Sometimes we find large organizations or entire nations where the people at the top appear to pretty much lost contact with reality, and certainly appear to have lost contact with the common people at the "bottom" of the pecking order.
There are a few spectacular cases where the top dog is simply criminal, and these tend to produce a stereotype that all Chief Executive Officers (CEO's) are crooks. That's not true, but as a story it greatly simplifies the peasant's world model and puts blame in someone else's yard.
There are also a few spectacular cases where someone on the bottom of the food-chain is criminal, some "welfare queen" or person clearly ripping off the system, and these also tend to produce headlines and a stereotype that all people on welfare are lazy bums and crooks. Again, this is not true, but it greatly simplifies the CEO's world model and puts blame in someone else's yard.
So, we have top management stereotyping, dehumanizing, even demonizing labor, and telling each other the story that production and business problems are all being caused by labor, despite our own heroic efforts to try to fix the problems they cause. They (labor) are primarily a cost and burden, but, of course, we, upper management, are the indisipensible saviors of the day. We just don't get respect for our 16 hour days and hard work, and are treated like dirt.
And, of course, we have labor stereotyping, dehumanizing, even demonizing management, and telling each other the story that production and business problems are all being caused by upper management, despite our own heroic efforts to keep the ship afloat. They (management) are primarily purely parasites, living off our sweat and labor, adding nothing and sucking out the resources we need to simply our jobs so they can live in palaces while we face foreclosure. We just don't get any respect for our 16 hour days and hard work, and are treated like dirt.
It should be clear that the above scenario is not going to produce very good working relationships between labor and management. From management's point of view, this is a problem because, darn it, output of the company still depends on labor rowing hard and doing a good job, which they don't seem to care about or be able to do on a day to day basis. From labor's point of view this is a problem because, darn it, the survival of the company still depends on management doing a good job steering, which they don't seem to care about or be able to do on a year to year basis.
So, we end up with companies that spend 49% of management's time fighting against 49% of labors efforts, leaving about 2% of the effort to actually accomplish the work to be done, which is almost impossible because of all the dust in the air and shrapnel flying around from the fight going on.
Now, there are additional problems in that the management of the company is probably based on obsolete and misunderstood competitive "Darwinian" models of thinking, where it is believed that the best way to increase company operation is to have all the manager's busy fighting each other for survival, and trying to put each other down all day. While Machiavelli may have been right that this strategy works for a King to have the contenders for the throne neutralize each other so they don't gang up on the King, it's a poor guide for models of countries or companies that depend on the effective working-together of middle management for success.
And it's a very poor guide for companies where the output depends on the people at the bottom feeling supported and safe to create workplace improvements without risking their jobs by doing so.
And, similarly, many companies believe that the pay of workers should be competitive with each other because, in some invisible way involving Adam Smith's hand, this will result in maximizing output. So, workers are motivated to look good and make each other look bad so that they can fight for the left over and diminishing table scraps and maybe "succeed" at getting theirs while others don't, or are laid off, etc. It's not really clear why anyone ever believed that was a very successful model, and it's not the model families use at home to decide which children to feed and which ones to neglect or kill off.
Aside from my babbling, however, the larger point is that companies like Toyota have abandoned the competitive model internally, with spectacularly successful results. So, maybe the old model worked in 1900 for whatever companies needed to do then, but the old model is broken in 2000 for what companies need to do to survive today.
Today, a good fraction of the wisdom and view of the world that matters is done at the bottom of the pyramid, not at the top. The knowledge workers and professionals have specialized training and education and skills and insights that are indispensable for making sense of the world and adapting successfully to the new world. The "cheese" has moved.
But even in companies that attempt to operate on the basis of "Theory Y" not "Theory X"
(as described by McGreggor), and even in a situation where most of the people are trying to do a competent and honest job most of the time, and where everyone knows the survival of the whole company is at stake, it still doesn't work. It doesn't click into place. It doesn't become vital and revitalized and refreshed and alive and productive and adaptive and agile.
Why is that?
Note by the the way that this discussion has nothing to do with outside forces that "cause" the company to be "unable to succeed." This isn't about "unfair currency values" or "unfair trade practices" or "high costs of health care" or "the burden of government regulations" or any of that. This is about what goes on inside the skin of the company or country that works or doesn't. These are things that we have 100% control over, if we really wanted to change.
But when we analyze why things breakdown and don't work, we can get distracted by specific instances, or by apparent problems over there that are really downstream of a larger root-cause problem somewhere else. The beast seems slippery and hard to get a handle on what to address, where, and how.
