Overcoming the Hierarchical/Functional Model with 10 Transformational Steps
In Chapter 1, I tried to explain why we need a Theory: without it, we would not know how to evaluate the outcome of what we do; we would not have any framework within which to gauge our actions, any backcloth against which to assess our decisions, or any guideline to drive our efforts toward any chosen goal.
A Theory is then needed when the radical changes brought about by digitization are foreshadowing a new organizational paradigm. It is one we can no longer afford to forego.
Digital technology is accelerating the ever-present quest that sustainably successful organizations must pursue: Quality of processes achieved through active Involvement of people that maximize the Flow of products, services, information, and money. Quality, Involvement, Flow.
The elements of the Theory we need must then be connected with how to achieve Quality, Involvement, and Flow and, for the purposes of this book, how they can help with the design of an organization fit for Complexity.
Some visuals can help.
Dr. Deming, long before anybody else, understood that organizations are systems, networks of components that work together toward a common goal. In order to achieve this in practice, such components must operate in a state of statistical predictability; otherwise chaos will ensue. Statistical Predictability is the result of processes, methods, and technologies operated by people sharing a precise epistemological framework, who are intrinsically motivated, hence cognitively involved in achieving a goal that betters their lives and the world around them.
(This is my quick take on the overarching vision for Quality that Dr. Deming put forth and, of course, I take full responsibility for what I write.)
Such a system, depicted in Figure 3.1, was first designed for the benefit of Japanese industrialists in 1952, and it was fueled over the decades by a continuously growing body of knowledge, eventually formalized in two pillars of management theory, Deming’s Out of the Crisis and The New Economics.
Figure 3.1 The organization as a system, based on Deming’s original 1952 “Production Viewed as a System”
The arrows represent the systems’ components (processes and projects the company is made up of) and in the circles I have highlighted its main features: focus on the client, feedback loops, emphasis on interdependencies.
Deming’s “Production viewed as a system” was the starting point for the Quality movement and ushers in those elements of integration, collaboration, and transparency that are the hallmark of digital technologies.
Dr. Goldratt, a physicist like Deming, looked at how to emphasize and capitalize on the systemic nature of the work of organizations from a different angle. Any organizational system, he argued, can be designed so as to have only one element (or very, very few) that determines the pace at which it pursues its goal. Such an element, strategically chosen, i.e., capable of offering the highest value, becomes the leverage point that maximizes what the whole system can produce. He called it the “constraint.”
In order to maximize the performance of the constraint (hence, the overall performance of the system):
- The constraint must be rationally exploited, meaning it must work constantly and on the most suitable mix of “products.”
- The constraint must be fully subordinated to—in other words every other component in the system must be designed and operated so as to enhance the performance of the constraint.
- A protection mechanism must be put in place to prevent disruptions (mostly due to statistical unpredictability of processes) to the constraint. It is called a “Buffer.”
Around these deceptively simple concepts, Dr. Goldratt built a wealth of practical applications; more importantly, he developed a comprehensive theory, the Theory of Constraints (TOC), to illustrate all the aspects connected with designing, managing, and evolving a constraint-based organizational system.
Deming and Goldratt grew up in different countries, had different cultural heritages, belonged to different generations, and so was the language they used to illustrate their views and findings. I venture to say that they had radically different personalities, just like their followers and comrades.
Integrating Deming and the Theory of Constraints
Why did Oded Cohen, a leading figure in the TOC world, and I feel the need for a cohesive and rigorous integration of Deming and Goldratt’s message? What is the purpose of the new resulting message that we called The Decalogue (meaning “10 steps”)? What emerges from the integration of Deming and TOC?
For me, at the time, the main drive was scientific: I wanted to see the thinking of these two giants finally unified and expressed in a language that would give justice to their common lifetime goal of providing a more meaningful, intelligent, evolved role for management and leadership. The outcome of this effort, I thought, could be a step ahead in the human quest for better reasoning.
