Are you using the OKR (Objectives and Key Results) framework in your organization to drive goals? Check if you’re leveraging their full potential!
Here are the top learnings that I acquired after several quarters of working with this model.
1. Do not decompose OKRs too deeply.
OKRs are not meant to be task lists or replace backlogs. Their purpose is to set a common direction for a wide group of people.
If you have dozens or hundreds of small, team-level OKRs in the organization, that purpose will get likely get lost. According to some OKR mentors, companies under 100 people shouldn’t decompose their objectives at all – just do them at the company level.
2. Make OKRs ambitious.
If you defined them right, you will likely not achieve many of the objectives – and that’s OK! This is a different model than SMART, for instance, where you are supposed to ensure that goals are realistic.
Here the confidence level around 50% seems to be the sweet spot. Much higher than that usually means sandbagging, and the organization ends up being driven by fear rather than by opportunity.
You will be afraid of feeling like a failure, or of your people getting burnt out by frequently missing the objectives. But get past that! It’s worth it, as only a very high bar will drive people to accomplish big leaps.
3. Don’t measure performance through OKR achievement.
This will automatically cause OKRs to not be ambitious. There will arise a pressure to make them easily, safely achieved.
Don’t compare teams based on that, don’t make bonuses dependent on achieving them, and don’t use them for individual performance reviews.
4. OKRs should be cross-functional.
If you have separate marketing OKRs, engineering OKRs, sales OKRs – you’re doing it wrong.
OKRs are about getting people together to achieve a common goal. If you split them between functions, you will end up with a siloed organization – which is likely exactly what you wanted to avoid when you started with OKRs.
5. Make OKRs measurable.
You need to be able to determine whether an OKR was achieved, and how much of it was achieved. Defining good Key Results in quantitative means is the key for that. They shouldn’t be binary, that often indicates tasks (‘do X’).
And you should be able to grade them, e.g. on a sliding scale of 0-1 (0.6-0.7 being a reasonable target). Ideally, even if you achieve 60-70% of the original (ambitious, remember?) goal, that will still be valuable.
A well-defined Key Result can be a great feedback tool throghout a quarter, when people need to see if their ideas are making the right impacts (moving the needle), or if they need to explore alternative ideas.
6. Measure outcomes, not outputs.
It doesn’t matter how many features are delivered in your product if they don’t make the expected impacts. A good OKR is oriented around the actual outcome (increase of usage, sales, satisfaction, conversion, etc.).
Instead of saying ‘deliver feature X’, seek a key result that measures the impact of that feature, e.g. ‘20% more users decide to try out the premium subscription’.
Sometimes you will need leading indicators as well, when the influence over the ultimate outcome isn’t direct, and that’s OK – but you can still express them through outcomes (like: ‘Double the number of subscription page views’) instead of outputs (‘create a fancy feature X that will attract users’).
7. Do not set OKRs top-down only.
A dictatorial approach causes disengagement, cancelling out the inspiring effect of an ambitious objective.
Companies experienced in OKRs claim that a combination of top-down direction and bottom-up feedback works best, with the sweet spot being around 60% in terms of objectives coming from the teams.
Another reason is that it helps with relying on data instead of stakeholder or customer opinions. Remember, majority of their opinions are wrong! Bottom-up involvement helps shape them in a less biased way.
8. Don’t give up and keep improving every quarter.
The first quarter after introducing OKRs will likely be horrible. People will think that it’s just another disguise for KPIs, and you will end up with many output-based, non-measurable goals, that are really just unvalidated feature ideas from stakeholders and higher management.
The second quarter will be a bit better. Patterns and antipatterns will start to emerge, learnings will appear in people’s heads. There will still be many mistakes, but more and more people will start to see the value and real purpose of this tool.
That has been my experience as well. While I initially felt that the concept makes sense, it took me a few months to really understand how to apply it and get value from it.
And once I did, I went all in. I even started to apply this approach in my personal life, and it has made a huge difference for me already. But that’s a story for another post.
What are your experiences and thoughts about the Objectives and Key Results framework? Are you interested in learning more? Leave a comment below.
