At Google, we are used to thinking from a strategic perspective. Help us communicate, quantify, and achieve those ambitious goals by using the Goals and Key Results tools.
Our actions determine the future of Google.
As we have seen many times, Google search, Chrome browser and Android, these product teams consisting of only a few percent of the company’s employees.
As long as we can set common goals and act in concert, we can subvert the entire mature industry in less than two years.
So, as employees and managers of Google Inc., how do we allocate our time and energy?
On these issues, we need to make conscious, cautious and wise choices, which is crucial.
This cautious choice is reflected everywhere in the OKR (Goal and Key Results Approach, a set of management tools and methods for identifying and tracking objectives and their completion).
At the same time, this approach is also an effective means of coordinating individual actions to achieve strategic goals.
We use the OKR management method to develop employee production plans and track their progress and program implementation.
At the same time, we also use OKR to coordinate the important events between employees and teams and the priority order of business.
In addition, we use OKR to help people focus on the most important goals and avoid being distracted by urgent secondary goals.
OKR is very ambitious, it is not gradual (not all must be implemented).
First, design an effective OKR
The design of OKR should be effective. Setting up an OKR that cannot be completed or cannot be controlled is a waste of time. It is just a form of formalism in management.
Effective OKR is an inspiring management tool that helps teams identify what is important, what is optimal, and what trade-offs need to be made in their daily work.
However, designing an effective OKR is not an easy task. You should pay attention to the following basic rules.
First, the goal is “what.”
Clear goals and intentions. Be enterprising, but recognize the reality. The goal must be tangible, objective, and clear. For a rational observer, it should be obvious whether the goal has been achieved. The successful implementation of the goal must be able to bring clear value to the company.
Second, the key outcome means “how to do it.”
Set measurable milestones and, if implemented, advance your goals in an efficient manner.
Be sure to describe the results, not the behavior (activities). If OKR contains words such as “consultation”, “help”, “analysis” or “participation”, such description refers to behavior (activity).
Instead, OKR should describe the impact these activities have on end users.
For example, “release the average latency and tail delay measurements for the six storage units of the distributed file system Colossus by March 7” instead of “evaluating the latency of the Colossus system.”
Must contain complete evidence. This evidence must be available, credible and sensible.
For example, the evidence should include: change lists, document links, notes, and published measurement reports.
Second, cross-team OKR
Many of the important projects in Google are from the contributions of different teams, and OKR is just right for coordination between teams.
The cross-team OKR should be done by all the teams involved in the OKR, and each team’s OKR should be clearly presented in the OKR of the entire project.
For example, if the ad development team, the ad operations maintenance team, and the network deployment team must all support the new ad service project, then each of the three teams should have their own OKR to describe their responsibilities in the project.
Third, OKR reading, understanding and execution
OKR mainly has two forms of expression, the promised OKR and the vision type OKR.
Committed OKR is the OKR that we will definitely achieve. It is our goal to adjust the working time and resource allocation to ensure that it can be achieved.
The expected score of the promised OKR indicator should be 1.0. If the score is less than 1.0, the reason for the unfinished part needs to be explained because it indicates that the team has made some mistakes in making the plan or executing the plan.
A team that can’t get a 1.0 point on a committed OKR needs to be upgraded quickly.
Under normal circumstances, upgrading is not only a thing that should be done, but also a must.
Regardless of the cause of the problem, there are differences in the setting of OKR, differences in its priorities, or the inability to obtain sufficient time/manpower/resources. Upgrades are the best choice.
It helps the team’s management to expand the options and resolve conflicts.
In fact, every new OKR setting inevitably involves a certain degree of upgrade, because resetting requires changing existing priorities and goals.
An OKR that does not need to change any organizational activity is a universal OKR. It is unlikely to be newly set, but may not have been recorded before.
If the promised OKR does not reach 1.0 when the deadline comes, then it needs to be re-examined.
This is not to punish the team, but to let the team know what mistakes they made in the planning and execution of OKR, thus helping them improve their ability to score 1.0 points in the implementation of the promised OKR.
2. Vision type OKR
The goal of the vision-based OKR is to exceed the capabilities of the team during the quarter.
The priority of OKR should inform the team members: Where should the remaining time be used after the team commitment type OKR is reached?
In general, a higher priority OKR should be completed before the lower priority OKR.
The vision type OKR and its associated priority goals should remain in the team’s OKR list until it is completed. When necessary, you can take them from one quarter to the next.
It is a mistake to remove them from the OKR list due to slow progress, as this can lead to confusion of target prioritization, misjudgment of resource availability or lack of integrity of the problem and cannot be fully understood.