The Decalogue, in its initial verbalization (1999), achieves the goal of laying out many of the prerequisites for sustainable wealth creation. Let’s say that it identifies “the pieces” of the puzzle, but it does not put them together completely. More work was needed, but I am grateful to both Oded and our publisher at North River Press, Larry Gadd, for taking the chance. Today, I stand behind the Decalogue, as it has proven to be scientifically and methodologically very solid.
The questions in front of me were the following: Why have so (relatively) few companies embraced Deming or Goldratt (let alone the Decalogue)? What was the missing piece in their hugely complex body of knowledge? What stands in the way of its adoption?
While we progressed with perfecting methods to facilitate the understanding of the Decalogue, my team and I, working on both sides of the “pond,” realized that any effort to promote the cognitive leap necessary to embrace the Decalogue would shipwreck against the prevailing organizational model: the Functional Hierarchy and the Silo mentality that it entailed. Before we look at that in more detail, we need to go back to the Complexity conflict.
Back to the Complexity Conflict
We use a conflict cloud to make a robust, cause–effect analysis of any situation in which we are stuck. When used in concert with the other Thinking Processes from the Theory of Constraints, the conflict cloud can help us generate powerful solutions that capitalize on our intuition, develop our understanding through analysis, and lead us to design a set of consistent actions that we can execute. The conflict cloud is a Thinking Process that everyone needs to learn. Let’s look again at the complexity conflict.
What are the fundamental assumptions that underpin this conflict (depicted in Figure 3.2)?
Figure 3.2 The complexity conflict
- The whole is equal to the sum of its parts.
- No new properties emerge from interactions among the parts.
- Interactions among the parts are always and only deterministic and linear (mechanistic view).
So, if we want to escape this conflict, if we want to open up to a new way of designing organizations that overcomes the alleged duality of Local versus Global and of Parts versus Whole and that replaces the outdated silos with something more in line with the evolving needs of work, but at the same time can be controlled, we need a new image and a protocol to sustain it.
My team and I began to imagine a way to visually represent the kind of organizational system that the Decalogue would support. The result was published in our book Sechel: Logic, Language and Tools to Manage Any Organization as a Network, in 2011.
The organization design we developed was a breakthrough in terms of how to manage organizations using the Knowledge from both Deming and the Theory of Constraints. We were assisted in developing this breakthrough by following the Thinking Process of the conflict cloud. Our new organizational design became a system, as depicted by Deming in 1952, but where the interdependencies are designed around a strategically chosen constraint, for example, Production, as in our illustration. This constraint is protected by a buffer whose fluctuations are statistically understood. We called the visual depiction of this organizational design “The Systemic Organization Cartoon,” illustrated in Figure 3.3.
Figure 3.3 The systemic organization cartoon
Let’s summarize: if we want to overcome the inherent limitations originated by a Newtonian worldview, we need to embrace the systemic elements of knowledge provided by Deming and TOC and we need to create a suitable organizational framework that allows this knowledge to be applied successfully.
The Systemic Organization Cartoon represents a visual injection (systemic solution) to the conflict of Complexity; this conflict is completely generic. Generic does not mean “overgeneralized.” It means that its validity is so strong that it can be tailored to different contexts. Finding the appropriate verbalization is key to surfacing the blocking assumptions that keep us stuck in the conflict. (For any readers passionate about physics, as the Decalogue methodology was developed also by a physicist, we have added to the appendix a divertissement applied to the duality wave-particle found in quantum mechanics.)
For instance, if we want to tailor the Complexity conflict more specifically to the context of organizations, it can be usefully re-verbalized as Hierarchy versus No hierarchy.
What is the prevailing organizational structure and what is wrong with it? The conflict is a blatant one. On the one hand:
A hierarchy fails to acknowledge three critical aspects of the life of a successful company: interdependencies, feedback cycle, and the customer. For these reasons, it is not suitable for sustaining a continuous improvement effort.
On the other hand:
A hierarchy seems to cater for accountability and provides a sense of control.
