The first step in designing system architecture is to define the system's boundary and appreciate the environment in which it intends to operate.

To define a system's boundary we need to understand the behavior of its stakeholders. A stakeholder of an organization is any individual or group who is directly affected by what the organization does and therefore has a stake in its performance. Therefore, we need to know the following: Who are the major stakeholders? What are their expectations? What are the desired properties of the system from their perspective? What is their influence? Which critical variables do they control (or influence)? For example, in a market economy the customer provides the operating income, the boss defines the membership, stockholders pay the capital, suppliers are the source of complementary technologies, and distributors provide access to customers. However, stake and influence do not necessarily go hand in hand; a high stake is often coupled with low influence, and vice versa. For example, customers with the highest level of influence (refusing their patronage) show a very low level of stake in the system. On the other hand, employees with a very high stake in the system often have low level of influence. Shareholders with very high influence have the least at stake. If unhappy, they simply take their money out.

As we alluded to before, the system's boundary is a subjective construct defined by the interest and level of influence and/or authority of the participating actors. Therefore the system consists of all variables that could be sufficiently influenced or controlled by the participating actors. Meanwhile, the environment in which the system must remain viable consists of all those variables that, although affecting the system's behavior, could not be directly influenced or controlled by the participating designers.

In this formulation the business environment must remain intact, so it is critical to get an appreciation of its behavior by understanding: How is the game evolving? What are the drivers for change? What are the bases for competition?Figure 9.3 captures the drama of the health-care system. The game is still evolving. The industry has yet to find alternatives for the three villains of third-party payer, cost plus reimbursement, and fee for service, which are assumed responsible for creating insatiable demand. Note how a counterintuitive event — General Motors' agreement with the labor union to cover the health-care costs of workers in exchange for a one-time concession on wages combined with President Johnson's Medicare and Medicaid programs — led to the “health-care bonanza” and a dynamic that no one has figure out how to deal with.

As for the change in the bases of competition, consider the fact that we are past the age of mass production based on the interchangeability of parts and labor. We have entered the mass customization era where success is defined by producing smaller batches of customized products at lower break-even points. In 1999 Honda's break-even point for a given model was two thousand cars, whereas Ford Motor Company needed five hundred thousand units of the same model car to break even. Economy of scale and mastery of a simple procedure was the cornerstone of mass production, but the multi-skilled knowledge worker is now the core requirement of mass customization. Globalization means that price is set in the global marketplace, therefore, it is considered an uncontrollable variable and the cost becomes the controllable variable (target costing). In the “cost plus economy” cost is considered uncontrollable and price is the controllable variable (target pricing). Efficiency in performing a routine used to be a virtue when the cost of automation outweighed the cost of labor. The digital revolution has reversed the equation. Increasingly cheaper computers have rendered the “narrow skill” and “simple procedure” approach obsolete. When it comes to performing a specific procedure the computer is the ideal actor. Knowledge workers of today are capable of putting together whatever pieces of know-how it takes to produce an integrated solution.

There was a time when problems could be neatly formulated and conveniently solved within the confines of a single discipline or department. Modern problems, however, are increasingly complex and interrelated. These problems are messy; they come in bundles and require a different approach. The knowledge workers of today are not only required to be competent in their own vocation, but they are supposed to be intimately aware of the total context and overall process within which they are to collaborate.

Many common sense statements invoked in the process of developing organizations prove counterintuitive. The surprise element is not because people are knowingly disingenuous. Emerging consequences contradict expectations because the operating principles are rooted in assumptions that belong to different paradigms. Such surprises are inevitable unless the underlying premises are surfaced and their eventual consequences are mapped out.

The purpose of business enterprise is essentially defined by its implicit paradigm. If the organization in question operates in a mechanistic mode, then it is considered to be a tool of its owner. Its purpose therefore would be to serve the purpose of its owner. Performance criteria of a tool, of course, would be efficiency and reliability. Purpose of organization in a biological mode of operation is survival, thus its performance criterion would be growth at any cost. Profit as the means of growth finds additional social value and overcomes the negative stigma of the mechanistic era. Since organization in a sociocultural mode of operation is considered to be a voluntary association of purposeful members, its purpose would be to serve its members and its environment by doing more and more with less and less.

