How we set short-term Objectives & Key Results for long term goals
For the digital product teams at Wellcome, the end of a quarter can only mean one thing: it’s OKR season.
Objectives and Key Results (OKRs) do an important job for the team. They align the team with what the priorities are for the next three months, they help us communicate expectations to stakeholders and they encourage us to reach for stretch goals. We set them quarterly. Having a 3-month drum beat for OKRs keeps us reactive and makes sure we are always focusing on what matters most. It also gives us a chance to reflect at the end of each quarter.
Every team will find their own path with OKRs. There is no ‘one way’.
We’re the product managers for two teams at Wellcome, and we’re going to take you through our process for setting OKRs in the rest of this article.
Setting Objectives in the Research Product team
Getting into a quarterly rhythm can be challenging. What about projects that span more than 3 months? How do you set OKRs within the framework of a longer term roadmap?
To manage this balance, we split OKR planning into two workshops. The first is about gathering ideas for areas of work, the second is defining the key results for those areas of work.
In the first workshop, we start by reviewing the long term vision and goals. In our team, this is the product vision — the overarching aim for the product, why it exists. For the Research Product, that’s:
Make Wellcome.org a helpful platform for everyone working to solve urgent health challenges facing everyone.
The website aims to:
1. Support funding seekers and grant holders with transparent information.
2. Be inclusive so someone can apply for funding regardless of their circumstances.
We then look at the context in which we’re working. We need to know what external expectations are and what business targets we need to hit.
We review our roadmap outcomes (the long term goals) and existing items on the roadmap. The team then adds further ideas for how to achieve those outcomes. We review all ideas, vote, and prioritise our work for the quarter.
The outcomes we’re focusing on for the quarter become the Objectives. Key Results can then be set depending on the work we’ve prioritised.
For example:
Roadmap outcome/ Objective: Researchers and Research Offices understand what Wellcome will/will not fund.
Ideas of how to support this:
A. Roll out changes to Grants Awarded search to make it easier to see the kinds of projects we fund.
B. Redesign scheme pages to ensure critical information is easily accessible.
In the second workshop, we look at the drafted key results and get feedback from the team.
Setting Key Results in the Explorify team
Once we have established the Explorify team priorities for the quarter (our Objectives) we ask ourselves:
“What can we measure to tell us whether we have made progress? Where do we want to be by the end of the next three months?”
This is the hard part. It’s an art and a science. Not every objective can be measured the same way.
Here is our team’s criteria for what makes a good Key Result:
- You can get the measurement during the quarter, not just at the end
- It’s a measure that you can expect to influence during the quarter
- It’s there to set an ambition and also to help you gauge if you are making progress. As a way to motivate ambitious goal setting, we consider 70% a success and 100% a sign that we need to set our sights higher
- Create tension between quant and qual. There is always the risk that if you are only measuring quantity, than the quality suffers (and vice versa)
- Should be specific enough that the goal posts can’t change during the quarter
Over time we have found two broad categories of Key Results: value-based and activity-based.
Value-based Key Results
Example:
Objective: The Explorify homepage helps new users get started.
Value-based key result: 70% of new teachers click on a suggested activity on the homepage in the first week of sign up.
Value-based key results are the ideal way to measure progress because they are directly linked to outcomes. These are the measures that help you know if you are on track to achieve your longer-term goal.
In the example above, we care about helping new users get started because we have a long term goal of improving our adoption rate. We know that the teachers’ first visit to the site is critical, and getting them to their first relevant activity will directly impact whether they adopt the product.
Our product goals are long term. They are ‘lag measures’, meaning that by the time we measure them it is too late to influence. We need measures we can impact in the short term that will tell us how likely we are to achieve our long term goal. Value-based key results give us a framework to focus on these lead measures.
Unfortunately, it’s not always possible to have this kind of key result. Really you need two things: access to the data and the ability to influence the measure. In the product world, it takes time to design and build a feature — often more than three months. This means you can only have a value-based key result if you will release the feature early enough and/or be able to iterate on it during the quarter in such a way that you can influence the measure.
Activity-based Key Result
Example:
Objective: Homepage is prepared to engage teachers at the start of the new school year.
Activity-based key result: Returning teachers can select a teaching topic and activity recommendations are using the topics
- 30% Design and interaction specifications ready for topics to be included in user journey
- 50% Back-end systems prepared for topics
- 70% QA: Front-end prompts new and existing users to choose a topic
- 100% Live on product and data is being captured
When you want to align team expectations for progress on an objective whose impact can not be measured in 3 months, the best approach is to break the key result down and ask, ‘What does 30% look like?’, ‘What does 50% look like?’ This way you can track your progress throughout the quarter.
While activity-based key result will reflect the tasks that the team are undertaking, this is not the best place to list out tasks. Here we are focusing on project stages of completion so that we can get a sense of progress against the objective. Another example tried in the past uses this template: ‘Complete 90% of the high priority tasks for project X’.
What we’ve learned
We know we won’t get it right all the time. It’s taken a lot of experimentation and practice to find a format for the team workshops that hits the right balance, measures that are specific enough but not over-complicated, objectives that aren’t binary, goals that are ambitious but not unachievable, and key results that aren’t a task list. These are easy traps to fall into and there’s been a lot of learn-by-doing for both our teams. Our advice to you is to have a go, and keep learning and adjusting each quarter.
