Recently I got back from 4 intense days at LSSC 2012 in Boston, which I have to say is the best conference I’ve ever been to for learning opportunities. It was exceptionally well run, with activities planned beforehand (Lean Camp and Action Kitchen), the best keynotes I’ve been to, excellent sessions (significantly better than last year). My thanks go to David Anderson and the rest of the organisers and volunteers. The venue was great, with very friendly staff, good food, rock solid wifi and a good amount of space. The addition of lightning and ignite talks in that post-lunch lull worked really well, summarising months of thought into minutes. The only improvement opportunity I can see for next year is to extend the Brickell Key Award Dinner to allow for more dancing opportunities.
The conference raised a number of themes for me:
- We should accept and embrace uncertainty in what we do
- We should be validating our learning, by forming hypotheses, acting and evaluating the outcomes.
- We should seek to understand other domains before we try to apply their processes to our own
- We are learning how to learn and how to do software development, not just what we are developing
- We are human, collaboration, reciprocation and motivation are intrinsic in most us, we should build systems accordingly
I’m going to follow a similar format for my post as Liz Keogh, who has already written up her thoughts on the conference, here.
Steven Spear talked about uncertainty, learning and the key to great organisations. By not having predictions, we rob ourselves of the opportunity to be surprised and to learn. He urged us to accept that we don’t know, yet be disciplined in making a prediction to allow us to correct the ignorance we bring to the situation. We miss many opportunities to learn by not admitting we don’t know and by band-aiding issues, this must start with the leadership of an organisation. The key to greatness is not in knowing the difference between the best and average organisations, but in how they transition over time. It is the pace of change and learning, in an ever increasingly complex world, which differentiates these organisations.
David Joyce asserted that we are still using the tools of over a century ago, which were built to solve different problems. The rate of change (and therefore of necessary learning) has increased significantly in that time, from every 3 years to every 3 months. His kanban board demonstrating how we should be validating learning was an excellent example of how much further we can take visualising the work and our value stream than the basic agile board.
Chris Shinkle put together a beautiful and touching session, in which I nearly cried, talking about his kids baseball team. Showing how important it is to put the emphasis on learning, to have fun and to change things when it’s not working.
Nigel Dalton showed us a glimpse of Lonely Planet, showing me for the first time how strikingly different visualisation of work can be outside of software development and gave me a few ideas of how to improve our boards.
Michael Kennedy talked about set based product development at Toyota and other organisations, emphasising the need to generate sufficient knowledge before designing. I didn’t manage to see the whole session, however I will most definitely be watching the video at least a couple of times. Whilst it’s still sinking in, I’m thinking about the implications of this talk on building capability, looking after and feeding forward knowledge and how this can relate to time boxed vs. flow based software development.
Gregory Howell continued the themes of uncertainty and us taking inspiration and ideas from domains we don’t understand. The assumption when we took ideas from construction, was that it was a complicated yet well understood domain with processes to match. Gregory explained that the processes were largely built around contract management and didn’t really work, 85% of construction managers overestimated the certainty in their projects (“they had the optimism gene turned fully on”). He described projects as a network of commitments and as a conversation of ends and means, both useful metaphors for how we work. In construction the key is knowing when people need to turn up and that the site is ready, therefore build buffers into the ongoing schedule and it is far more important to work on the interfaces between people rather than each job in order to go faster.
Bob Charette gave me a interesting perspective on risk, expanding upon going up the down escalator metaphor from Waltzing with Bears, the two key ideas being, “If there was no change, we wouldn’t need managers” and that profit is the result of exchanges of risk and opportunity.
Jeff Patton gave a genuinely hilarious session (following a similar toned one from Benjamin Mitchell based upon the laughs coming from the room) about the myths and misconceptions about Kanban, including part of an episode of South Park. Make sure you watch the video when it is released, it summarises much of what we need to improve our teams now. I learnt value is hypothetical, identify opportunities and then measure where is came from at the end and that Process != skill, we should not forget that no process will save an under-skilled team.
Mike Burrows talked about the effects on risk of using Kanban. Time makes a nonsense of prioritisation, we’ve all had critical bug sit around for years and variability means that sequencing is important, therefore we should make the variety explicit.
Yochai Benkler gave the densest keynote I’ve ever experienced, all whilst we were suffering hangovers from the Brickell Key Awards Dinner the night before. The twitter stream drastically slowed, showing how hard it was to summarise this deep a set of content. I’m finding it difficult to refine it here, watch the video, it will significantly effect your view on motivation and social systems. Recent research has shown that conscientiousness and collaboration are inherent in the population, we inherit it in our genes. Our traditional systems have been based upon the rational actor model which doesn’t deal well with uncertainty and imcomplete information and that people will do the least to get the most, which for most of the population is untrue. Therefore we need to move from systems based upon closed control to open intrinsic motivational systems. Different forms of reciprocity allow us to understand our behaviour in social situations, e.g. giving blood. And my favourite quote from the conference: “Roles allow us to have a different morality to usual”.
Don Reinertsen covered the idea of centralised vs. decentralised control, arguing it is always a combination, that the idea that a system has to be one or the other is the “tyranny of the exclusive or” and digital thinking. Uncertainty does not demand decentralisation, for example, chess. He used forest fire fighting and the military to demonstrate effective combinations of centralised and decentralised command, based upon the importance of doctrine which leads to mindful individuals and teams, who use principles and an understanding of intent, in combination with situational awareness to make good decisions. I will definitely be doing more research into this area, in order to improve my understanding of building a skilled capability in order to deliver when needed, a subject close to my heart.
Jim Benson explored the world of cognitive biases and the danger of us believing that we know reality. The Kanban board is merely physical manifestation of the gemba and a decaying desperate grasp upon reality, not reality itself. Our goal is to share our incorrect world view, in order to elucidate our assumptions. Significant to me, is the idea that cognitive dissonance causes suggestibility, which might explain why I am all to often too easily influenced. He also put to bed, in the best way I’ve seen, the myth that home/work balance is possible, they are “two system conjoined by us”, a simply beautiful expression of the futility of trying to separate them.
Following this conference, my head is brimming full of ideas, around validated learning, how we approach training/early learning for new teams, leadership and building capability, expect to see posts on these subjects soon.
It was a pleasure to see people I don’t see often enough, Simon & Saffron Bennett, Jabe Bloom, Eric Willeke, Liz Keogh, Derek Wade, Karl Scotland and David Joyce, in particular and to make many new friends, too many to mention here, including Jim Benson, Tonianne DeMaria Barry and Gerry Kirk on my last morning in Boston.
This is only a very brief summary of my learning from this conference, I haven’t even mentioned Lean Coffee or Lean Camp. From what I’ve heard, the sessions I didn’t attend were also very good, I’ll defintely be watching the videos. If I haven’t convinced you yet, make sure you watch the videos when they are released, otherwise I hope to see you in Chicago next year as I want the opportunities for learning that this conference gives to continue and for as many as possible to experience them. As usual, please feel free to comment, feedback is always welcome.