As stated so elegantly by Milton Friedman (1962), ultimately “business of a business is business”; therefore, for designers of business architecture understanding the vision of the desired future of the enterprise and its business model are of the utmost importance.

The initial sketches of the vision of the desired future usually can be put together after the first iteration of interactive design. Although we are told that visions are escape mechanisms of daydreamers, without a vision there will be no sense of direction. Without a vision all possibilities would have equal values; there would be no basis to judge the relevancy of the emerging opportunities.

The business model, on the other hand, defines the way a business generates value, creates a deliverable package, and exchanges it with money. Today amazing originality is shown when developing new business models, which is radically influencing the traditional business concepts. Consider, for example, the case of the search engine Google. By providing free service to a group of consumers it can make so much money from advertisers that it exceeds the value of the giants listed in the Dow Jones index. Creating a multi-billion-dollar business to create, package, and sell operating systems independently from computer hardware, as is done by Microsoft, was inconceivable for those of us who worked for IBM during the 1960s. In the early 1990s, I was involved with a Fortune 100 company as a consultant in the acquisition of a $10 billion business. The total sum of money that all service providers — legal, financial, and management — charged the client was less than $500,000. This fee was based on actual time spent multiplied by the hourly rate. Today the same task would cost over $200 million based on a business model that works on percentage.

In general, business is usually defined in terms of three dimensions: a know-how or technology, which is transformed into a set of tangible products or services and delivered via an access mechanism to its target customers or markets. Business architecture defines the nature of relationships among the three dimensions of technology, product, and market.

Traditionally one of these dimensions has been designated as primary, forcing the other two into subordinating roles.

When the product defines the business, technological requirements and the markets to be served are determined by product characteristics. Alternative technologies are sought for making the product, and different markets are sought for selling the product (Figure 9.4).

The success of a product-based business is usually measured by the success of its product divisions, which is why everything is easily compromised and subordinated to the product. However, when the potential value of a given technology somehow exceeds the value of the product it supports, the business faces a dilemma.

Unfortunately, experience shows that, in these cases, technology is always compromised. In a product-based business, the product managers are the bosses. They see technology only as a competitive advantage for their product; therefore, technology will never get the chance to realize its potential in the context of a product-based business. Quite a few cases support this assertion, but the most conspicuous is the fate of the first Apple operating system (the famous graphic interface). This phenomenal operating system was used exclusively to sell Apple computers when its potential had been much higher than the maximum potential of the initial Apple products it supported at the time. This was proven by Microsoft, whose claim to fame, Windows 95, was an imitation of the initial Apple operating system used by all computer producers. It was reported that IBM spent billions to lay off about 200,000 people — some say the best available knowledge workers. But it did not consider capitalizing on its advanced “digital technology” and state-of-the-art “C4 technology” to become the major player in the most potentially explosive market of the time — “electronics packaging.”

When the market is used as the basis for defining the business, the characteristics of the market determine the product mix and the type of skills and technologies required to produce it (Figure 9.5). Procter and Gamble, for example, provides as many of the products sold in supermarkets as it can profitably make.

However, the trauma of the defense industry, when its market collapsed overnight, was another confirmation that the price tag on selecting a single pattern of existence is very high. The industry's inability to take advantage of its access and knowledge of emerging Internet technology was not surprising. Managing technology requires a different set of criteria than those necessary to manage products and/or markets.

Finally, when the business is technologically defined, a variety of products are developed around a given core technology, using the same base of know-how, and sold in different markets (Figure 9.6). For example, 3M defines itself as being in the “sticking” business. The company revolves around technologies used to bind different things together (from Scotch® tapes and composite structures to electronics packaging).

Using its materials science and processing technology, it continually searches for new products that it can produce and offer in different markets. Managing a technology-based business not only needs broader planning vision, but also the ability to use the same knowledge in different contexts.

The success or failure of each approach depends on its compatibility with the emerging competitive challenge at the time. Because competitive games change over time, they force companies to switch their strategic emphasis from one dimension to another. Since each strategy has organizational implications in terms of authority and responsibility, changes in strategy require changes in the organizational structure and the dominant culture of the business. As competition intensifies and the changes become more frequent, this unidimensional approach becomes more ineffective. Shifts from one dimension to another in search of the effective competitive base cause organizational turmoil and strategic confusion. The waste and frustration associated with periodic restructuring have necessitated a search for alternative solutions.

