Earlier this month I attended David Anderson’s Kanban Leadership Workshop in London. It is intended to help experienced Kanban or lean practitioners lead change initiatives in their own or client’s organizations.
We started off playing Version 3 of the getKanban game which was cool. I had played version 2 before, but did not get totally involved as I was in charge of the graphs, but this time I was fully involved. There was so much content it is hard to decide what to mention. The two biggest things for me were the idea that it doesn’t do any good for the process geek to present the perfect idea and expect everyone to simply accept it and change, as people get emotionally invested in their roles and what they are good at, and don’t want to leave that behind. This is obvious but something I need to be reminded of, as I can get impatient when compromising with what I perceive to be suboptimal ideas. The other main point for me was the discussion about high trust and low trust culture. Kanban is supposed to help a high trust culture emerge, but I think it is more of a symbiotic thing. I think there needs to be some commitment and movement towards a high trust culture before Kanban can be fully utilized. It seems to me like there might be a point where Kanban could be used to reinforce a low-trust culture, if that is the dominant culture present, and those involved perceive a need for e.g. more rules, formality and control.
The discussion on Kaizen vs Kaikaku was enlightening. I did not know about Kaikaku before this workshop, and I had been feeling a need for something like that. The signals that our Kanban system were sending me seemed to be calling out for something stronger than minor kaizen improvements. A lot of Kanban is described as a fairly mechanical process, and the book provides a lot of practical information, backed up with a lot of science from queuing theory and lean manufacturing etc, but I like to think a bit beyond that, and look at it as a system of unpredictable, emotional human beings, so I was delighted to see a lot of focus on the cultural, psychological and sociological elements of Kanban at this workshop. For me that is where the the leadership comes in, realizing that cards on a wall and some pretty graphs are just the easy bits at the beginning. We covered so many anecdotes, other books and I guess concepts, that David finds interesting and relevant for leaders in lean software development. It went way beyond Kanban itself.
There was a lot on where to start, and avoiding calling it a "Kanban Initiative". One extremely important thing that I have to do at work is to change the way we calculate our predictions based on measured data. It turns out cycle times are not normally distributed, so using standard deviations are not appropriate for working out SLAs etc. Instead I intend to look into something called Shewhart’s Method. I have had a bit of a google already, but I think I will need to ask the community for information about it.
There was so much more, I can’t cover it all.