As proponents of Lean Six Sigma, perhaps we should all consider how to design our own systems to respond to the pull of demand, not just teach the concept to others. All too often we see training delivered in batches, well in advance of meaningful project work, and not aligned with the skills necessary to execute that project work.
If the whole point of a Lean Six Sigma or Operational Excellence program is to raise organizational performance through realized process improvements, then that activity should march to the drumbeat of demand for project work.
Traditionally, high fixed costs have led to large training batch sizes, and the need to "fill a class" often translates to the selection of rather iffy projects in order to justify the training, or purely to achieve some certification.
You may recognize this as the tail of the distribution wagging the dog — prospective black belts or green belts in search of projects rather than important projects searching for a way to get completed. It's largely a function of trying to fill training classes to spread the fixed costs.
But isn't it inherently inconsistent to teach Lean using a big batch model?
Breaking through batch thinking requires two elements: 1) recognizing that training people who are not yet assigned to work on a value‐creating project is a false economy, and 2) reducing the high fixed costs of training so that smaller lot sizes make economic sense. That's where blended learning enters the picture.
One of the drivers behind shifting the economics of training is a variable cost model that allows on‐demand training in lot sizes of one. A second driver is the high effectiveness of blended learning: somewhere between 20% to 40% more effective than training delivered through classroom instruction (based on extensive academic studies and confirmed by MoreSteam client data from 2008 to 2015).
The first step toward leaning out the learning model is to recognize the rightful demand signal, which is the project work to be done, rather than the people looking to be trained.
Here's what we propose:
1) Prospective project work flows out of the strategic imperatives as required to close performance gaps (see previous post on this subject).
2) For each project, are there people capable of executing the work successfully?
3) If there are capable people, there is no need to train — just do the work.
4) If not, identify candidates and (5) train them in a curriculum consistent with the work that needs to be done. If it looks to be a complex project, then the training curriculum may need to be more involved, but if the project appears to be less involved, why over‐train? An adaptive on‐demand learning model should respond to a pull signal for more complex tools as needed, as opposed to batch training that guesses about the needs, and then pushes excess complexity.
The leanest learning delivers only what is needed, as needed, where needed.