Interactive systems architecture uses product, market, and technology in an interactive mode. It recognizes the necessity for achieving competitive advantage in all three dimensions of market, product, and technology (Figure 9.7).

The objective is to capitalize on the totality of the value chain and actively generate synergy among the three dimensions. An interactive architecture is based on managing the interactions among technology, product, and markets.

The emerging multidimensional modular structure, with a different business model and reward system for each dimension, dissolves the need for subordination or suboptimization around any one business dimension, eliminates periodic restructuring, and shifts from one dimension to another in search of an effective competitive base.

In an unpredictable, turbulent environment, the viability of any design depends on its capability to explore and exploit emerging opportunities all along the value chain. These opportunities, which emerge out of interactions among technology, product, and markets, remain inaccessible to unidimensional cultures and architecture. A platform that identifies the basic dimensions of the multicultural architecture is the starting point from which the initial elements of the value chain will evolve. As the business gains maturity and stability, new elements will be added.

In formulating the strategic intent of an enterprise we must remember that competitive advantage is a dynamic and relative phenomenon; it is different in different contexts and in reference to different classes of users. For example, the element of time, as a basis for competition, is not the same for supermarket shoppers at 10 a.m. and those at 5 p.m. Often the former are spending time, whereas the latter are interested in saving time. What might be an advantage in one context could be a disadvantage in another. This explains the success of convenience stores despite their higher prices.

Now recall the concept of the experience curve, which is volume-based learning. An important outcome of learning through an experience curve is process control. The volume-based control system is usually achieved in a given context after using considerable time and resources. When this context is changed or even modified, the whole knowledge that was gained in controlling the particular process is lost. Now assume that, somehow, we have developed the ability to make this know-how transferable to different contexts and applications without having to go through the time- and resource-consuming notion of the experience curve each time. If we can really do this, then we have created a core competency, or an unmatched opportunity for competitive advantage. Formulating and developing state-of-the-art knowledge that can be operationalized in different contexts such as process technology, is, in my opinion, a much more profound way of formulating a competitive strategy. For example, adding digital technology to miniaturization, as new core competency, has provided Sony with a whole new dimension in its competitive strategy.

Strategic intent can be formulated as a core competency. In this context a core competency is the attribute of the organization as a whole; it cannot be housed in a single division. Creation of the ability to transfer knowledge to different contexts requires multiple sources of learning and application.

To serve its own interest in an exchange, a purposeful system ought to have something to give back to its environment. Therefore, it needs access to a group of potential users with purchasing power who have a need or desire for the product or service it can produce. A true customer, in reality, is a pain in the neck. He/she has something you want, and his/her satisfaction, in competition with other sources of supply, is the price you have to pay to get it.

Selecting a product-market niche is the first step in defining an enterprise's function and designing its architecture. We need to answer the following questions: Whose problem are we trying to solve? What solutions are we offering? How will we access the target customers? And, finally, will the target customers have sufficient purchasing power to pay for this solution?

Competitive advantage is an attribute of outputs produced by an enterprise. It is a difference that makes a difference, for a given class of users, in affecting their choice. And it must be transferable to a value for the provider.

To select a desired product-market niche, it is necessary to differentiate the customer base. There are many different ways to segment a market. Each segment reveals something new about the nature of the market and the behavior of the target customers. Use as many criteria as you can to differentiate user characteristics and identify their purchasing habits. The most useful segmentation is the one that identifies:

Using the traditional bell-shaped distribution curve to differentiate the customer base assumes that the majority in the big category alone defines the market (Figure 9.8). The behavior of this single group, considered normal, is used to determine the desired properties of outputs, while behavior of the smaller groups, considered nerds, are conveniently ignored as insignificant.

Struggle for a share of this monolithic market leads to a game of intense competition between big powerful players. However, increasingly it seems that the nerds are taking over, and the bell-shaped curve is somehow flattening out (Figure 9.9). Targeting and understanding behaviors of the smaller groups that are emerging rapidly provides the best, and sometimes only, window of opportunity for a new player to successfully enter the game and skillfully avoid the intense competition at the early stages of the entry to a new market.

