The old approximations of reality don’t work any more

Reading Time: < 1 minute

There’s a deep issue with Scientific Management (Taylorism). And Lean and ToC. What got you here won’t get you there.
Treating  the world as made up of defined repeatable work (Taylorism, CMMI, Lean Six Sigma,  ITIL…) or made up of flowing left-to-right streams  (Lean, ToC, DevOps CD…) are approximations of reality that only have acceptable levels of error when the world is sufficiently stable over time.  It’s not now.

Lean and ToC work in a factory production line and DevOps CD works for enterprise deploy because they are highly constrained systems where linear flow approximations still work. Beware of other contexts.

You can’t measure work (or variation in work) when every transaction is different.

You can’t measure much about knowledge work when it is invisible and created by collaborating groups not individuals.

You can’t optimise flow unless it is steady state, or at least predictable,  mathematically modelled.

You can’t model flow when it exists in a complex network of autonomous co-creating agents.

These complex adaptive systems are VUCA.  You can’t know what they are doing with precision, not predict what they will do next with any certainty.