The conflict rests, then, on the idea of control and how this idea is translated into a consistent, and consistently measured, organizational system.
Let’s reorganize these thoughts using the Conflict Cloud Thinking Process. It is clear that the hierarchical model is inadequate and prevents organizations from unleashing their true potential. The traditional model for control is a hierarchical/functional structure (Figure 3.4).
Figure 3.4 B and D in the hierarchy conflict
The reasons why a hierarchical structure exists are the assumptions between B and D (Figure 3.5).
Figure 3.5 The assumptions between B and D
Increasing our capacity to listen to the customer so we can satisfy the needs of the market leads us to NOT adopt a hierarchical structure (Figure 3.6).
Figure 3.6 C and D’ in the Hierarchy Conflict
Figure 3.7 shows the pieces of the conflict we have looked at put together with the assumptions.
Figure 3.7 The Hierarchy Conflict with assumptions
The two positions are in conflict because we believe the assumptions in the box on the far right in Figure 3.8.
Figure 3.8 The assumptions between D and D’ in the Hierarchy Conflict
This verbalization of the assumptions makes evident why the Systemic Organization Cartoon is an injection completely consistent with the teachings of Deming and Goldratt:
- Control is exerted by understanding statistical fluctuation in the buffer.
- Design is Systemic, customer focused, and based on accelerated flow through tight collaboration among different competencies.
- Global and Local optima are reconciled by subordinating the whole system to the constraint, hence avoiding the fallacies of cost allocation, functional budgeting, individual performance appraisal, etc.
The Systemic Organization Cartoon is the “injection” (systemic solution) to the conflict. What transforms this injection into reality is an operational plan that gives life to the Cartoon. This is what the Decalogue does.
The 10 Steps of the Decalogue
What follows is a brief explanation of the 10 Steps of the Decalogue.
Step 1: Establishing the goal of the system, the units of measurements, and the operational measurement
Without a goal, there is no system; and without clarity on what to measure in the system and how to measure it, talking about a goal becomes lip service. GAAP accounting utilized to support managerial decisions is the enemy number one of productivity. EBIT, EBITDA, EPS, and any form of GAAP-derived measurements totally miss the point of what a company should strive for. If the goal of a company is connected in any way with making money, then all we need to know is:
- What comes in (Sales);
- What goes out to purchase materials and services that go into the products we sell (TVC, Totally Variable Costs);
- What we need to make the system function (fixed costs + investments), Operating Expenses (OE );
- The inventory (I ) we need to keep in the system to ensure that we always have enough “material” to produce and ship.
These basic variables are connected in the following way:
Sales minus TVC = Throughput (T) T minus OE = Net Profit; T minus OE minus I = Cash Profit, the physical money we see in the bank account before tax. Indeed, from year to year, I becomes Delta I.
Throughput is overwhelmingly more important than Inventory and OE because it can potentially grow far more than OE, and I can be reduced only so much. Sadly, the whole GAAP effort revolves around “understanding OE ” with devastating effects on decision making. Let us say this clearly: The industrial world is where real wealth can be created, and industry needs to learn how to measure its performances based on T, I, and OE. Banks, stock market, financial institutions, and the guild of accountants must understand that their job is to support true wealth creation by fostering in industry the pursuit of these measurements instead of thrusting upon them anything different.
Over 30 years ago, Dr. Goldratt explained these simple concepts, and all the financial and industrial catastrophes we have experienced in the last 25 years are, directly or indirectly, the result of not having understood them. Throughput is the most powerful way to link the industrial and the financial worlds, and the accounting that goes with it provides a whole new and meaningful scope to the endeavors of the accounting profession.
It is also critically important to measure how the system performs in its effort to deliver to customers. We can have a pretty good handle on it by measuring the on-time delivery and assigning a dollar value to any delay; we call it Throughput Dollar Day (T$D). The longer the delay, the higher the T$D. Similarly, we want to know how much cash we keep trapped in Inventory to achieve an optimal, ideally zero, T$D, and we call it Inventory Dollar Day (I$D).
