Impressions from Lean Systems and Software Conference 2012 Boston

The buzzword of the conference seems to have been “Lean Startup”. It permeated into many talks (including mine) in two main aspects - One was the classic product/customer-focused Lean Startup as an alternative narrative to Lean/Agile. The other was taking the ideas of fast cycles of Validated Learning and adopting them as a narrative for the approach to change.

This came up in Jeff Anderson’s ambitious and thought-provoking or even provocative talk about the transition participation engine as well as in my attempt to “fix” continuous improvement.

Continuous Improvement Is Broken - LSSC12 on Prezi

Topics like gamification tied into this a bit, and I think we barely scratched the surface on the discussion about what forms of gamification are appropriate for engaging people in lean/agile change, if any.

I had the chance to talk to Troy Magennis about Simulations and although skeptic, I see some promise in the usage of simulations not only for forecasting and risk management but also for studying and choosing experiments to run as part of the improvement cycle. Interestingly, LeanKit Kanban announced they intend to integrate this simulation capability into their Kanban tool. I’m looking forward to seeing how that works out.

Getting back to Lean Startup, there were several discussions trying to compare Set Based Concurrent Engineering with the Lean Startup Minimum Viable Product Split Testing approach. For me Set-Based is aimed at reducing technical risk trading money for accelerated time and robustness while MVPs are aimed at reducing business risk avoiding investing money in products/technologies that are irrelevant. So one basically would try to start with a Customer Development approach using MVP. If the business model is proven, continue to implement the technology.

If an initial MVP is very expensive/risky to the point of requiring Set-based, I would try to use a non-product MVP first (The Pretotyping concept makes a good differentiation between different kinds of ways to “fake it until you make it”). If I feel good enough about the business model, but not yet sure, I would indeed use Set-based focused on my next level of MVP (but not the full first release of the product) to accelerate learning and reduce technology risks on the way to the MVP. Then launch my MVP and fine-tune it using the Build-Measure-Learn cycle. If indeed I find Product-Market fit AND there are major technological aspects still missing I might use Set-based again.

To sum up, they can probably be used together in different phases of the product/business lifecycle. But it is the MVP that should drive the need for Set-based. In general, validating the business need BEFORE focusing on various technological choices is a point that cannot be  emphasized enough based on what I heard in sessions and offline discussions.

I finally had a chance to meet Claudio (@AgileSensei) and see his A3 talk, which was magnificent as expected. I really need to integrate more of this thinking into my work. I’m especially going to focus on the usage of A3 to accompany management workshops I use for initial thinking and studying how to approach a change initiative for an organization.

Don Reinertsen hit a home run again with a fabulous session about Decentralized Control, the last chapter in his Principles of Product Development Flow book. He presented concepts like the reality of uncertainty and how to deal with the need for both decentralization as well as alignment using Doctrine, Commander’s Intent as well as great examples such as Firefighting being more disciplined and effectively executed than IT work.

This was also a highly interactive session with Don engaging the crowd, including Andrew Fuqua who happens to be an actual certified burn master commenting on the immense complexity rising from the fact that fires affect weather which in turn affects fires.

This chapter in the book is one of my favorites as I find it a great mature way to explain the need and value of some self-organization. It comes at the end of a dense book so I think many people miss it which is too bad. I had a chat with Don about some ways to make this specific content more accessible to the masses…

The Brickell Key Award for 2012 was awarded this week. After a nerve-wrecking evening, the winners were announced - my friends Arne Roock and Jim Benson. Congratulations for a well deserved celebration of their contribution to the international Lean/Kanban community.

Another announcement this week was reorganizing the Lean Software and Systems Consortium into  the Lean Systems Society, a change that strives to emphasize even more the openness and nurturing of new ideas and approaches instead of freezing current models. I’m proud to be one of the founding fellows in the society. 

Beyond these highlights there were many conversations and interesting sessions, and I head home energized full of ideas and thoughts looking forward to share them with the AgileSparks team and my clients. It is as usual great to meet many old friends and make new ones and hear many appreciations for work I’ve been doing. Cannot wait for LSSC13 in Chicago (29/April-1/May 2013, the new JW Marriott)!

Oh, and the Holy Land Kanban book sold quite well at the book corner… Inviting readers to write reviews/comments - keep it honest!

Improving flow means fewer bottlenecks, faster feedback, and less wasted capacity. Explore flow advisory or let’s talk.