Nassim Taleb has written a nice article on Edge that builds up a theory of what my gut feeling is. Decentralised organisations, in my opinion, work better because they rely on the brains of all the people involved instead of on the brains of a few managers, with the rest executing the work. Orchestrating the work of many is harder than just chopping up the plans of one individual to delegate it to several others. This is a sort of complexity that I touched upon in my previous blog post in the context of irregular work hours.

Taleb’s enters the discussion from a different direction. He compares “tinkering” with strategy based work, in a very theoretical way. Tinkering allows you to try out a lot of things quickly and, if you do it correctly, to maximise your best case result. You can make notes on what works, keeping them as “options” for future work. By trying many approaches and working short-term, you stay flexible and you can opt out of bad decisions early on (and put more effort into good ones). In this way, Taleb shows that rigid planning and strategy don’t really pay off. Interestingly, the contradictory approach of working short-term allows you to get to your long term goals faster and more effectively. It could explain why a rigid planning system in the Soviet Union wasn’t competitive to the decentralised market principles of the Western world.

I like Taleb’s way of explaining things. It starts out feeling like a dry theoretical article. In fact, he works from the world and introduces a number of scientific techniques to assist him.

Edit: I just found this article on FastCompany that indicates the same thing. You can’t be sure of what is going to happen in your industry. So, stop that five year plan already and start some serious experimenting. More ideas and experiments mean more chance to hit something useful. That’s why companies like Google give free time to employees to work on their own projects and the chance to continue when a project looks promising. It’s how gmail got started.