Step 2: Understand the system (draw the interdependencies)
If we do not know who does what, what the inputs and outputs are, and how everybody’s work is connected, then we are not managing. As we have mentioned, Dr. Deming declared that “Business schools teach you how to raid a company, not how to manage it.” It is disconcerting to discover how little we know about our system until we have a clear picture of how our interdependencies are laid out and how neglected by top management this issue is. Through mapping out all the main processes of an organization, Step Two provides the foundational elements of understanding that will enable the building of a truly effective Quality System.
Step 3: Make the system stable (understand variation and its impact on the network)
Dr. Deming used to say that if he had to reduce his message to Management to just a few words, he would say that their job is all about reducing variation. Dr. Deming’s unrivaled contribution to the science of management comes from having understood the importance (and all the implications) of a body of statistical studies developed by Dr. Walter Shewhart of Bell laboratories in the 1920s. Shewhart found out that any process is affected by variation, which is enemy number one of Quality and reliability; however, this variation can be attributed to either common causes or special ones. Distinguishing “noise” from “signal” was then critical to devising actions aimed at managing this variation. Dr. Shewhart developed an important part of the Theory of Variation known as Statistical Process Control (SPC) and a very useful mechanism for measuring variation called a Control Chart (also known as Process Behavior Chart). His work was foundational for the improvement of productivity, first at Bell and then in a myriad of organizations countrywide, and certainly served as a springboard for the gigantic work of Dr. Deming.
Dr. Deming realized that, in order to manage the variation associated with a process, people make two kinds of errors:
- They attribute variation to a special cause when it is instead due to a common cause.
- They attribute variation to a common cause when it is instead due to a special cause.
The ramifications of these errors are endless and still today plague the way companies make decisions; Statistical Process Control (SPC), the main body of knowledge from which the Theory of Profound Knowledge has evolved, is largely ignored at Top Management and Corporate level. SPC is neither a purely mathematical tool nor a conventional financial management tool. The process behavior charts mirror the outcome of our managerial decisions, often taken with some kind of local optima in mind. The image they portray is not the reassuringly deterministic one that accountants and financial people are accustomed to; on the contrary, what these charts display are predictable or unpredictable ranges of oscillation. These oscillations very often reflect, mercilessly, the conflicting confusion that dictates our choices, and the course of action that their interpretation calls for flies in the face of “conventional wisdom.”
Understanding SPC, let alone using it properly and accurately, requires a paradigm shift in the way we look at data and make sense of them for business decisions. From a mathematical standpoint, process charts are based on an average dispersion statistics and the approach they use is the 3-sigma one for the calculation of limits. What has always generated confusion about SPC charts is that, although connected with probability theory, they do not work because of it. The essence of the charts is in their predictive role and in the possibility that they provide to build an epistemological approach to management based on prediction.
Control charts capture the most fundamental feature of the work of individuals and their interaction within an organization, the variation associated with processes. In building a systemic organization based on the Decalogue, SPC has a central role.
Step 4: Build the system around the constraint (subordinate the organization to the constraint)
The new kind of organization that is based on managing variation and constraints is a network of interdependent processes with one common goal where we have achieved a good level of statistical predictability. It can be successfully managed, but the question is, how? Dr. Goldratt’s main contribution to the Theory of Management has been to point out that any system is limited toward its goal by very few elements, the constraints. If we identify them, we can manage them following the steps of focusing that he developed. The Decalogue, leveraging the intrinsic stability of a Deming-based system, suggests that the constraint can be “chosen” (one constraint) instead of being identified. In other words, we can always decide which constraint it is strategically more convenient to focus on and build the system accordingly. However, we can do it only because we have already built a system made of low variation processes; this is why we can safely design our company around a strategically chosen constraint. Instead of cycling the five focusing steps of TOC, namely, (1) identify the constraint, (2) exploit the constraint, (3) subordinate to the constraint, (4) elevate the constraint, and, if the constraint has moved, (5) go back to step (1), we can make the system grow by appropriately choosing the constraint and sizing the capacity of all the feeding/subordinating processes coherently. Again, this is possible only because the variation in the system is low.
