How to write great OKR objectives
Objectives are the hardest and most important part of OKRs to get right
Objectives and key results (OKRs) are a great way for you to align and provide feedback to teams about how they fit into the strategy of your organization. While I’ve seen them implemented many times both well and not well, the difference is usually based on where your organization is in the “journey” of using OKRs.
This is because OKRs aren’t something you just pick up and start doing. It takes lots of practice to get from the concept to execution. I’ve found that writing great objectives is really hard. This is because of the focus on outputs from the team, rather than understanding why you are doing something. For many leaders this is a huge change in mindset.
In this article I will talk about how objectives fail and fall short of their intended purpose. Then I’ll talk about how teams learn to write great objectives and examples of the steps along the way. Learning about these steps will help you focus on iterative improvement of objectives instead of fixating on what is a perfect objective one. Since they need to be adopted to your circumstance it is highly nuanced.
I hope this article helps you speed through these stages faster if you are aware of them and to write great objectives sooner.
In OKRs, perfection is definitely the enemy of progress, so let’s get started!
Failure and progress of objectives
Objectives can fail in many ways:
- Lists of things to do — this is the classic issue with output driven objectives.
- KRs in disguise — a “smell” for this is having a number in the objective.
- No progress is made — if they can’t find leverage over the outcome because it is too broad.
- Aren’t worked on — was it inspirational enough for people to think about it.
I’ve found that people progress through the following steps when trying to write and interpret objectives for their teams:
- As initiatives
- As a business impacting metric
- As progress towards an outcome
- As end state outcome
- As inspirational end state outcome
It is a learning experience to get through these steps but can be sped up if you are aware of them.
As initiatives
This is the starting point for most organizations. The progress of objectives starts with a very simple problem: you think about the initiatives you need to do rather than the results that you are aiming for. It is a natural part of the work to think differently about the way that teams connect with strategy. I’ve found that they are so used to having leaders tell their subordinates what they need to do that they don’t think of the reason for doing something.
When you are first asked to make objectives you back into them by taking their most important work and putting it as an objective.
For example, say a team is working on recommendation engines at a movie streaming service. If they are asked to write some OKRs they may create an objective as initiative like the following:
Deploy updated recommendation microservice that takes a user’s Tweets about movies into account
As you can see, this is not an objective. It is some work that the team was going to do anyway. If you know the reason why you are doing the work you may not need to do this initiative if you find something better.
The first question I would ask if “why” are you doing this? This will usually get us to the next manifestation of an objective for this team…
As a business impacting metric
Objectives are commonly mistaken as something that will make the leaders of an organization happy. Those leaders often talk about the business impact that features will have in all-hands. This includes more revenue, less cost, more customers. They are all the business impacts that you and your executives want but aren’t good objectives.
The hypothetical team on the movie streaming service might think of the following after their initiative driven one earlier:
Reduce member churn by 10% with updated recommendation service
What is wrong about these types of objectives is that they really should be KRs rather than an objective. Business impacting metrics are following metrics not leading ones usually. They tend to show that we have made a good change and can show progress but not the objective. You can put these aside as possible KRs later.
If you see a number in the objective it is a hint that this objective may actually be a KR.
As progress towards an outcome
After this stage we usually make an important breakthrough and we get to a real objective: one that is focused on a real outcome for someone. A way to get to this point faster is to focus on a particular problem that a particular person has rather than any system you are building. When you force the perspective to be problems your customers, partners or employees have you can much easier understand what is the objective that would solve that problem.
Once you make the transition to talking about people’s problems you may mistake progress towards the fixing of that problem, rather than the end state itself.
For example, the movie streaming service group may draft an objective like this:
Enabling members to get recommendations related to what they are talking about on social media
This is a great start towards the right objective but it is really just the process towards what the team really wants.
As end state outcome
It isn’t too far to an actual end state once you are targeting the progress towards an end state. You just need to follow it to the it’s conclusion that you want for some person.
As an example, the video recommendations team may settle on the following objective.
Member’s find the content they have been talking about at the top of their watch list
At this point, I’ve found another common issue at this stage is to make an objective too broad. The right level of breadth is hard to make but it should be focused on what the team believes they can make change around.
For example, this video streaming recommendations team could have gone with something like this:
Member’s find new content to watch faster
This is way too broad for their ability. They need to focus it down to what is an appropriate trade off for their team, not the whole company.
As inspirational end state outcome
The last step is often overlooked but very important. We need to make our objectives be inspirational and aspirational to motivate teams. We need these objectives to be something that is memorable in prioritization discussions and part of the regular consideration of whether they are working on the right “thing.”
For example, the video recommendations team could rewrite their objective to get at why this is important to people:
Member’s are excited to watch content they have been talking about
The prompt is more about emotion and qualitative experience that is exciting for the team. You should be figuring out what is motivating so that we are more likely to remember and consider the objectives when making hard decisions.
Bonus: as a networked, inspirational, end-state outcome
At this point in the journey we will want to consider an additional step: networked into other OKRs for the organization and other teams. In a previous article I made the case for networked OKRs, rather than those that fit into a hierarchy.
OKRs aren’t just a way for you to set a team’s goals more effectively but to also create a dependency graph with other teams that you may depend on (or that depend on you).
In the video recommendations team they could share their objective with other teams:
- A user experience that needs to think about how people see the new recommendations
- A marketing team that may be trying to reduce churn
- An organization-wide objective if it is a high priority for everyone
You should always be considering how your work fits into the larger organization and how you support those other teams.
Now you just need to figure out your key results…
The road from initiative-driven objectives to those that are outcome-oriented and inspirational isn’t something that happens overnight. You need to continuously reflect and think about what your team is really trying to do.
You should be always looking to make your objectives more helpful for getting feedback from the product strategy to the teams that are using them. The best way to do this is to make them networked, inspirational, and outcome-oriented. If you don’t you will find that OKRs don’t meet the promise they could have been for your teams.