Traditionally, organizational theory deals with two types of relationships: (1) responsibility (who is responsible for what) and (2) authority (who reports to whom).

Structure, so conceived, can be represented by a two-dimensional chart in which boxes represent responsibilities and levels and lines represent the loci and flow of authority (Figure 9.10).

The criteria used for dividing the whole into areas of responsibility, and for determining their relative importance (line of authority), represent the major differences among organization theories. These criteria, not surprisingly, have evolved primarily around three components of a system: input (technology), output (products), and environment (markets).

Depending on the nature of the competitive game at any given time, a priorities scheme is used to designate one component as primary and the other two as subordinates. For example, when the ability to produce was the defining factor of competition, input became the primary concern rendering market and product subordinate to manufacturing. The marketing era saw the shift of emphasis from production to market and thus subordination of manufacturing to marketing. It seems that a self-imposed unidimensional concept of organization has prevented the realization of a multidimensional alternative.

For a majority of designers, the unidimensional mode of organization based on structurally defined tasks, segmentation, and hierarchical coordination of functions seems the only acceptable way of organizing work. A predominant management culture continues to value command and control very dearly and considers any form of variety in the organizational structure unacceptable, wasteful, and at best impractical.

As discussed earlier in this chapter, the multidimensional structure assumes that the three common criteria — input (technology), output (products), and environment (markets) — are complements. Treating them as interdependent dimensions and managing their interactions eliminates the need for periodic reorganization when a change in the competitive environment necessitates a change of emphasis from one orientation to another, for example, from products to markets, or vice versa.

The viability of any organization depends on its ability to adapt actively to the changing requirements of the emerging competitive game. The ability to adapt requires some form of flexibility and responsiveness, which in turn demands that some degree of redundancy be built into the system.

A modular structure embedded in a multidimensional scheme can achieve the required level of flexibility to create an adaptive, learning system by shifting its attention from micro-managing the parts (power-over) to macro-managing the interactions (power-to-do).

Power-to-do is what organization is all about. It should not be confused with power-over. Power-to-do is the foundation of organizational potency and duplication of power, whereas power-over is about domination. Potent organizations are not built on impotent principles. Power-to-do multiplies when it is duplicated in special purpose modules. These modules enjoy considerable freedom as long as they meet the interface and functional requirements of the larger system of which they are a part. In any system it is differentiation that keeps the system alive and potent. Organizations that differentiate and integrate create real value for themselves and others.

Multidimensional, modular design, in my experience, is the most practical means of handling complexity and uncertainty. It makes it feasible to implement a complex design without getting lost in the process.

Multidimensional modular structure consists of a set of distinct, but interrelated, platforms. Each platform represents a dimension of the system signifying a unique context, mode of operation, and predefined set of performance criteria and measures. Each platform hosts a set of special purpose modules with the same set of behavioral characteristics. Relationships and the interfaces among platforms are explicitly defined, integrating them into a concept of the whole. Parts operate as independent systems with the ability to be relatively self-controlling and act as responsible members of a coherent whole with the ability to respond effectively to the requirements of the containing system.

For example, a technology platform provides a friendly environment for component builders. Component builders are modules that usually host a core technology and therefore require a different mode of management and performance criteria and measures than those necessary for managing and controlling the marketing modules.

The organization, so conceived, then becomes capable of expanding or contracting by the addition or deletion of replaceable modules that have the means of vertical and horizontal interactions. The resulting mode of organization is capable of redesigning its structure and redefining its functions, allowing it to exhibit different behavior and produce different outcomes in the same or different environments. This means work can be organized in a variety of ways, and indeed, an organizational choice does exist. Figure 9.11 is an outline of the multidimensional modular design.

The essential characteristics of a throughput and organizational processes were discussed in Chapters 5 and 6. Recall that throughput processes are those directly concerned with the actual output of the organization. Organizational processes, on the other hand, are concerned with creating integration, alignment, and synergy among the organization's parts. In this context, the planning, learning, and control system is an integral part of designing architecture.