So, we can look to basic principles and ask if we have the basics right, before worrying about advanced skills. I'll assert that, so far as I can see, this is where most of the breakdowns occur -- on things so basic that no one's attention goes there. What's breaking is things so "little" that they seem "negligible", except that we know from feedback loop theory that, after you apply the power of selective compound interest, what is "little" and what is "big" can completely swap places.
But, there is one basic principle that has to "be right" or at least be functioning at all for any adaptive organism to do just that- adapt to reality around itself. That principle is that the fundamental cybernetic loop has to be functioning.
There has to be a goal-driven feedback control loop in place that continously adjusts efforts based on another continous stream of sensory data regarding the outside world and how closely we fit to it now. This is basic common sense. You can't leave out any of these parts and have a system that adapts to the external world. You have to be able to see the external world. You have to be able to see where your own body is. You have to be able to sort out how close you are to the external world, and which way you're offset now. You have to be able to push on some muscles to get a combination of your own body and the world's responsive echo to shift a little. And you have to learn from experience what actually works today, not what used to work 30 years ago.
So, figure 1 shows the basic loop in a pyramid shape, to capture a little of the fact that there are more people at the bottom of an organization than at "the top" -- although even there we've biased the picture by not calling the parts the "outside of the circle" and the "center of the circle" and implied that there is a vertical dimension of status, privilege, and power that has to be there, which is a whole different question.
Now, we can learn a huge amount by letting go of the idea that management, or labor, or "the company" that is the primary actor or agent of change, which focuses our attention on "who" should be "leading" change. (Then we also have parasitic consulting companies who come in and assert that only THEY are qualified to lead change, please pay up front, no guarantees.)
In a complex adaptive system matrix or substrate or medium, the agency that does anything is not the parts, it is the whole. The pilots do not fly the plane - the cockpit team flies the plane, which is why 74% of commercial aircraft "accidents" occur the first day a new set of highly competent individuals is assembled and starts team-formation while actually flying a real plane, not a simulator. (hmm, that suggests a solution.)
But, saying "the whole" or "the team" or "all of us need to work together" is fluffy and not helpful.
And, it's not entirely accurate. The "whole" of us could be in a big heap, not an organized lean, mean, fighting machine. There are lots of ways to have "wholes."
What actually is needed is not even directly visible in physical space, and not something you can hit with a hammer, although you can indirectly measure it.
What actually is needed is a feedback control loop, what I've called an S-Loop. And it doesn't really matter to the S-Loop which people or computers or liquids or waves it flows through, so long as there is a closed, self-aware, self-restoring, goal-seeking loop that can be traversed hundreds of times in a row.
This is like the picture I had a few days ago of a person filling a glass of water. During the "filling process" the "fill-the-glass" control loop is steering, not the hand, not the eye, not the water. It's as real, and as right and as wrong to say that "the hand controls the faucet" as "the water level controls the hand". Once you make this abstract loop, it comes alive and takes over and the parts that make it up, or used to make it up, are really now, like a concert pianist, present but almost simply a channel through with something is flowing.
It is very possible to have companies without a vertical control loop, although they will be slowly or rapidly self-extinguishing and non-adaptive. General Motors comes to mind as one where management hasn't listened to labor for decades.
But, I'll assert that it is not possible to have an adaptive, agile, successful company without this feedback S-loop in place. There are no parts that can be left out, almost by the definition of "adaptive." If the world changes, the company has to detect the change, try to figure out which neuron or muscles is attached to what, push or pull on them, and change until it has adjusted to match the new reality. That requires a closed, convergent, goal-seeking feedback loop if the outside world keeps shifting somewhat unpredictably - and all complex adaptive systems do.
Our basic problem in constructing an adaptive assembly of people, then, comes down to putting a structure in place that will generate and then identify, value, and protect its own internal S-Loop. If that piece isn't in place, everyone might as well just go home.
We have two difficulties with this task, aside from even realizing it's a task. First, we have legacy mental models of "management" that work to break the "top" apart from the "bottom" of our assemblies. That snaps the loop apart into two dysfunctional pieces, and is a losing strategy.
Second, even if management and labor are on the same wavelength and try to make this work, they have a problem of distortion caused by distance. If an organization has 14 layers of mid-level managers that have to "process" and "add value" to the perceptual stream going uphill, by the time it gets to "the top" it bears no resemblance to "ground truth". And, by the time a CEO's clear and direct orders and intentions have been disassembled and "interpreted" by 14 levels of middle management going down the chain, they make no sense at all when reaching the bottom.