Step 5: Manage the constraint (protect and control the system through Buffer Management)
The performance of this system is ensured by its predictability but we need a mechanism to protect it and control it. This is provided by the buffer and by its management. What is a buffer? Individual processes exhibit variation; two or more together do too, let’s call it covariance. The effect of covariance is a cumulated variation that can result in any sort of combination of these variances. Regardless of how little individual variation the processes of the system have, we do need a mechanism to protect the most critical part of our organization, the constraint. A buffer is a quantity of time that we position in front of the constraint to protect it from the cumulative variation that the system generates. Simply put, in synchronizing the processes that deliver the output of our organization, we ensure that what has to be worked on by the constraint gets in front of it “one buffer time ahead.”
Is there a precise and unequivocal recipe to size the buffer appropriately? No, and it doesn’t matter. If you understand control charts, the sizing is fairly obvious. The real importance of the buffer is not in the protection from disruption, certainly not an irrelevant issue, but in the possibility to exert control over the functioning of the whole organization: If our processes are all predictable, then we can truly use the buffer as a mechanism to gather succinct, effective, and comprehensive information about the “state of synchronization” of our company. Indeed, if we monitor the buffer with the use of a control chart, then the control becomes real insight, the goal of management.
The management of the buffer entails a Copernican shift in the way we mean control because it prompts decision makers to rely on understanding and knowledge rather than on hunches; it forces them into leveraging arguably esoteric statistical studies instead of salt of the earth accounting; it constrains the unbridled vibrancy of managers into the straightjacket of rational thinking. As a result of those considerations, over the decades, many of the repeated attempts to introduce buffer management to Top Management have failed.
Step 6: Reduce variation of the constraint(s) and the main processes
Step Six is obviously connected to Step Five but less obviously to Step Seven. Clearly, we do understand the impact that variation has on our system and the need to reduce it but when push comes to shove, we are not prepared to continue to work on variability reduction. Why not? The answer is in our ability and desire to understand the purpose of system management.
The culturally disappointing translation of Deming’s Philosophy into a plethora of “Kaizen-like” management techniques has transformed it from a vision of the world based on innovation and wealth creation into an efficiency game. If, on top of this, we continue to view our company in “functional” terms, then reducing variation simply means reducing costs. Of course, no function would ever easily surrender to that because it would imply “cutting the budget”; hence any serious attempt to reduce variation is nipped in the bud. A relentless effort toward continuous reduction in variation can only stem from a systemic vision of our company and the understanding that only this reduction would provide the insight needed for triggering real jumps in performance. The way to link a relentless, focused, and companywide variation reduction crusade to financial performance is through the adoption of a suitable organizational structure.
Step 7: Create a suitable management/organizational structure
At the time of publication of the book Deming and Goldratt: The Decalogue (1999), Step Seven was not clearly and comprehensively elucidated upon.
It was clear that, without a suitable structure, the realistic possibility to sell all the capacity of the constraint would be hindered by local optima considerations. In other words, the design of a suitable structure was a prerequisite for enabling the true expansion of the system. Today it is much clearer that the natural way to see the work of an organization is in terms of process and projects. In the years that followed, further research and application led to the development of the Network of Projects organizational design as a means to improve and optimize the performance of the whole system.
Step 8: Eliminate the external constraint (sell all the capacity the system has available)
When we design a system that caters for a high degree of process predictability and synchronization, where control and protection are ensured by buffers and where all the policies, behavioral, and measurement “constraints” are dealt with by an appropriate organizational structure, we do so to maximize sales. The most important part of the chain is the customer, and any company should always be designed to ever improve its ability to satisfy its customers’ verbalized and hidden needs. The Decalogue, if understood and embedded in the appropriate structure, should very quickly reveal capacity that is not currently being sold.
