Gartner's new view of enterprise architecture provides a solid base for new research into the discipline and gives our clients insight into the best practices for building a well-defined, well-aligned enterprise architecture in a mature, productive enterprise architecture program.
Table of Contents
In December 2004, Gartner announced the acquisition of Meta Group — one of our closest competitors. On 1 April 2005, the transaction was consummated, and we welcomed Meta's vibrant architecture practice into our research community. One of our first orders of business was to unify our research positions. We convened a working group of analysts with a mandate to identify and resolve positional differences, with an eye toward maintaining legacy continuity for both sets of customers. However, we recognized that this was an excellent opportunity to advance our thinking in the enterprise architecture space, and we expected new dimensions and insights to be developed as part of the unification efforts.
Gartner and Meta always agreed, in broad strokes, on the principles of enterprise architecture. We agreed that it is a "top-down" discipline that derives the optimum constellation of business, information and technology to support the business strategy. We agreed that any "solution" required business, information and technology components to interoperate in support of business capabilities. We agreed that the future state architecture should be developed before the current state is documented, and that no architecture was useful unless it included a gap analysis and an actionable road map for getting from where you are to where you want to be. We agreed that "architecting" was only a small part of the job, and that much of an enterprise architect's time is spent strategizing, communicating, leading and governing. We agreed that good enterprise architecture does not happen in a vacuum, but that it is one of several strategic planning disciplines that organizations should practice to align their technology with their business strategy.
However, there were also some significant differences. Meta's approach to architecture focused more on the process, while Gartner's focused on the framework — the theoretical constructs used to organize an enterprise's thinking around the subject of architecture. We found that Meta's process carried with it an implied framework, whereas Gartner's framework had evolved an implied process. Reconciling these secondary products carried some challenges, because we had to flesh them out in more detail to understand where the similarities and differences were.
We believe that successful enterprise architecture programs are process-focused, so we started with the Enterprise Architecture Process Model. Our objective was to accurately reflect the activities associated with the initiation and implementation of an enterprise architecture program. The result is a multiphase, iterative, nonlinear model focused on enterprise architecture process development, evolution and migration, along with governance, organizational and management subprocesses. It represents a synthesis of best practices of how the most successful organizations have developed and maintained their enterprise architectures, and it is described in detail in "Gartner Enterprise Architecture Process: Evolution 2005."
With the process model defined, we turned our attention to the enterprise architecture framework. Enterprise architecture is a complex subject with abstract components — frameworks are important because they provide a context within which the organizational thinking can be structured, and consistent use of a framework in all components of an enterprise architecture program is a best practice. A good framework will define the components of an enterprise architecture and the relationships between them, providing the architecture team and the organization a set of shared semantics and concepts with which to describe their architecture. We adopted an aspect-oriented approach to our framework, deliberately compatible with IEEE 1471 — Recommended Practice for Architectural Description of Software-Intensive Systems. The 2005 Gartner Enterprise Architecture Framework defines a "business context" consisting of the business strategy and external trends that provide the overall context for the enterprise architecture. We advocate the development of a minimum of three interdependent viewpoints: a business viewpoint, which is concerned with the processes and organization of the business; an information viewpoint, which is concerned with the information that runs the enterprise; and a technology viewpoint, which is concerned with the hardware and software components that support the enterprise. The aspect-oriented approach allows for the articulation of additional viewpoints, should the organization require them. An important aspect of the 2005 Gartner enterprise architecture framework is the recognition that the "solution architecture," where the systems that support the enterprise are actually specified and designed, takes place at the intersection of the viewpoints. Gartner's new framework is described in "Gartner Enterprise Architecture Framework: Evolution 2005."
Although the process model and the framework have changed significantly, components of both approaches were valuable enough to keep. One such component is the legacy Gartner concept of "bricks" — those elemental technology building blocks. We felt that the life cycle management concepts that were embedded in the brick were worth retaining. Indeed, a similar construct existed within Meta's framework. Another construct that we retained is the architectural pattern — a concept that both organizations had written about extensively. Patterns are the best vehicle that we know of for communicating the guidelines of the architecture to developers and implementers. We explore the concept of patterns and how it fits into the new framework and process in "Enterprise Architecture Patterns: Combinations That Repeat."
No discussion of enterprise architecture would be complete without an articulation of how it fits in with the other planning disciplines of the enterprise. No enterprise architecture is effective unless it is implemented, and no enterprise architecture is implemented unless the vision of the future, guidelines and standards are embedded into the decision-making processes of the organization. We explore the relationship between enterprise architecture, IT strategic planning, IT portfolio management and enterprise program management in "Enterprise Architecture Improves IT Planning Synergies."
The research described above is just the beginning. Enterprise architecture is a wide topic area with much opportunity for additional research and commentary. Our new view on enterprise architecture provides us with a solid base for expanded research into all of the aspects of the discipline — and it provides our clients with valuable insight into the best practices for building a well-defined, well-aligned enterprise architecture in a mature, productive enterprise architecture program.