Last time on this Terminator inspired series of blog posts, I wrote about goals, gave a crash course in OKRs, talked about my previous company's desire to implement OKRs and then finally described how I got entangled in that implementation.
Hopefully you enjoyed reading it, otherwise why in gods name did you come back for a second helping?
Well, you're here, so I might as well make the best of it. I promised to write more about our OKR adventures and I'm a man of my word, so this post will detail our approach and the processes that we established to get the OKR ball rolling.
Without further ado, let the rambling commence...
Our Mission Is To Guide You
When we started with OKRs, we didn't want them to be another one of those classic process change failures, where people roll their eyes, nod and then continue to do exactly what they were doing. You know the ones, where a tone-deaf leadership team pat themselves on the back for a job well done but nothing actually changes because people just don't care.
The reason should hopefully be obvious: you don't get real change unless people are engaged, and people don't get engaged if they feel like something was foisted on them with no rhyme or reasoning. Its much better if they opt in voluntarily, or at the very least, have been consulted and their opinions and thoughts taken into account.
As a result, we shied away from engaging with traditional management roles when it came to pushing out OKRs across the organisation. Instead, we established a small group of people who were mostly not managers, pulling them from a variety of areas and roles.
We called them the OKR scouts, named as such because they were to "scout" the way ahead for others. They were to be grass-roots champions, helping their own teams adopt the new patterns first and then branching out to help others.
It was a good plan. The scouts were mostly volunteers and were engaged and interested, which is a great combination. They wanted to improve the way that their teams did things and deliver more value for the business.
Leading by example, the scouts created OKRs of their own, communicated them to the rest of the organization and then tracked them religiously. They weren't the best OKRs, but they were a great start.
Objective: We as a business are aligned and heading in the same direction
Key Result 1: At least 70% of the organization have walked the OKR trail by 6 September 2019
Key Result 2: In a company wide survey, 90% employees report that their work for FY20 Q2 aligns with either FY20 or FY20 Q2 company OKRs. At least 90% of the company replies to this survey
Key Result 3: OKR retrospectives are completed across all teams and the top 3 areas for improvement across the company are identified and actioned by the end of October 2019
Key Result 4: In October 2019, a random sample of 20 staff can all: a.) Name their team's OKRs, and b.) Articulate how their day to day tasks are contributing to the team OKRs"
But it wasn't enough for the scouts to have OKRs. In order to roll the concept out across the rest of the business, we needed something to help align our teams.
Trust Me On This
When it comes to alignment, company level OKRs are critical. They help to set the tone and direction and act as a guiding star, allowing teams to orient themselves correctly to make sure they are heading in approximately the same direction.
I don't actually know what the process for creating and agreeing to the company OKRs was, as I didn't take part in it directly. What I do know is that the executive team eventually surfaced with two sets of OKRs: one for the entire financial year and one for the impending quarter.
Its a pretty daunting task for a team to create OKRs for the first time, so I can't imagine what that felt like at the company level. They certainly looked stressed, that's for sure.
What we got out the other end was...a good first attempt. A set of OKRs relevant to the financial year and then a narrower set of OKRs for the coming quarter. There was a lot of room for improvement (incomplete key results, unclear objectives, too many "focuses"), but it was still a vast improvement in both clarity and transparency vs what we had before.
More importantly, the company OKRs were enough to get the rest of the business started on creating their own.
Stay Functional Until The Mission Is Complete
Before we could do that though, we needed to clearly define our teams, to figure out how many sets of OKRs we had to create. It turns out, we had a lot of teams (~27), and there was some back and forth about which ones needed to have OKRs and which didn't.
Having already run OKRs in the experimental teams and somewhat proven the hypothesis, we decided that every single team needed to have their own OKRs. In retrospect, this might have been a mistake, because it meant that the relatively small team of scouts had to then facilitate the creation of many many OKRs across the entirely organisation.
Regardless, in order to create their own OKRs the process that each team went through was similar:
- Take the company OKRs
- Have a brainstorming session to come up with potential objectives
- Vote on said objectives to order them
- Repeat the process in order to create key results for the top voted objectives
- Groom said key results to ensure they are measurable and meaningful.
The final step in this creation process was a review, where each team presented their OKRs to the executive team to check alignment and quality. This was pretty onerous, given the number of teams that we had, but it was extremely useful as a sanity check. After the review, came the unveiling/announcement, where each team quickly went through their goals for the next quarter, describing them to the rest of the organisation in a company-wide town hall.
For tracking, each team recorded their OKRs in their own templated Google Sheet which was publicly visible across the company. The contents of those sheets were rolled up into another sheet to provide an entire company view. It was a pretty neat system, showing all of the latest progress and scores for each OKR across the entire business.
For the scouts, once we'd helped everyone form up their OKRs our job was far from done. We had to then help them follow the scoring process, which needed to be done every two weeks.
Speaking of scoring, it is where you evaluate your current progress towards the OKR (by measuring the key results) and then decide whether or not you need to change your actions in order to achieve your goal. Its a critical part of the process, because without it its easy to fall into the pattern of "set goal, forget about goal, wonder why goal wasn't hit".
Having scored frequently throughout the quarter, at the end we did final scores (i.e. was the OKR accomplished), discussed them during a company town hall, did a retrospective and then, armed with new information, rebooted the process and started again.
Stay Here, I'll Be Back
There was obviously more to our process than what I've written above, but its enough to illustrate how we approached the roll-out. I went into as much detail as I could, hopefully finding that sweet spot just before your eyes glazed over.
We iterated on our OKRs for a few quarters, making small incremental improvements (i.e. slightly better tracking sheets, recruiting more scouts to help share the load, setting better OKRs by learning from our mistakes, etc) and generally starting to feel like things were falling into place.
That's not to say there weren't obvious cracks showing, but we were confident that given enough effort and patience, said fractures could be repaired.
Little did we know that our nascent OKR process would soon be faced with a challenge that would shake it to its very core...