OKR examples: How to formulate good Objectives and Key Results

We are asked again and again for examples of good OKRs. We are generally reluctant to use examples, as we find it much more helpful to focus on improving the basic understanding of the rules behind formulating good objectives and key results rather than referring simply to examples. At a glance, examples are easy to understand and appear quite obvious. However, to transfer this understanding and OKR formulation to your own situation is not so easy if you don’t understand the underlying principles first.

In order to illustrate the application of OKRs more easily, we have compiled some OKR examples that can be regarded as good OKRs according to our framework. 

In the search for the ideal objective, we look for self-contained states in the future that can be judged as accomplished or not by an objective third party at the end of the quarter. The key results are the drivers of accomplishing this end-state. They increase the probability that the objective will become reality (but do NOT show whether the objective was achieved or not in the end). 

Let's assume that we want to offer electric scooters as a new mobility solution in large cities to enable customers to quickly and comfortably cover that "last mile" between a public transport station and their final destination. It is important to note that the quality of a good OKR set can never be assessed without concrete context. The identified topics are relevant for the OKR content, as it would not be helpful to convert general KPIs into KRs and to assign them target values. Keeping this in mind for our scooter company, a possible company-level OKR set from a product perspective could read something like this:  

Users can use a scooter at any Munich subway station at any time. 

• Increase the number of available scooters to an average of 1,200 per day

• Improve the prediction probability of motion profiles by 25 percentage points 

• Increase scooter operating range by 7 km per battery charge 

• Reduction of maintenance time per scooter from 90 to 35 minutes 

Alternatively (or additionally), an OKR set from the perspective of an app to use this scooter service could be the following: 

With the use of the app, new users can use spontaneously discovered scooters immediately

• Increase the location accuracy to 25 meters 

• Reduce verification process time required from 7 to 3 minutes 

• Implementation of at least three new payment methods (eg., credit card, Paypal, Apple Pay, etc.) 

• Reduce the error rate of QR code recognition by 7 percentage points 

If you Google-search examples of OKRs you will find a plethora of different formulation styles, just as you will find different opinions about the acceptable maximum number of OKRs (we insist a maximum of five objectives with a maximum of four key results each). You will also see there are a wide range of opinions about the quality of the content.

 Manybad examplesof OKRs can be found in similar forms to the following:   

O: Improve customer satisfaction 

• Increase the NPS to over 8.0 

• 2000 responses to the annual customer satisfaction questionnaire 

• Conduct 50 telephone interviews 

• Present an action plan for the next quarter 

 O: Achieve record sales with increased profitability 

• Reach 1,000,000€ revenue in the third quarter

• Achieve €100,000 revenue in two new markets 

• Increase gross profit from 54% to 67%

 Why do we think that these are bad examples of objectives and their respective key results? In our framework the objectives should be (as mentioned above) clearly defined, finite states in the future. They are not permanently desirable or continuously valid goals: these we refer to as “evergreens.” Increasing sales or customer satisfaction is alwaysrelevant and good. However, these do not help the company and its employees a) to focus, nor b) to define clear statements for the expected outcomes of the coming quarter. Since KPIs also play an important role in our framework, the KPIs do the task of reflecting this constant effort to improve customer satisfaction, increase revenue, etc. The objectives, on the other hand, refer to concrete targets to be achieved within a quarter, which have a positive effect on KPI(s) like a gross profit rate for example.

Our approach differs fundamentally with the view that key results increase the likelihoodthat an objective will become a reality; they do not indicatethatan objective has been achieved. Through the cause-and-effect relationship between objective and key results, the most critical drivers of an objective’s success are analyzed and key hypotheses are defined. Through this, both participants of a direct-report relationship have the same understanding of what will determine success. If there are different opinions between these two parties, they should be discussed and aligned in advance of the quarter. Subsequently, the teams and employees are completely free to decide how they want to achieve their Key Results. Through a clear and analytical approach, the "what" is explicitly defined together, whereas the "how" is completely left to the team members to define. This is the foundation for complete freedom for teams to run independently for three months since the key drivers of success have already been clearly identified.  

The following can be said about the afore-mentioned bad key results:  

• The NPS is a KPI whose target value in this example would have been better suited in the objective itself. If one follows the theory that the KRs would be defined as "…as measured by," then this would have been the only meaningful KR in this OKR set. From our perspective, though, this is an indicator that we would track in a KPI set, instead. 

• We understand “2000 responses to a survey” to be more of a task than a KR: As it is stated, the answers to the survey could also be negative. As such, they are neither a driver of the O in our OKR perspective, nor would they be a measurable indicator for the quality of customer satisfaction in the "…as measured by " ideology. 

• The 50 interviews should also be seen as a task and not as a driver or indicator of success.

• The presentation of an action plan for the following quarter is not ideal for two reasons. First off, the agile mindset does not aspire to makeplansfor the future and then implementthem in the following quarter. Instead, agile methodology tries to define a much smaller action plan and then implement the first possible result within this quarter. The traditional way of thinking about first "planning" then "implementation" corresponds to the waterfall model and should be avoided. Secondly, the key results should be positively influencing the objective in the present quarter as much as possible. A plan that will not be implemented until the next quarter will certainly not have any effect in this quarter. As a rule, the preparation of a plan is to be regarded as a task. In addition, it can be assumed that the responsible team (hopefully!) has a plan, even if just a rough one, which it can now implement in steps. These steps allow for immediate learning so that the lessons can be applied in the following quarter.

• Our experience has shown us that revenues do not serve as a good key result on company-level. When this happens, the entire company is left in the dark about how these sales are to be achieved. As a rule, this does not contribute to the sustainable success of the company. Instead, a short-term focus on revenues often leads to problems in the medium-term. In the above KRs, the first revenue metric is an indicator of success; therefore, a KPI from our perspective. 

• The second KR represents a subset of the first KR and therefore cannot be achieved independently. If the second KR fails, the first KR will most likely, also fail. This drastically reduces the probability of accomplishing the objective since you will now have more than one failed key results. 

• In this example, the gross profit serves as the second "measurable" dimension within the objective which is being addressed in the KRs. The goal should not contain a content link, since a final optimization parameter must be provided in situations of quick decision making. If the likelihood of achieving the goal becomes less, the manager would not know whether to focus more on sales or more on profitability. A more helpful statement would refer to profit or growth, since it would make a clear strategic statement as to whether profit or growth should be increased. The typical "reflex" of a CEO is: "Both!" If both are possible, of course, do so. But if you cannot do both, then you need a clear prioritization to define where your efforts should be directed.

• We feel much better, if we include the measures leading to these KPIs in the OKRs while tracking our business results in a rolling forecast.

Of course, the best examples come from real life.  Therefore, we always offer to comment on and revise the examples you contribute here. Simply post your own formulations in the comment section below and we will be happy to give you feedback and find lots of practical examples for good objectives and key results.