This morning I read with amusement the headline in the Wall Street Journal that BMW was recalling 345,000 cars including 5,800 Rolls-Royce Phantoms [BMW now owns RR, its no longer a classic British company]. The problem is potential issues in the power brake system — although I did not read the article for the specifics of the issue. Don’t really want to. But reading the headline triggered recollections of the endless problems Boeing has had with the 777, the parade of other car companies, computer companies and other organizations with production problems rooted in the segmentation of the production process into loosely connected and potentially unrelated organizations.
I wonder if we are seeing the consequences of organizational ‘divide and conquer’ where the tasks are divided up, communication across boundaries is via specifications and segments are outsourced in an attempt to increase profits? I see this model at play in many organizational problems (and governments) coupled with the other widely applied model called ‘economies of scale’.
In ‘divide’ the task (produce a car with certain characteristics, for example) is divided into a series of sub-tasks — design, procure, assemble, sell and so forth. Building a real machine takes time, so often the work is modeled on computers. Specifications for components are worked up and turned over to purchasing — who then source the components based on the specifications. And so on. The procured items are audited with statistical quality assurance methods to ensure the process is being followed to meet the specifications. The opportunity for problems to arise are manifold — the computer models may be less than perfect, the procurement specifications could have degrees of freedom that allow items to meet the specification and yet be unsuitable for the purpose [Boeing, eh?], the assembled components could interact in unanticipated ways, etc.
The net result is that the product or process falls short of its intended suitability of purpose — but management will still look concerned while rewarding themselves for a job well done. And the consuming public suffers. At the root I suspect is the loss of control over the production process and the disassociation of the workers at every level from what they produce. By applying the ‘divide’ strategy the organization has in a sense submitted itself to a ‘death by a thousand cuts’ losing control of pieces of itself until finally it is nothing more than a purchasing consortium unrelated to what made it a success in the first place.
Closely allied with the ‘divide’ strategy is the ‘economies of scale’ approach. This idea comes from industrial production processes where making a large batch of something ended up reducing costs allowing a cheaper product to be produced. But like ‘divide’ it is widely misunderstood and even more widely misapplied. The problem is that for making a larger batch of steel or cement one needs larger production equipment or process changes to allow continuous production. So the ‘suits’ look at this and think — if we amalgamate local governments or hospitals we can achieve lower costs by economies of scale, ignoring why the concept worked in industry. And so large scale changes are made that in the end result in poorer service and higher costs — a loose, loose change for everyone concerned. But since they rarely measure the results (and even more rarely back off when an intended result does not materialize) this fatally flawed strategy is pursued to the point of collapse.
What has been forgotten is that organizations are not industrial processes. Human groups do not scale like batches of steel for a simple reason — communication for information and control. The larger an organization grows the more potential links there are between its components and the larger percentage of its energy is consumed in the communications process. Hence the software industries’ rule that adding resources to a late project tends to make it later.
I suspect this all comes down to a reiteration of Schumachers ‘Small is Beautiful’ idea. As businesses and governments grow they become less capable of accomplishing their objectives until they reach the point of paralysis and ultimate collapse. I think this was called growing beyond their level of incompetence. The world is filled with examples at every level. There is probably a natural scale to organizations where the gap between the decision makers and the product is small enough that some sense of quality can be maintained. We have a small cheese plant up the road — they make wonderful cheese. But I suspect if they were absorbed by some huge corporation over time their product would become as tasteless as the stuff sitting on the grocers shelf (which is much cheaper, by the way).
There is another lesson to be learned as well — try to understand WHY an idea was successful in its original area. And don’t blindly assume that taking an idea and applying it without the supporting conditions is going to produce a good result. Or at least measure the results and have the courage to back off if its not working.