Mastering Complexity

Adding Coherence Throughout Your Business with Dependency Structure Spreadsheets

Stephen Denker

• Publisher: Productivity Press
• Year: 2017
• ISBN:
  9781000730128
• Edition: 1
• Topic: Business
• Subtopic: Operations
• Index: Business

Mastering Complexity is designed to help readers develop and maintain logical, adaptable solutions to business problems. It demonstrates how to visualize business dependency connections with a square spreadsheet called a Dependency Structure Matrix (DSM). Once the links representing dependencies are visualized in a DSM, it can be used to improve th

Section 1

Work Has Structure

By relieving the mind of all unnecessary work, a good notation sets it free to concentrate on more advanced problems, and in effect increases the mental power of the human race.

—Alfred North Whitehead

1

What Is Structure?

Does our organization functional structure match the logical dependency among its elements? If not congruent, it costs us time and money. If we visualize our dependency structures, we can improve our business or know better how to cope.

Some think the only structure in a business is its organization chart. An organization chart is a valuable administrative convenience. But, it should not be confused with the what, why, and how of the business. All too often, it is the organization chart, not the business, that is managed (see Figure 1.1).

Others think structure means the design of organizational workflows and business processes.

• In this book, structure is any pattern of logical relationships among key components.

What we call structure shows how elements are logically connected together. Structure shows whether one part needs another. The meaning of why they do is called semantics. Together, they describe our problem.

• We make this distinction so we can extract and study the dependency structure of a problem separate from its meaning.

Every business has these structures. They might include our whole organization, its employee attitudes, the product quality, the ways decisions are made, or hundreds of other factors. The word structure comes from the Latin struere, “to build.” Dependency structures are built from the choices we make—things inside or outside our organization or anything that matters anywhere. Structures are built from the choices we make over time.

FIGURE 1.1
How dependencies travel throughout an organization and beyond.

When we ask, What happens if x changes? we begin to see how an element might be part of a dependency structure. Dependency structures occur everywhere.

• If we can make them visible, the implications are enormous.

Actions cannot be considered independently. If we are used to thinking in terms of tasks rather than their underlying dependencies, we may need to recast our point of view to move from a logistics view (managing tasks) to a structural view (managing dependencies).

Every business consists of networks of dependency structures, of persons and their dependent activities. Why they do what they do is called semantics. Together, structure and semantics describe our work.

• The structure and its semantics determine how our business performs.

As dependency structures become larger, diagrams drawn to illustrate relationships can become cluttered with arrows and connecting lines.

“To assemble an annual financial plan for its 3,500 U.S. stores, Sears collected data scattered across many computers. It took a 100-square-foot diagram [emphasis added] to describe their 300-step consolidation process” (Business Week, October 29, 1996: 131). A simple DSM spreadsheet model could have captured all the relevant paths in one clear spreadsheet. A DSM spreadsheet concisely visualizing both entire enterprises and their local connections on one page would clearly show how one thing triggers a cascade.

Figures 1.2 and 1.3 are other examples of how complexity that could overwhelm our ability to comprehend is made more manageable when converted to a DSM spreadsheet visualization. The process of how to accomplish this is illustrated in the chapters that follow.

Our client, in preparation for reworking data entry capabilities in anticipation of possible Y2K (year 2000) computer errors as the twentieth century was ending, diagrammed their enterprise as shown in Figure 1.2. I joined the project. Our first task was to draw the equivalent DSM spreadsheet shown in Figure 1.3. With the DSM spreadsheet visualization complete, immediately our client saw three errors in their model.

• Correcting their graph took some time.

• Correcting the DSM spreadsheet visualization took less than one minute.

With DSM spreadsheets, we can show clearly how individuals are connected. We can understand how their relationships are dependent. More important, we can systematically revise and optimize these dependencies one at a time or as a whole.

In this book, we first treat simple structures. As illustrated problems become more complex, we introduce the tools we need. With these tools, we can consider the concepts as applied to larger systems.

Let us get started!

FIGURE 1.2
Original Y2K enterprise diagram drawn as a graph.

FIGURE 1.3
Original Y2K enterprise diagram drawn as a DSM spreadsheet.

2

Why a Structural Model?

How can we know that a proposed organizational change will be an organizational improvement?

• By looking at the organization’s logical structure.

Using a DSM spreadsheet, we can make this structure visible and its key logical elements self-evident. A DSM spreadsheet helps us articulate, verify, and reorganize structures.

• We use the general term element to denote a row (and its column) in a DSM spreadsheet.

In a DSM spreadsheet, each element will be represented twice—once as a row and again as a column. Elements are listed down the left-hand side of the DSM spreadsheet. The same list of elements is placed along the top of the DSM spreadsheet, from left to right. Their row order and their column order are identical. Each element’s row and column intersect at the DSM spreadsheet diagonal cell.

• Elements in a DSM spreadsheet can represent many entities: tasks, physical objects, time, human resources, a person’s point of view—anything. And, they need not all be the same type.

Also, we sometimes refer to DSM spreadsheet elements as nodes. A graph is a set of points, or nodes, and the lines between them. When the lines have a direction, they are called arcs in a directed graph. A DSM spreadsheet and directed graph are alternative, equivalent ways to map dependencies. A sequence of arcs from node to node is a path.

Some nodes might be arranged in parallel paths because they do not depend on each other. Others might be arranged in sequence because each depends on the results of one before.

If a path eventually returns to a node, the path is a circuit. Circui...