Just like the titular machine, I'm back, and ready to share more stories and thoughts about my first foray into OKRs.
The last post was all about our roll-out; it talked about scouts, company OKRs and the processes involved. It was a fairly succinct description of how we tried to introduce OKRs. Tried is the important descriptor here, because at the end of the day, we did not succeed.
But the failure that I keep alluding to will have to wait for a few more paragraphs yet, because there were challenges along the way that are worth talking about first.
Then we can get to the failure.
I Am Unable To Comply
Remember when I wrote about the scouts? That grass-roots team of champions from across the entire organisation who were engaged and interested in OKRs and wanted to see their teams use them for good?
Remember when I said we specifically avoided involving traditional management roles, like those managers that are responsible for other managers or people in leadership positions who were a step or two removed from the teams?
Yeah, that was a mistake.
It turns out that people listen to their immediate managers and continue to follow their directives and work towards their goals, even if you give them explicit permission not to.
Not all that surprising in retrospect, but hindsight is 20/20.
What we failed to do was bring those leaders along with us. In fact, we might have even frozen them out a little in our efforts to engage with the teams and their members directly, which I suspect left them feeling a bit left out, a bit angry and a touch bitter.
By not engaging with them and helping them to tie their existing objectives into the new world of OKRs, we accidentally built a barrier that just made it harder for ourselves, and it showed. Motivated people within the teams would try and get OKRs off the ground and would create excellent goals working with the scouts, but as the quarter rolled on, those goals would languish because no-one in the team was paying them any attention.
The areas where OKRs had the most engagement and consistency were the ones where the manager was bought in and actively pushing them.
I still think the idea was sound (i.e. don't make OKRs a purely top down exercise, give teams the freedom to choose their own goals with the condition that they align with the company), but our execution was definitely lacking.
Unfortunately, even ignoring that issue, there were other challenges.
Anger Is More Useful Than Despair
We didn't do OKRs for very long. Looking back at the artefacts from the time, it looks like the total duration was three quarters and that is being generous by counting the first, experimental quarter as one where we "did" OKRs.
So, three quarters of goals.
Three quarters of mostly failing to accomplish those goals.
Our OKR rollup featured a sea of red and orange (i.e. scores less than 70%), which is never fun to look at. There were definitely successes and cases where teams started rough and then changed tact and hit their goals, but they were the exception, not the rule.
The thing is, that is normal and expected, at least at the start. When you first start setting highly transparent and measurable goals, you're going to be bad at it and that's okay. Talking about, aiming towards and measuring your bad goals and their bad key results very much feels like failure.
That's not to say you haven't accomplished anything though, far from it. There are still plenty of victories to wring out of even the lowest quality, lowest scoring OKR, and that was the message that we tried to send.
I'm not sure if it helped.
Its a solvable problem as well, but it takes time. Reflecting on why a set of OKRs didn't work and then iterating towards a better outcome is a core part of the process. We just didn't spend enough time failing to be able to improve enough to reap the benefits. Sticking through that sort of stuff requires discipline. An unerring belief that you are on the right path.
That takes effort and willpower though, both of which are finite resources.
Desire Is Irrelevant. I Am A Machine
Fatigue was a real killer. Beyond the emotional fatigue of setting goals and not hitting them, there was real intellectual fatigue in adopting a new way of doing things.
Though we tried to minimise the total time required, it still required many hours, spread out across the quarter. The initial brainstorming, selection and grooming process was tiring and intense and then it didn't let up, because after that it was straight into regular scoring, which itself requires thought, consistency and discipline.
As we hit our second quarter, the frequency of scoring dropped off dramatically, even at the executive level. The scouts tried harder, but that just moved the fatigue to a different place (the scouts themselves) and didn't really help.
I have two theories here:
- This was just the initial hump and we didn't get over it. With more effort and more discipline, eventually the fatigue would have dropped
- OKRs were fundamentally not delivering any value and people saw that, so started tuning out, saving their brainspace for other, more important things
But fatigue wasn't our death knell. It was a contributing factor sure, but I believe that the real end of OKRs for us was a little something that happened in 2020.
We'll Meet Again. Go!
COVID-19 was a crap time for everyone and with it came a lot of stress in a lot of different flavours. We were spoiled for choice when it came to stress: fear of getting sick, fear that the people you love would get sick, breakdown of routine, inability to connect with others in familiar ways, lack of toilet paper, etc. It really was a smorgasbord of shitty experiences.
With the cracks already starting to show in our nascent OKR process, COVID-19 was the hammer that shattered the entire thing.
As we went into the planning process for FY20 Q3 (i.e. around March/April 2020) the shit hit the fan and we dropped everything, immediately went into fire-fighting mode on multiple levels and fell straight back to the way we'd always done things: without OKRs.
I have the theory that had OKRs been more embedded into our organisation at the time when COVID-19 hit Australia proper, we would have just used them to deal with the problem at hand. But they weren't and we didn't.
In fairness, its not like we just sat back and despaired. We did a lot of different things (flipped to working remotely, added features and functionality to help our customers, changed the amount we worked to save money), but none of it involved clearly set and communicated OKRs.
But its not all doom and gloom and the death of OKRs and the breakdown of civilization. I've still got more words in me that need to get out, so come back next time to read them.
Spoiler alert: I didn't stop doing OKRs.