Shipulski On Design

When the output cannot be predicted, that’s uncertainty.  And if there’s one thing to be certain of it’s uncertainty is always part of the equation.

With uncertainty, the generally accepted practice is minimization, and the method of choice is to control inputs.  The best example is a high volume manufacturing process where inputs are controlled to reduce variation of the output (or reduce the uncertainty around goodness).  Six Sigma tightens the screws on suppliers, materials, process steps, assembly tools and measurement gear so the first car off the production line is the same as the last one.  That way, customers are certain to get what they’re promised.  Minimization of uncertainty makes a lot of sense in the manufacturing analogy.

But there’s no free lunch with uncertainty, and the price of all this control is inflexibility.  The manufacturing process can do only what it’s designed to do – to make what it was designed to make – and no more.  And it can provide certainty of output only over a finite input range.  Within the appropriate range of inputs there is certainty, but outside that range there is uncertainty.  Even in the most well defined, highly controlled processes where great expense is taken to reduce uncertainty, there is uncertainty.  Even the best automotive assembly lines can be disrupted by things like tsunamis, earthquakes, epidemics and labor strikes (100% certainty doesn’t exist).  But still, in the manufacturing context minimization of uncertainty is a sound strategy.

When the intent of a process is to do things that have never been done and to bring new things to life, minimization of uncertainty is directionally incorrect.  Said a different way, creativity and innovation demand uncertainty.  More clearly – if there’s no uncertainty in the trenches, there’s no innovation.

The manufacturing analogy has been pushed too far from the factory.  Just as Six Sigma has eliminated variation (and uncertainty) from things it shouldn’t (creativity work), lean and its two uncertainty killers (best practices and standard work) have been jammed into the gears of innovation and gummed up the works.

Standard work and best practices were invented to reduce variation in how work is done with the objective of, you guessed it, reducing uncertainty. The idea is to continuously improve and converge on the right recipe (sequence of operations or process steps) so the work is done the same way every-day-all-day.  By definition, innovation work (the process steps) is never done the same way twice.  The rule with best practices is simple – it should be reused every time there’s a need for that exact process.  That makes sense.  But it makes no sense to use a best practice when a process is done for the first time.

[Okay, the purists say that all transactional elements of innovation should follow standard work, and theoretically that’s right.  But practically, the backwash of standard work, even when applied to transactional work, infects the psyche of the innovator and reduces uncertainty where uncertainty should be bolstered.]

Uncertainty is an important part of innovation, but it should not be maximized (it’s as inappropriate as minimizing).  And there is no best practice for calculating the right amount.  To strike a good balance, hold onto the fact that uncertainty and flexibility are a matched pair, and when doing something for the first time flexibility is a friend.  And when standard work and best practices are imposed in the name of innovation efficiency, remember it’s far more important to have innovation effectiveness.

Image credit – NMK Photography

Comments are closed.