I may have found a way to improve OKRs. Well, at least for me. Maybe it works for you as well.
After failing dramatically with my professional OKRs in the first quarter of this year (hint: I was no exception in my team), I want to take a moment to evaluate how the practice works for me, and how it doesn’t. Let’s do a Perfection Game with OKRs!
I also appreciate that I am forced to think of measurable results, and to evaluate my own performance on those metrics. It’s what professional athletes do all the time. So why not expect the same from professional workers?
How OKRs Could Be Improved
OKRs, according to the online video, have a quarterly rhythm by default. However, my work is highly dynamic and I feel the need to also do something that happens monthly. My team member Sergey felt the same and experimented by doing OKRs strictly on a monthly basis, but he found those timeboxes to be too small. I need a solution between those extremes.
I found that committing to, and evaluating, several unrelated things in one synchronized quarterly cadence is too rigid for me. Some things can be done in two months; others are better stretched over four. I feel the need to decouple my little projects, similar to software teams decoupling features by switching from Scrum to Kanban. I need a continuous flow of OKRs.
I found the self-assessment at the end of three months a bit demotivating, because it said I had “failed” all my targets. It would be more accurate to say that my ambitious targets cost me a bit more time than expected. I dropped some projects, agreeing that my results were smaller than I hoped for, but still satisfactory. Other projects I keep working on because they’re doing fine and just need some more time.
I commit to no more than three slots on my projects/OKRs board. For example, I may commit to create new keynotes, sell more books, and promote Happy Melly membership. The tasks in these projects get a high priority for several months. All other activities get lower priority.
As usual, for each project I define an objective and several targets (key results). However, each target must be defined as something that I can measure/evaluate every month. For example, the targets “sell 1,000 books per month” and “have 500 paying members” are valid. I can check those each month.
I evaluate the scores on my targets each month (not once per three months) and then I decide which scores are good enough to have the project replaced by another. For example, a score of 58% for “sell more book” may be disappointing, but it could be good enough to drop the project in favor of another. But I could also keep the project and aim for a higher score next month.
When projects get dropped, I fill the newly available slots with other projects. This means I can commit to new OKRs any month, and I can keep them in focus for as long as I want. The special attention trickles down into my GTD/RtM task list, where I usually keep prioritized projects capitalized to remind myself of my current OKRs.
With these small adaptations, I hope I will achieve better flow of my projects instead of requiring that they are in sync. It also means that I have a monthly instead of quarterly cadence, while not aiming for OKRs to be successful within a timebox of just one month. Finally, I like the idea of not “failing” when my estimates were simply too optimistic. I “fail” only when I drop a project with a low score in favor of something else. This matches the idea of “I tried but it didn’t work”.