At one level, a lean transformation means improving existing systems and building new systems where required. People and organizations cannot avoid systems. We all work within systems, be they regulatory, financial, logistical, political or eco. These systems influence each other to varying degrees and at different rates. Some systems are natural, others emergent, yet others deliberately designed by humans. Lean transformations succeed when we take a long-term approach, recognizing that the systems that deliver our profits are interlinked collections of processes both within and outside of our four walls. It is a lot to think about.
One of the least appreciated sayings of Taiichi Ohno is on the topic of systems and how we should build them. He advised
Build a system that loses money if you make defects.
Even for a man who liked wordplay and was at times deliberately obtuse to force his students to find the answer for themselves, this is an odd one. Why would we build our system to lose money? This is the man who said “costs exist to be reduced” and changed the equation from “price = cost + profit” to “profit = price – cost”. Was it a roundabout way of saying “work on reducing your defects?”
Perhaps he didn’t really mean “build a system”. Maybe the idea was to install discrete mechanisms, methods or means. For example, visual, electrical or mechanical triggers to stop the line when we detect defect are a staple of lean management. When we limit the amount of work in process (WIP) and create defects, we feel the pinch of the shortage. Did Ohno just mean to say, “Make defects painful so that you will act to prevent them”? No doubt he would agree with that sentiment, but Ohno specifically used the English “system” as a borrowed foreign word, when he said “Build a system…” (悪いものを造ると損をするシステムを作れ).
If we are to take him at his word, we need to use a wider lens and view systems he wants us to build as a broad collection of processes. Systems are made up of stocks, flows and the actions we take to move value through from start to finish. Ohno advocated taking slack out of the systems to expose problems that are known but not felt. No doubt this saying was in support of his message to reduce WIP, make flow rectified and laminar, and thereby reduce wasted work associated with WIP. In today’s language, we need to identify value streams, reorganize our processes and people around them, and track our true costs including defects through value stream costing.
But why stop there? There are many critical systems in our society in which the makers of defects do not lose money. In healthcare, physician malpractice is protected by insurance. The loss is spread out across all insured parties, rather than felt by the individual doctor or hospital who makes the grievous error. Liability insurance serves a purpose, but it can also create “slack” in the system, resulting in a lack of error-prevention action in the system. A rise in insurance premiums is a “loss” that is not in proportion with a medical defect. When politicians make defective decisions, the losses are socialized and spread across current and future generations in the form of increased taxes and debt and/or eroded infrastructure and service levels. When political systems allows losses to go unaccounted for, defective decisions continue. Where do we harvest the losses of a particular defect? As a rule of thumb, the later we detect a defect the larger and more distributed is the loss.
The words “if you make defects” in Ohno’s quote suppose a production process. But what about the types of work in which bad results, rework, trial-and-error and so forth are the norm? Software bugs are an unavoidable byproduct of coding. Laboratory work in biotech create outputs that is not the desired result for the downstream process. Should we take Ohno’s words to apply only in the context of established, predictable processes in variation is never good? Or can we apply the notion of defects in research and development environments, and build systems that highlight these losses? Some view this as a rhetorical question with a clear, binary answer. Others say it is silly to even attempt to apply such a paradigm so far outside of its original context of Toyota.
Taking Ohno’s saying at face value, most of us don’t have to work very had to do what he is asking of us. We already work in systems that make defects. They do indeed erode our profits. But if we were to build such a system, what would we do differently?