That set of discrepancies then feeds into and sustains the perception of each "side" that the other side must surely be composed of incompetents and morons and criminals.
Here is just a tiny bit of math to see how powerfully this effect compounds, and how even, say, a 5% distortion at each level adds up.
With even a single pass of the feedback loop, with 14 levels going up and 14 going down, and each level applying a 5% distortion and noise, the result could be how much?
The equation would be distortion = (1 + 5%) to the 28'th power, minus 1, if they were all coherently distorted in the same direction. That number works out to be about 300%.
First of all, that 14 levels thing has to go. Cut it down to 5 levels of management, as Toyota did, and the numbers get better, but not good enough.
In fact, you have to get distortion at each level down to 0.5% or less, (99.5% accuracy) before the total loop distortion, per pass, is down to 5%. You can see why Toyota is obsessed with accuracy.
But this is accuracy in the vertical dimension, accuracy of the control feedback loop that should be running the company, not management, not labor, not some person or committee.
Travel time up and down the loop matters too. Like "takt time", the cycle time for the vertical loop better be down as far as possible, preferably down to hours, not months, and if possible down to minutes.
How is THAT possible? first, delegate everything possible to as near the front lines as possible, so that many decisions either can be decided at the front, or near the front. Same way your body has override loops that let your hand pull away from a hot stove even though your brain hasn't even gotten the news that you're being burned yet. Where you can, make everything decidable and steerable at the bottom 2-3 levels. Like the coast guard, let each ship's captain have discretion about what actions to take without having to check in with mom first.
But, to really fly, you want the distortion even lower, to simplify the loops task of trying to figure out how distorted the round-trip signal is and adjusting for that. You can tell when you hear your own words come back twisted about how far bent things are.
But the key to making this work is the virtues previously sustained by religion - honesty, sincerity, transparency, surfacing problems instead of distorting the 'bad news" by hiding them, etc. That will really cut down distortion a lot.
Other factors that are pretty much required to cut distortion involve letting to of the overpowering stereotypes that groups have of each other that prevent them from hearing what the other group is actually saying. That's where I started this post and where I'll end it.
Prejudice and hatred and anger cause dramatic distortions in ability to see straight and to think straight and to hear what someone else is saying. Sometimes humility is required and willingness to let go of a cherished misconception that keeps trying to distort the facts to fit the model, instead of vice-versa.
This is the most important loop that everyone be "kept in". This is the loop that "being out of" is more or less a death sentence, at least metaphorically.
But, so far, it hasn't been identified explicitly as a loop, or the reason it is important made so obvious. That's what I was trying to do today.
There can be many other problems a company can have, but those problems could be solved if the core S-loop were fixed first. So, in terms of priority, figuring out where this loop has taken up residence and shortening it and making each step cleaner and care and feeding of this loop is as crucial as care and feeding and QA on the assembly line loops.
It will require that people at "different" levels actually psychologically "touch" each other and accept each other's role in the overall picture. That involves letting go of the "we" and "them" mentality and seeing the "us" in everything.
I left out the picture but have to go now. I'll finish this later. I think it's a very simple, very clear, very helpful model of what to look for and diagnose and troubleshoot first in a company that is not healthy or is failing to thrive or perform up to expectations.
This is also 100% consistent, so far as I can tell, with the U.S. Army Leadership Field Manual (FM22-100). One major goal of that whole doctrine is to establish the vertical loop that lets fragments of news flow up bill and be synthesized and slightly adjusted in that light new orders come downhill in response. When the S-Loop functions and takes over, it doesn't eliminate the hierarchy, but it floats above it and channels through it and makes it, in some senses, almost irrelevant. Like the concert pianist, you can almost just watch the music flow through you and be amazed because it's like it's doing itself now.
That's the end goal, and it's achievable -- it's just a fully functioning, fully unclogged and unblocked and undistorted S-loop, on a very large scale. Until that works, nothing else
matters. Once that works, all the other problems can be solved.
That's the first thing to address if results are not up to expectations. If someone knows a way to achieve externally responsive adaptation without that loop, I'm all ears. I simply can't imagine how it could be done. Adaptation IS a feedback loop. You can't leave the loop part out.
What we haven't done is realized that and assigned some people to Quality Assurance to make sure the loop didn't fall apart when no one was looking.
w.
No comments:
Post a Comment