Josh Seiden and I are writing a new book on OKRs. If you want to stay up to date on how that’s progressing, join the book launch group to read early previews, provide feedback and help us choose a title: www.okr-book.com.
The majority of the writing on OKRs on this blog has been focused on the daily activities of teams working with objectives and key results. It’s also dealt a lot with the tasks and mindsets that need to change with leadership and management teams if this human-centered goal-setting framework is going to succeed. One thing we’ve not yet discussed here is how these two personas come together in a cadence that supports the needs of both. What follows is a rough draft for what the OKR cycle might look like in a typical organization. I am eager to hear your thoughts, feedback and suggestions.
Why do we need a cycle?
We work in a continuous world. Every system – product, service, customer experience – is evolving as new information is learned. Our ability to sense that information is unprecedented. This gives us the ability to respond to this change as quickly as we want. Linear mental models are outdated. Instead, thinking about our work, and especially the way we measure success as a continuous cycle serves us well in this environment. OKRs need a cycle because teams are explicitly required to do product discovery work to truly harness their benefit. Trying to predict well into the future what we’ll work on, what we’ll learn and how we’ll adjust course is impossible. A cyclical mindset helps teams build an approach to their work that provides maximum agility. Let’s take a look at the OKR cycle.
As mentioned above, this is a rough first draft. I’m eager to hear your feedback in the comment.
Define strategy
This is a fundamental first step and is required to be done by the leadership of the organization. To align all the teams in a similar direction your strategy needs to be clear and well understood broadly.
Identify strategic key results
Once the strategy is in place, in most cases it can be used as the company’s strategic objective. Next, leadership must determine the strategic key results. What are the high level measures of success for the business (based on the strategy) in the coming 6-12 months. This gives the teams a specific target to support in their work.
BU’s and Teams set their OKRs
Having understood the strategy and the specific key results their leaders expect, teams can now start coming up with OKRs for their own work. In most cases, these OKRs are supporting the organizational strategy. They function as leading indicators to the strategy but focus on goals the teams can achieve as independently as possible. Once set, teams share their proposed OKRs with their leadership teams to ensure alignment and accurate focus.
Determine solution hypotheses
OKRs don’t tell teams what to build. They tell teams what behavior to change. The teams therefore need to brainstorm what possible features, products, services, initiatives, etc might help them change that behavior. There are a variety of ways to facilitate this type of activity. It should always include any existing backlog items that the team still believes may help them achieve their new KR goals. However, it’s important to make sure that the team isn’t reverse engineering their backlog to meet their new goals.
Begin product discovery work
Since our solution hypotheses are unproven to some extent we now have to de-risk them. To do this the teams start talking to customers, reviewing data and analytics reports and running lightweight experiments to see where they were right about their solution hypotheses and where they might need to pivot. The learning informs the backlog prioritization and helps the team divide their time between delivery and discovery work.
Discovery/delivery
With learning flowing in, the team can now start building and shipping features. Over the course of the cycle the percentage of time dedicated to delivery and discovery will fluctuate. Each activity generates insights that help the team determine whether their original prioritization holds for the quarter or if there needs to be a pivot of some kind. At all times, the determining factor for any decision is progress towards the team’s key results.
Quarterly check-in
At the end of the cycle (quarterly in this case, but adjust to a cycle time that makes the most sense for your team) the team gets back together with their leadership to review progress. Everyone who comes to this meeting should be well aware of what the team has done, learned and decided in the previous cycle. This should be a forward-looking meeting, not a review of the past efforts. The conversation in this meeting focused around what the team plans to do with the knowledge they’ve acquired in the last quarter. Do the OKR’s still make sense? Are there fundamental changes to the plan? This is where those decisions are made with a time horizon of “the next cycle” as the container.
..and the cycle repeats itself
Assuming the strategic goals stay consistent for a while, teams should be getting through their OKR cycles multiple times during that time. In most cases these are annual (strategic) and quarterly (OKR) cycles. The shared theme around which this cycle rotates is human behavior. Are we positively impacting the behavior of the humans we serve? And if the answer is no, we adjust what we’re working on based on the insight at hand at that time. Shorter cycles are better in this case. The shorter the cycle, the sooner we can course correct AND the less we’ve invested in an idea (making it easier to pivot). Once the company’s strategy shifts, teams will have to spend a bit more time realigning their goals again before kicking off another set of cycles.
What do you think? Have you seen this model work? What models have worked for you and your teams? I’d love to hear your thoughts on this.