There are often many challenges, disagreements, and arguments in social media about various aspects of Lean management that center upon data. For example, people will argue about the data on “Lean failure” — whether it is accurate, reliable, or just plain wrong. Or, something stated as fact is challenged by others who demand to see the data before they will accept the fact. Then there are opinions, grounded in the reality of one’s experience — one’s sensory data. Some people with decades of first-rate experience are challenged on their opinions as if they have little experience — no data.
These conflicts are caused by a type of cognitive bias, data bias, wherein one must see data to accept that a problem exists, and demand that someone produce the relevant data — usually knowing that is unlikely to happen. Data bias seems to be the residual effect of the big company way of thinking of the last few decades: “We make data-driven” decisions.” Allied with that is the banality: “What gets measured gets managed.” In my experience, arguments about data, the demand for data, etc., are usually either a delay tactic or a desire to protect one’s interests. Either way, it is dishonest.
One of the great things we learned from Toyota is to observe with our own eyes (and other four senses), grasp the situation, and take action. In many, perhaps most, cases, data (the numbers) are not needed. In other words, “Go see.” However, just because someone goes and sees does not mean they are seeing much of anything. Observation is a skill developed over many years — decades, in fact. So two people, one with developing observation skills and another with highly developed observation skills, will see things differently. So it is with “Lean failure” and other Lean problems.
Data bias is inconsistent with TPS way of thinking. Namely, data bias causes delays in taking action or results in maintaining the status quo. There are other problems with data bias:
- Problems can appear long before data becomes available. When a house is on fire, do you measure the temperature and height of the flames before you begin to extinguish it? No.
- Data is not gathered for myriad types of problems. Does that mean the problems do not exist? No.
- Lots of data is gathered for the wrong problems. Does it respect people to have them work on the wrong problems? No.
- Lots of data is gathered for things that are not a problem. If you believe in the dictum “What gets measured gets managed,” then you are doomed to managing metrics instead of solving problems.
You can have the facts of a situation without data. But having data does not mean you have the facts of the situation.
Computers, software systems, and databases (spreadsheets) have surely contributed to widespread data bias. But just because sensory information is not captured by databases does not mean it isn’t data, and it does not mean there is no problem. Returning to the house on fire example, Lean has been a house on fire for 20 years. It’s problems — profound misunderstandings, misapplication of methods and tools, precious few Lean transformations, absence of things like standard work, misuse of A3 reports, odd combinations and mutations, etc. — were apparent long ago. Countless observations confirm it, whether via company visits, conference presentations, bad books, weird articles, and so on. The data? It is still missing. And it will remain missing.
Lean’s long slow slide towards irrelevancy is due to inaction despite the abundance of observable evidence. In general, Lean as it is practiced today is far weaker than how it was practiced 25 years ago. It has been overrun by compromise and concession in order to accommodate classical management.
In Lean-world, online (social media) and offline (conferences), recognition, celebration, and agreement are much more highly valued than engaging the underlying problems that should be discussed. Is our goal to recognize, celebrate, and agree, or get better at understanding the problems associated with Lean transformation processes (and countermeasures)? For all the alleged scientific thinking in Lean-world, conferences are merely feel-good sessions that highlight success stories while the struggles and failures remain hidden — but everyone knows they exist. This, unfortunately, damages Lean’s credibility.
Scientific conferences on Lean should focus on the big problems and invite people to present their findings based on their in-depth analysis — much like what would occur at a physics or humanities conference where people present their research findings. Data does not always come in the form of numbers, as some would like to see. It also comes from observation, experience, experiments, analysis, and logical arguments. And these usually come more quickly than numbers and allow immediate action to be taken. Time, so important in TPS thinking, has lost its meaning in Lean.