Fourth, OKR set errors and traps
Trap 1: Failure to correctly distinguish between promised OKR and visionary OKR
On the one hand, if the promised OKR is treated as a visionary OKR, it generally increases the risk of failure.
The team may not value it or change its business priorities to focus on the implementation of the OKR.
On the other hand, mistakenly using the vision type OKR as a promised OKR will artificially create obstacles, and as a result, the team will not be able to effectively find a path to achieve the OKR, and may lead to a reversal of priorities, making commitment to fulfilling commitments. Type OKR personnel turned their attention to the vision type OKR.
Trap 2: OKR as always
The setting of OKR is often based on what the team believes can be achieved without changing anything that is currently being done, rather than based on what the team or customer really wants.
Trap 3: Vision-type OKR setting
The visionary OKR usually starts with the status quo and needs to effectively answer the question: What else can we do if we have a surplus of labor and a little luck?
In addition, there is a better question: If our resources are not restricted, then what kind of world will I (or our customers) live in in a few years?
Generally speaking, when the prototype of OKR is just developed, you can’t know how to achieve this goal, which is why it is called the vision type OKR.
But if you can’t understand and clearly express the goal state that you ultimately expect to achieve, there is no possibility to achieve it.
Touchstone: When asking customers what they want, does the company’s expectations have met or exceeded the customer’s needs?
Trap 4: Carrying weight
Committed OKR should consume most of the team’s resources, but not all of them.
The promised OKR plus the resources consumed by the visionary OKR should exceed the full resources of the enterprise. Otherwise, they form a perfectly matched promised OKR.
If a team can complete all OKRs without having to use all of the team members or resources, it can only indicate that they are either hoarding resources, setting no challenging goals, or both.
This has a good inspiration for top management: reassign people and resources to teams that can use them more efficiently to ensure that they are used to their best.
Trap 5: Low value target (OKR unattended)
OKR must reflect clear business value, otherwise there is no reason to waste resources to achieve it.
A low-value goal means that even if the goal is achieved and a score of 1.0 is achieved, no one will notice its implementation.
An example of a classic low-value goal is to increase CPU utilization by another 3%. This goal alone does not directly bring value to users or Google.
However, the roughly related goal of “reducing the number of cores required for peak query by 3% without changing the quality/delay time and being able to return redundant kernel resources to the free pool” has obvious economic value, which is A good goal.
Touchstone: Under reasonable circumstances, can OKR achieve a score of 1.0 when it does not provide direct end user benefits or economic benefits?
If you can, you can reset the OKR to focus on tangible benefits.
Trap 6: Key results pointing to promised goals are not ideal
The OKR includes the expected result (target) and the measurable mark (key result) required to achieve the result.
The key is that in the key results set, all key results scores are 1.0, and the target score is 1.0.
A common mistake is that the setting of key results is valuable, but it is impossible to achieve the goal through effective cooperation.
This type of error is often confusing because it allows the team to avoid the hard commitments (in terms of resources/priorities/risks) needed to achieve critical outcomes.
This trap is extremely harmful because it prevents people from discovering the resources needed to achieve the goal in time, and cannot find in time that the corresponding goal cannot be achieved as planned.
Touchstone: With all the key results getting 1.0 points, is there a possibility that the true intention of the target is still not achieved?
If this is the case, you will need to add or re-set key results until they are successfully completed and are sufficient to ensure that the target will also complete successfully.
V. Verify the reasonableness of OKR
1. If you can write them all in less than five minutes, then their quality may not be too good, so you need to think more.
2. If the goals set are not inherently consistent, they may not be mature enough.
3. If the key outcome is expressed in internal team terms (such as the release of Foo4.1), then it may not be good enough.
What really matters is not the release, but its influence. Why is Foo4.1 important?
A better argument would be to “increasing the registration rate by 25% by releasing Foo4.1” or simply saying “increasing the registration rate.”
4. Apply real data. If every key result is made on the last day of the quarter, then you may not have actually completed the plan.
5. Ensure that key outcomes are measurable: an objective metric must be assigned at the end of each quarter.
“Improving the registration rate” is not a good key outcome. Relatively speaking, a better expression should be “to achieve a 25% increase in daily registration rate by May 1.”
Make sure the metrics are clear and clear. If “1 million users” means 1 million users in the entire target period, or is the weekly active users reaching 1 million?
6. If there are important activities in the team (or an essential part of completing the goal) that are not covered by OKR, then add them.
7. For larger organizations, the OKR needs to be graded; for the entire team, a high-level OKR is required; for the sub-team, a detailed OKR is required.
Ensuring “horizontal” OKR (projects that require multiple teams to participate) to support key results for each sub-team