Paul “Bear” Bryant
Recently we moved from an annual planning process to a quarterly large scale release planning event. There was a lot going for the idea of the new quarterly planning process:
1. Smaller batches of projects to plan 2. Less wasted time on projects that would never get worked on 3. More collaboration between teams and stakeholders, because we held the planning event in a big room with everyone invited 4. It only lasted 2 days, so the amount of preparation was dramatically reduced
What’s not to love? Our previous annual planning process started in February or March and continued non-stop until late September or even further (once it went through Christmas). The amount of waste in the process was truly astonishing. There were multiple reviews by executives, red-lining exercises, requests for more estimates. It was interminable.
So anything that offered us a way out of that madness was OK with me. We modeled the planning event on the SAFe style large scale planning events. It seemed like a slam dunk. Get everyone in the room, facilitate like hell for two days, and out pops a bouncing baby quarterly plan. Easy.
We held the event and it went pretty well. There were a lot of missing requirements and people had to get used to participating in a high intensity collaborative event, but in the end, we got the job done and walked out of the room after day 2 with a relatively stable set of commitments for the upcoming quarter. So we did a retrospective and started planning for the next event in 3 months. Just like it says on the bottle: re-apply, rinse and repeat.
And this is where things started to go wrong. Some folks were concerned that they hadn’t been prepared enough for the first event, and they decided that they should begin their planning for the next event earlier. While this is not a bad idea on the face of it, there were some consequences that we didn’t see coming. Soon, everyone was screaming that they needed to start planning – in as much detail as possible. All of this happened within about 2 weeks of finishing the first planning event. So the end result was there were a lot more meetings and planning done before we had our second large scale planning event.
So we do the second event, and yes, things were improved. There were more requirements, the teams came to consensus faster, and in general things looked pretty good. However, there was one rather alarming development: executives added a bunch of projects to the backlog the day before the planning event. They were “must do” projects, but they didn’t have any requirements behind them. So we sucked it up, put on our big boy pants, and dealt with it. By the way, I’m not recommending that as a coping strategy.
So of course, product management and others were even more panicked this time to get started on requirements early. So now there was literally no pause in the planning between events. After we finished our second planning event, then we immediately began work for our next planning event.
Did you see what we did there? That’s right, we started off with a brief but effective planning event that replaced our year round nightmare. Then for the next event we added more preparation. And for the next event we added even more…until our quarterly planning process has become virtually indistinguishable from our disastrous old year round planning process. We’ve come full circle!
So what went wrong? I have my suspicions, and I’d like to point out that’s all they are, so take them for what they are worth. I think that in order for changes like this to be effective, there needs to be corresponding changes in the expectations for what you get out of the event. If you change the planning process, but the expectations for detail, control, and commitment/accountability don’t change, then you are going to run into problems. In other words, if the process changes, but the outputs and the expectations around them don’t change, then it’s very likely that sooner or later you will end up right back where you started.
I think you see this most commonly in organizations that implement a change (adopting agile, kanban, or anything else) and the top level management aren’t bought into the change. They are going to be looking for the same outputs that they’ve always had. They aren’t going to accept anything else. That puts pressure on the system to return to the status quo. Eventually, the change may still be there in name (as in our case), but effectively you have returned to the original system.