Tuesday's idiotic post: Knowledge as capital

And now, in the spirit (as noted before) of "See, I did so write something on Tuesday!" ... here is what I wrote then.  (sigh)
__________

Knowledge as capital, and why it matters

X's website lists several key requirements for turning rigid organizations into Dual Organizations, but he misses one.  As you set up agile teams, you have to ensure that each team has the right number of deep subject matter experts.  I'll explain why, but it'll take me a minute.  Please bear with me.

Years ago I worked in exactly the kind of organization X describes – not because we made it on purpose, but because we were a small, growing company at that point in our life-cycle.  We never hit the big time; but we were all engaged, we made innovative products, and it was a lot of fun.  X is right about all that.

We also made some big mistakes.  Being a Dual Organization avant la lettre didn't make us smarter than everybody else.  So some projects were successes while others were train wrecks.

We learned from our mistakes.  That's what you hope for.  We all want to work for a "learning organization".  Only ... when an organization "learns", it looks a lot like adding process and overhead.  Some checklists here, some prestudies there – all for good reasons, all to avoid last year's cosmic screw-ups – and soon you're buried under a mountain of process.

Can you avoid being buried?  Sure, if your people are experts.  An expert won't need most of those checklists, because he knows at a glance to do this and not that.  When he does fill out a checklist, it goes faster because he knows instantly that this section has to be green and that section is non-applicable; so he can focus on what matters and skip the rest.

But where does he get his expertise?  From experience ... which often means years of trial and error.  Do it wrong, it blows up; do it right, it works.  Repeat this a thousand times and you're an expert.

In other words, doing things right always requires time.  You can push that time onto the whole organization, by making your process idiot-proof: it takes longer to get through a project, but anybody can do it.  Or you can throw away the detailed process and trust your experts: but then you have to cultivate them over years till they have the judgement you need. 

Either way, working successfully relies on an irreducible quantity of knowledge as a kind of capital that underwrites the venture.  And the only coin to buy this capital is time.

This doesn't mean Dual Organizations are impossible.  X says they can be built by design, and I believe him.  If you graph the trade-off of process-vs-expertise against total time, there has to be a local minimum that involves some of each.  My only point is to recognize this.  Every agile team has to have the right number of deep experts whose judgement they can trust implicitly.  Not too many – too many cooks can spoil any broth.  But not too few, or you risk driving into a lamppost.