In today’s business environment, the amount of information being created is far greater than at any time in history. And there appears to be no end in sight. The beauty of systems thinking is that it is ideal for managing dynamic organizations that are rich in information.

Most businesses have a plethora of sophisticated tools for forecasting and analysis, the results of which are fed into elegant strategic plans.

Dynamic complexity takes on various forms. It is present when the short-run and long-run effects of an intervention are dramatically different. Or when an action has one effect on one part of a system and a different effect on another part of the system. Or when obvious actions or interventions produce nonobvious outcomes.

As Senge puts it, “The real leverage in most management situations lies in the understanding of dynamic complexity, not detail complexity.”² Consider the dynamic interplay of forces in the U.S. economy. As witnessed in the financial collapse of 2008, a new type of investment that entered the system caused an eventual imbalance that wreaked havoc on many innocent parties.

Unfortunately, the typical approach to increasing complexity is increased analysis. Senge points out that “Simulations with thousands of variables and complex arrays of details can actually distract us from seeing patterns and major interrelationships.”³ Most people tend to battle complexity with more detailed solutions. However, this is the opposite of systems thinking.

Senge says “The essence of the discipline of systems thinking lies in a shift of mind. It requires seeing interrelationships rather than linear cause-effect chains, and seeing processes of change rather than snapshots.”⁴

The free flow of information is the very lifeblood of organizations. And the larger the company, the more critical it is to keep it flowing. As information moves, it becomes self-generative. As information is sent, received, and processed, new information is created. In this model, feedback loops take on a different meaning. Traditional feedback loops were designed to measure deviations from the norm. The goal was to detect problems or ensure compliance, thereby ensuring the stability of the system. However, to accommodate an adaptive system, positive feedback loops are necessary. These may create short-term disturbances, but they magnify information that creates the impetus to move a system forward.

Fundamental truths for all systems regardless of their type include these assertions:

Understanding of systems is achieved through identification, modeling, and analysis of relationships and interactions among the parts of a system—a distinctly different and more in-depth analysis than is possible with structural models of a system. Systems modeling is performed by representing the parts of a system and the interactions among those parts.

The most basic concept of systems theory is that a system is a collection of interacting things. The use of the word thing avoids the context-based connotations that might occur with terms such as entity, object, or component.

Things in a system are of many types. They may include (but are not limited to) entities that are familiar to data modelers, objects that are familiar to object-oriented systems analysts, and components as they are understood by software developers. Things in a business system encompass artifacts such as resources, capacities, limits, gaps, goals, desires, actions, results, plans, processes, rules, standards, and much more.

Influence is a behavioral characteristic of interaction. Interaction between two things in a system is directional—one thing has influence on another thing. System behavior is important to understand why things happen in a system and to predict what may happen in the future. Analysis of influences is the key to understanding system behavior.

Influences are of two types—same direction and opposite direction. A same-direction influence means that the values of two things move in the same direction when change occurs: When employee morale increases, employee productivity goes up. An opposite direction influence means that the values move in opposite directions: When employee stress increases, employee productivity decreases. Exhibit 8.1 illustrates how these two examples are modeled. Note that a plus (+) indicates same direction and a minus (-) is used for opposite direction.

The diagramming technique is called causal loop diagramming because real understanding comes from understanding the system as a whole. Cause and effect is typically not linear; it is circular with a sequence of influences producing a feedback loop. Loops are closed structures that represent a sequence of system interactions without a beginning or an end. A loop may contain any number of interactions greater than one. Feedback is a characteristic of loops in systems.

Feedback is a process by which the results of an activity or action are returned to the actor in a way that influences the behavior of that actor. Positive feedback occurs when the cumulative effect of all interactions in the loop is one of growth, amplification, or acceleration. Positive feedback loops are often called reinforcing loops. Negative feedback occurs when the cumulative effect of all of the interactions is stabilization or equilibrium. Negative feedback loops are also known as balancing loops or goal-seeking loops.

Exhibit 8.2 illustrates both kinds of feedback loops. Note that the kind of feedback loop—positive or negative—is indicated using a polarity symbol at the center of the loop. Polarity describes the positive or negative feedback property of a loop. Determining loop polarity is relatively easy. Simply count the number of subtractive interactions in the loop. An odd number indicates negative polarity, and an even number, positive polarity.

