assess the likelihood of reaching your objective.
Key Result 1: New users can create an account in under 1 minute.
Key Result 1 provides clear criteria for evaluating results. If signing up for the service takes 30 seconds, expectations have been surpassed. If signing up takes one minute, expectations were met. However, if signing up takes 5 minutes, the target was badly missed.
To make evaluation clear and unambiguous, specify a grading scale. The scale should specify a range: no progress (0.0), minimal progress (0.3), expected result (0.7), or exceeded expectations (1.0). It is also possible to have a binary result: incomplete (0.0) or complete (1.0).
Key Result 1: New users can create an account in under 30 seconds.
• 0.3 (minimal progress): User signup takes 2 minutes.
• 0.7 (expected result): New user signup takes 1 minute.
• 1.0 (exceeded expectations): New users can sign up in under 30 seconds.
Evaluating the result against this key result is easy. In fact, it makes deciding on whether or not to pursue a certain solution easy. If a solution will only reduce the signup process to 3 minutes, it might not be worth pursuing unless it is a stepping stone to a better solution. This formulation leaves the team free to choose the best solution — which might mean buying a solution rather than building it. Tell your team what you need, not how to get it.
Astute readers will note that we need data to grade ourselves. In fact, if we do not have a way to measure the current account creation time, how can we establish success or failure criteria? Our first OKR might be to implement a system to collect measure account creation time. OKRs will help drive us towards more objective measures. Remember, behaviors are optimized for the perceived measurement criteria. Be sure to incentivize the correct behaviors.
Poorly written key results lead to discussion and debate over the intent of the key result and its associated evaluation criteria during the review process. The goal is not to write the tersest possible key result; it is instead to write the clearest possible key result. If the Key Result had been articulated as “improve the account creation process” how could a team determine what work should be done or evaluate their progress?
Developing OKRs
The OKR development process is iterative and should include the teams and people who will be working to deliver the OKRs. This is a back-and-forth negotiation. If the business sets objectives teams do not feel are attainable, everyone is set up for failure.
The business leadership team should first specify the top-level business objectives and key results. The department and team objectives and measurement criteria (key results) should then be negotiated until the people doing the work are confident they can deliver the promised results and the business leaders are satisfied the commitments will help reach the business’ objectives. If the commitments do not meet the business’ needs, additional investment may be required to meet the objectives.
When negotiating OKRs, there should be cross-team debate and discussions on the importance and priority of work. These discussions are critical to ensure resources are focused on the most important work and everyone understands the priorities. The back-and-forth across teams should be similar to the back-and-forth between departments and the company level OKRs. Every team’s outcomes should impact the top-level OKRs.
Developing OKRs is challenging. The first pass will likely be either 1) ambiguous and vague or 2) a task-list. Iterate to work through this stage. Improve the objective by asking the question: “what is the business value provided?”
The same rule applies to key results. Almost certainly they will be vague and hard to evaluate or over prescribe the “how.” The goal is to make key results easy to evaluate, as discussed above, but not prescriptive. Avoid locking a team into a specific implementation. Teams should be focused on delivering the outcome, not the current idea about how to deliver that outcome. By focusing on the outcome, teams can explore a wide range of solutions.
Applying OKRs
Applying OKRs is a continual process. Higher level OKRs (company level, for example), might be defined on an annual basis and refined each quarter. As you move down into the organization, OKRs will likely be defined and reviewed quarterly.
After the OKR definitions are well written, the teams are confident they can deliver them, and stakeholders believe the OKRs support the needs of the business, there should be periodic reviews. For quarterly OKRs, a review frequency between weekly and monthly works well.
The periodic reviews are used to assess the likelihood of delivering each key result. It is imperative that stakeholders know if they need to make alternative plans to achieve objectives a team was supporting, but will fail to deliver. If OKRs are aligning resources correctly, a team missing their targets should result in other teams being impacted — due to the hierarchical nature of OKRs.
Common Patterns
Over time, common patterns will emerge. Here are a few to watch for and suggestions on how to address them.
A team consistently missing their “expected results” (0.7s) but hitting their “minimal progress” (0.3s) indicates a planning issue. The team is being overly optimistic, which might be because they do not know how to correctly quantify their work. It might also be an indication the team is overloaded and needs additional resources. In either case, management should hold them accountable for setting realistic expectations in the OKR Development Phase, or to deliver their promised results.
Setting the same OKRs each quarter with subtle language changes is another indicator of an issue. This issue commonly appears after a team fails to make significant progress. The lack of progress may be due to poor planning (failure to break work down or think through designs), frequently changing requirements (making progress difficult or impossible), or too many interrupts (such as production support). The takeaway in this case is that no meaningful progress is being made and the cause should be explored and addressed.
Teams that consistently deliver their “expected results” (0.7s), and especially those that consistently deliver their “exceptional results” (1.0s) are not being aggressive enough. Encourage the team to be more aggressive. Occasional “minimal progress” (0.3) results are expected and desirable!
Teams working on a long-term initiative can still utilize and benefit from quarterly OKRs — treat the initiative’s business value as the objective. It is important to note that it is a long-term effort, because the objective may change very little each quarter. The initiative can then be broken into smaller deliverables, which become quarterly key results.
The OKR process is meant to help resources focus on the work most valuable to the business and reduce low-value work. If a team is consistently missing their OKRs and performing work that does not align with their OKRs, either the planning process is failing or the team does not understand that the OKRs are set to support the business’ most important needs.
Conclusion
Project management has many techniques, each useful in different situations and for different purposes. OKRs are one tool in the toolbox. Good OKRs improve alignment within and between teams, identify business operations issues, highlight resource deficiencies, and focus efforts on work that provides the most business value.
Learning to write and apply OKRs takes practice and continual improvement, but after a few quarters the results will become apparent. OKRs quickly identify teams that are planning and executing well while still being aggressive. The converse is also true; teams that are struggling will be evident. Highlighting struggling teams allows management to assist teams that need the most help. The result: a higher functioning organization capable of consistently delivering results.
—
We use OKRs at Real Kinetic to align our team and deliver value to our clients. The Real Kinetic team is passionate about helping our clients develop great engineering organizations. Learn more about working with us.