The autumnal equinox and the end of another quarter can only mean one thing: it’s time to look back at our OKRs and determine how we’ve done, what our performance says about our ability to impact our operating context, and what we might do next with respect to both our goals and the OKR process itself. Since my very first OKRs, I’ve now done this with dozens of teams across a wide variety of industries (genetics, banking, emergency medical services, logistics) and I’ve noticed some emergent good practices. Here’s the process I currently follow to consistently create useful conversations and capture valuable learnings.
Too often teams simply close one quarter and roll straight into the next one, sometimes taking existing OKRs into the new quarter, unchanged. This is a huge missed opportunity. Just as we use What? So What? Now What? to perform our weekly check-ins, the end of an OKR period, is a good time to refelct on how our activities have helped us to realise our most important objectives. Have we had the kind of impact we had hoped? What assumptions have we proven or disproven? Where did we get lucky or unlucky? What should we be doing differently? This kind of reflection is critical to improving our approach and achieving better outcomes.
Here’s a quick breakdown of some of the steps I take with folks:
The short version…
Two weeks before the end of the quarter:
Read on for more detail. Smaller organisations may not need such a structured or thorough approach but this process is robust, repeatable, teachable, and customisable within your organisation.
You can close OKRs whenever you’ve decided you’re not going to do any more work on them. It might be because you’ve delivered your key results, or decided to drop this OKR entirely, but usually it’s because the end of the OKR period (typically a quarter) has rolled around. If you’re using individual timeframes for specific OKRs you would close them at their expected end date.
For quarterly OKRs, start the process a couple of weeks before the end of the quarter. Teams may say “Wait! There’s still time to do more!” but don’t count on last minute miracles or heroics. Firstly, those miracles often don’t happen. Also, would you want to run your organisation around 11th hour miracles, anyway? Secondly, it’s actually quite important to provide a little downtime and space at the end of quarter to wrap up, reflect, and reset before ramping up for the next big effort. Think of it like a 3 month long agile sprint.
Scoring is useful but only in the sense that it opens the reflection process. The real power of closing OKRs is in the reflection and learning. It’s fine to have low scores on your key results but it’s not fine to have no understanding of why they’re low and no plan to do things differently in the future. OKRS are a learning laboratory. Use the experience of attempting nearly impossible things to learn how to one day deliver those nearly impossible things.
Scoring key results shouldn’t require much debate. If you’re just starting with OKRs you might not even bother with scores.
If you’ve picked clear, easily measurable key results then scoring may be as simple as opening a dashboard, looking at a report, or simply counting some quantity of things that have happened this quarter. For example, if we wanted to increase monthly revenue from $80m to $100m, or go from 1200 to 1800 customers, or increase paid memberships from $30k to $50k of monthly recurring revenue then everyone should probably know the latest values immediately because we have he data and we’ve been looking at these numbers constantly throughout the quarter.
For slightly less measurable, more subjective key results – for example, 100% of staff can explain our product to a stranger on the street – I like to use a modified “fist-to-five” vote where, on the count of three, each team member holds up fingers to indicate how well they think we’ve achieved a specific measurable key result. They might hold up a fist to indicate a score of zero or an open palm to indicate we’ve delivered this OKR at 100% or something in between. Doing it on the count of three helps build psychological safety and avoids group-think where everyone agrees with the team lead. It leads to robust conversations about outcomes and learnings which we’ll cover next.
If you’ve been using my recommended approach of measuring confidence throughout the quarter, you may notice at some point that your confidence rating actually becomes the outcome rating. In other words, if you were 80% confident that you’d reach a key result, then you might expect your final score to be around 80%.
To understand the score, it’s important to discuss and capture a few bullets or sentences about what actually happened with respect to this key result. Did get faster at closing deals or growing proteins in the lab? This is a useful step in making sure the team really does have the same definition of success and how the reality aligns to the indicator in the key result. It’s also useful for external observers to know what happened.
It might sound blasphemous but the metrics and scores really don’t matter much. Especially when you’re just starting with OKRs, don’t put too much emphasis on the numbers. Yes, we absolutely need to deliver results, and OKRs help make this happen, but it does so by creating a learning environment where we can test hypotheses and understand how much control we have in the world and what actions of ours truly have the greatest impact. This, for me, is the most powerful aspect of using OKRs and the hardest to practice. This is where you can start to embed a culture of curiosity in your organisation and make every attempt to do anything – regardless of whether it “succeeds” or “fails” – into a powerful learning event.
To do this with OKRs, I like to ask a few questions around each key result to stimulate discussion such as:
From this conversation you’ll want to distill down a few bullets or key learnings that will affect how you might approach this (or a similar) goal in the future. You may not have learnings for every key result but try to get into the learning mindset as much as possible and think about what you would share with your colleagues in other teams.
To summarise, you want to understand what happened, what it means, and what you’ll do next. Much like What? So What? Now What?.
After each team has a chance to look at their OKRs and follow the steps above, I like to encourage multiple teams – perhaps the whole organisation – to get together and share their most exciting wins and powerful learnings. Much like the OKR Festival, you can run multiple sessions simultaneously if required and give each team time to share:
After hearing from each or several teams, invite teams across a given context or department to synthesize the most important themes and learnings related to this set of OKRs. What does this tell us about the wider picture of our operating environment? Does our industry move faster or slower than we expected? Do we have unaddressed dependencies across teams? Should we reorganise ourselves or pivot? Use this discussion to consider some big multi-team moves.
Take some time to investigate the OKR process, itself. Every organisation does it slightly differently. While there is plenty of good practice, I believe it’s important to continuously inspect and adapt the process in a way that amplifies the strengths of your organisation and addresses the weaknesses.
Look at questions like:
Discuss in small groups and consider: what policy decisions and actions can we take to make the OKRs process work harder for us?
Like anything, using OKRs gets better with practice. Take time to reflect and improve and you’ll be richly rewarded!