Individual feedback loops are a step toward understanding cause and effect, but they only scratch the surface. Often the interactions among loops provide real insight into system behaviors by breaking down stovepipe views of the parts of a system. Exhibit 8.3 illustrates this principle with only one minor change to the diagrams shown in Exhibit 8.2. The new model shows a connection between the two feedback loops. Finding these kinds of connections is the first step to developing a holistic view of a system.

In reality, a system consists of many loops and many interactions among those loops. It is that total system view that helps to achieve depth of understanding and real insight into the behaviors of complex systems. The intersection nodes—those that participate in two or more loops—are the core of system complexity, and they provide the greatest opportunity to discover side effects, hidden influences, and unintended consequences.

Determining the boundaries of a system model can be challenging. Every system is a part of some larger system. Therefore, it is possible to continue modeling infinitely. The time to stop modeling is when enough knowledge and information has been acquired to satisfy the purpose of the model. Stopping too quickly, however, presents a risk that side effects and unintended consequences might be overlooked. Exhibit 8.4 illustrates the nature of this challenge.

The discipline of systems thinking includes several archetypes—generic models that represent recurrent patterns in systems. The names of the archetypes are fascinating in themselves: accidental-adversaries, fixes that fail, drifting-goals, tragedy of the commons, and so on. But even more interesting is the clear and certain relationship that exists between these archetypes and the patterns seen in time-series analysis.

As a sociocultural example, consider the conflicting goals and activities of national security versus workforce economics as they relate to U.S. immigration policies. In a more business-oriented scenario, consider an example where two people are managing separate but related software development projects. Cooperatively they have agreed to develop shared and reusable software components, where practical. Yet each of them, when faced with schedule pressures or conflicting needs, chooses to build local custom components.

In behavior-over-time analysis, accidental-adversaries graphs as two activities—X and Y—which both experience accelerating growth early in the time scale. As local optimization limits success potential of both activities, they each decline in the later stages of the time scale. Exhibit 8.6 shows the graphical pattern of accidental-adversaries.

Consider this scenario: A company is scheduled to negotiate its sales revenue budgets annually. In one year, actual revenue would significantly exceed the budget, creating pressure to increase budget in the following year. The higher budget in the second year caused actual revenue to fall short of budget, creating pressure to reduce the revenue budget in year three. What are the implications of this seesaw budgeting pattern continuing over several years?

In behavior-over-time analysis, drifting-goals graphs as mildly oscillating patterns of both the current state and the desired state. Current state increases slightly over time, as desired state experiences a slight decrease.

The cold war is an obvious sociocultural example of escalation. Competitive pricing is a common business-oriented example. It is common for retailers to advertise that they will match any competitors’ price. What would be the eventual outcome if two retailers each established a policy of beating the other’s best price by 5 percent?

In behavior-over-time analysis, escalation graphs as two activities—X and Y—that each grow in a stair-step pattern, with each as the driving force for the next growth step of the other. Exhibit 8.10 shows the graphical pattern of escalation.

Free phone promotions in the wireless telecom industry are a sociocultural example of fixes-that-fail. A business-focused example is a company that is losing customers due to long wait times at the customer service call center. To improve customer retention, the company decides to outsource call center operations. The early result is a visible reduction in wait times and a corresponding reduction of customer attrition due to call center waits. After several months, however, the customer retention rate flattens and begins to trend again toward attrition. What may be the cause, and what fundamental solution may resolve it?

In behavior-over-time analysis, fixes-that-fail exhibit an oscillating pattern of increase followed by decrease. Each of the increases coincides with the introduction of a symptomatic solution. Each decrease that follows is the result of unintended consequences of the fix that become visible only after some delay. It is common that the time intervals between cycles decrease over time and that the amplitude of each wave also shrinks. Exhibit 8.12 shows the graphical pattern of fixes-that-fail.

One-hour photo developing is an example of limits-to-success where the limiting factor is the emergence of digital photography. Limiting factors may appear in many different forms, including capacity constraints, market saturation, aging product lines, emerging technologies, resource limits, and so on.

In behavior-over-time analysis, limits-to-success exhibit a growth curve that shows early acceleration, followed with deceleration and eventual flattening over time. Exhibit 8.14 shows the graphical pattern of limits-to-success.

A common example here is overuse of temporary labor to balance workforce capacity with workload demands. Temporary labor satisfies the immediate need to increase capacity. But, when used repeatedly, the percentage of the workforce that is classified as temporary grows, and many “temporary” workers become a permanent part of the workforce. Ultimately issues of Fair Labor Standards Act compliance, benefits eligibility, and such emerge—sometimes resulting in legal action and financial penalties.