Another way of looking at this issue is the following. Let’s say that, day one, the rate of sales of a company starting its Decalogue journey was such that some shipments were missed and constant fire-fighting would create friction between production and sales. The Decalogue would call for a disciplined process mapping effort aimed at understanding process variability and the choice and management of a suitable constraint as well as the devising of a coherent measurement system. Moreover, in order for this level of synchronization not to be hindered by the local optima inclination that functional organizations invariably undergo, we would design an appropriate, coherent organizational structure. Almost invariably, the constraint would shift outside, become an “external constraint”: our capacity, what we can realistically design, manufacture, and ship, would become greater than what we were currently capable of selling. At this point, it would be blatantly obvious that our real understanding of the market is woefully limited and we, in truth, do not know how to sell. Goldratt developed a powerful and detailed Thinking Process approach to managing the external constraint. This is a very critical point in the pattern of a successful Decalogue implementation. Why? When left to behave in a traditional local optima way, salespeople can singlehandedly jeopardize any systemic endeavor unless we integrate them organically into the way the company operates.
Step 9: Where possible, bring the constraint inside the organization and fix it there
The Decalogue approach to management is based on process stability; indeed the most critically important part of the system is the constraint. Hence, we want to ensure maximum predictability, especially on the constraint. Clearly, when the constraint is external such predictability is more difficult to achieve. This is the reason why, whenever possible, we want to manage an internal constraint. Moreover, this is also the easiest way to make the system grow without stirring chaotic company dynamics.
The need for constraint reliability is so strong that even when organizations have a virtually unlimited internal capacity, like supermarkets, we should always elect to appoint an internal constraint and subordinate the whole organization to it. The growth of the system would then happen through a systematic, orderly, and relentless exploitation of the capacity of the constraint. When such a capacity is not sufficient to meet market demand, we would first increase the appropriate non-constraint areas to make them capable of subordinating to the constraint and only then would we elevate (increase the capacity of) the constraint. Once again, the name of the game is process predictability.
Step 10: Create a continuous learning program
The possibility for a Decalogue-based management system to produce results over time rests on the ability that the organization has to continually learn what is needed to constantly improve its performances.
Learning does not happen in a vacuum and cannot be based solely on individual desire. Learning and personal development must become part of the way the organization functions, and the change associated with it must become a way of life for the company. Learning cannot be “installed,” nor can it be forced on people. It must be a personal choice, but also an integral part of the way the company has structured itself to conduct business. Learning and the self-development that comes with it must be promoted companywide and from the Top Management, but must come from a designated and empowered source. Over the years, we have come to call it a Center for Learning.
The Decalogue is founded on the principle of continuous improvement. Not only does it employ the Plan, Do, Study, Act continuous improvement cycle designed by Deming, but the entire 10 steps embody that cycle pattern in the way they are carried out. This can be seen clearly when we map out the 10 steps of the Decalogue onto the PDSA cycle itself (Figure 3.9).
Figure 3.9 Steps of the Decalogue mapped onto the PDSA cycle
Summary of Chapter 3
- In order to embrace and make the most of the changes that digital technology is bringing, companies need to focus on improving Quality, Involvement, and Flow.
- Organizations are systems, networks of components that work together toward a common goal.
- The key focus areas for managing a company as a system are managing variation and managing a strategically chosen constraint.
- If we look at the conflict of Hierarchy versus No hierarchy, we see that the 10 transformational steps of the Decalogue provide a valid solution to shift away from silos and toward an organization as one whole, synchronized system. The Ten Steps are:
1. Establish the goal of the System, units of measurement, and operational measurement
2. Understand the system (map the interdependencies)
3. Make the system stable (predictable)
4. Build the system around the constraint
5. Manage the constraint - Buffer Management
6. Reduce variation at (of) the constraint and the main processes
7. Create a suitable management/organizational structure
8. Eliminate the “external constraint” – sell all the capacity the system has available
9. Where possible, bring the constraint inside the organization and fix it there
10. Create a program of continuous learning