In behavior-over-time analysis, shifting-the-burden shows an oscillating pattern of erratic growth of a symptomatic solution. A corresponding (but not always graphed) pattern of oscillating decline in the viability of a fundamental solution occurs simultaneously. Exhibit 8.17 shows the graphical pattern of fixes-that-fail.

Consider the example of two departments in a company that are competing for priority of information technology (IT) projects. The marketing department needs data and technology to get a 360-degree view of customers and the marketplace and to manage effective marketing campaigns. The research department needs modeling and simulation technology to drive innovation of new products. Both projects are initiated at similar times. In a span of a few months, the marketing department illustrates success with campaign effectiveness metrics. In the same short time span, the research director has only anecdotal justification for the simulation project. The demonstrable success of marketing is reasoned to justify assigning more IT resources to marketing projects, which takes them away from research projects.

Success-to-the-successful graphs as two activities—X and Y—with divergent patterns. The activity to first demonstrate success (illustrated here as X) shows a growth curve, while the other shows a corresponding decline. Over time, the gap widens. Exhibit 8.19 shows the graphical pattern of fixes that fail.

Consider the example of a company that depends extensively on the subject and domain knowledge of one person. Initially that person provides valuable knowledge that fuels growth of programs, products, marketing, sales, and quality. As each area grows, demands on the expert increase to the point where he or she cannot keep pace with demand. Ultimately the demands become burdensome, the job becomes unrewarding, and the expert resigns.

In behavior-over-time analysis, tragedy-of-the-commons illustrates three variables. As seen in Exhibit 8.21, two activities—X and Y—exhibit early and steady incline followed by late and rapid decline. The common resource of limited capacity exhibits rapid growth of demand (coincidental with the growth peak of the two activities) followed by very rapid decline.

The things that accumulate in a system are called stocks. A stock is an accumulation of something in a system—either concrete and tangible things (i.e., dollars or widgets) or abstract and intangible things (i.e., knowledge or morale). Tangible stocks are accumulations of consumable resources. Intangible stocks are accumulations of catalytic resources.

A stock changes through the influences of flows. A flow is an action that influences a stock by increasing or decreasing the quantity of the stock. Flows are of two kinds: inflow, which increases the accumulated quantity, and outflow, which decreases the quantity.

The relationships of stocks and flows are graphically represented using stock-and-flow diagrams. Exhibit 8.22 shows a simple stock-and-flow diagram for the accumulation of workforce capacity.

The diagram notation is:

The rest of this section builds on the simple example shown in the workforce capacity model. It is important to mention, however, that stock-and-flow models are not always as simple as one stock with two flows. Stock-and-flow sequences may involve more complex and interrelated stocks and flows, as shown in the materials to shipment sequence illustrated by Exhibit 8.23.

External influences often affect the rate of a flow. In stock-and-flow modeling, these influences are known as converters. (The term converter may seem odd for this concept right now. However, it is standard stock-and-flow terminology and will make sense later.) Connectors link converters to flows as shown in Exhibit 8.24.

A systematic process of working from a CLD to create stock-and-flow diagrams uses these seven steps:

The act of creating stock-and-flow models from causal loop diagrams also serves to test the causal models and make them more complete. It is common, for example, to discover influences previously not modeled when analyzing rate of flow and identifying converters that affect the rate.

Labor budget is a stock whose inflow is labor cost allocation rate. The cost allocation rate is a converter that affects hiring rate, which is an inflow to workforce capacity. Similarly, outstanding orders is a stock with the inflow of order received rate. In both instances the converter link is a flow-to-flow connection. The stock itself is never used as a converter. The result of this analysis is greater insight that may add understanding and detail to the CLD.

Here is where the term converter begins to make sense. The units of measure vary among the three stock-and-flow sequences. Labor budget is measured in dollars. Workforce capacity is measured in FTEs. Outstanding orders is measured as days to complete. A conversion formula is needed to describe the influence of labor cost allocation dollars upon hiring rate FTEs—in other words: How do dollars convert to FTEs? Similarly, there is a need to determine how days to complete is converted to FTEs for the influence of outstanding orders upon workforce capacity.

But further consideration suggests that the BI connection is much deeper than simulation and prediction. The power of BI is in the ability to deliver insight, gain understanding, enable reasoning, support planning, and drive innovation.

The example shown in Exhibit 8.27 is identical to that of Exhibit 8.26 with only one exception. The model in Exhibit 8.26 does not link order received rate to labor cost allocation rate—the converter that is shown as a dotted line.

This example illustrates: