New Book Review: "The Service-Oriented Enterprise"

New book review for The Service-Oriented Enterprise: Enterprise Architecture and Viable Services, by Tom S. Graves, Tetradian Books, 2009, reposted here:

At the outset, Graves explains that "everything in the enterprise is a service. Everything the enterprise does is a service. Everything the enterprise delivers is a service – even products are services, in a sense. And the enterprise itself is a service. Everything is a service. That's the key idea behind the service-oriented enterprise: a view of the enterprise in which everything is seen in terms of services and their interactions and interdependencies, providing consistency and simplicity everywhere, and creating new space for agility and innovation in the enterprise." It is important for the potential reader to understand the definitions that the author explains along the way within this text, because like his earlier work, "Real Enterprise Architecture: Beyond IT to the Whole Enterprise" (see my review), he provides his own interpretation of industry terms, most of which add value to his discussions. And unlike some other enterprise architecture texts (within the space of which few are currently available in the marketplace), these interpretations are based on the author's personal experience in the field. In explaining his goal for this book, the author mentions that "although there's a fair amount of theory, the keyword here is practice: the aim is to give you something that you can use".

While still presenting a substantial amount of theory, this reviewer thinks that this goal was largely met. As a consultant architect, this reviewer realizes both that heavy use of examples is bound to greatly increase the size of the text, and that materials not generally available on the bookshelf (i.e. through training such as workshops and conferences) can be relatively expensive, and the general practice of Graves to write expanded white papers such as this fits a niche in this space that provides introductory enterprise architecture tooling. This text is divided into three core sections: (1) a review of key terms such as "enterprise architecture" and "service-oriented architecture", as well as some of the core metaphors underlying the service-oriented enterprise, (2) a description of core ideas concerning both the structure and relationships between services, as well as their relationship to the structure of the enterprise, and (3) an illustration of application of the principles outlined in the second section. The bulk of the book resides in the latter two sections, with more weight allotted to the second section on principles. Graves also provides an appendix on building a "function model" in Microsoft Visio, which the he presents in the third section on "practice" as "a layered list of business functions – a visual summary of what the enterprise does in functional or service terms".

As in his earlier work sited earlier in this review, the author provides the reader with straightforward terminology. During this initial section, Graves indicates what other recent authors have noted in this space, such as Jeanne W. Ross, Peter Weill, and David C. Robertson in their "Enterprise Architecture as Strategy: Creating a Foundation for Business Execution" (see my review): "It's still a new discipline: the term itself was coined barely a decade ago." Graves furthers his reputation for providing abundant sidebars throughout this book, a practice that this reviewer highly encourages him to continue in future efforts. Many of these sidebars consist of case study snippets. The first case study reminds this reviewer of one of his recent clients, for example. What is meant by "architecture" in the context of the "enterprise"? Graves explains that "an all-too-common answer to that last question – typified by the comments of the architecture manager at a recent client – would be 'IT functions and data'. When I explained that the business of her employer – a government department in the social-services sector – was a great deal wider than that, and hence her architecture needed to cover people-based processes, legislation and a great deal more, she replied, 'Yes, I agree with you: architecture is all about IT functions and data'. We went round that loop a couple more times before I gave up: there wasn't much point in trying to discuss real enterprise-architecture with someone who wouldn't allow herself to hear about any aspect of business beyond her own domain of IT. Oh well."

Another aspect that this reviewer particularly enjoyed was the author's discussions of "machine and organism" and "services in the living enterprise", in which Graves derails the common business metaphor of business as a machine that was implanted in the psyche of the business world by Taylor's scientific management outlined in texts such as Christopher D. McKenna's "The World's Newest Profession: Management Consulting in the Twentieth Century" (see my review). Whereas in the machine metaphor, where all of the components that comprise an enterprise must be carefully put into place after being carefully designed and constructed, and every task and resource (including people) have specific functions, the organism metaphor describes the enterprise as being "composed of a myriad of interdependent services, working together, as a community, towards a common purpose", much like cells within the human body, similar to how David A. Taylor described object-oriented technology in his "Object-Oriented Technology: A Manager's Guide" (see my review), although concentration on low-level objects can easily lose sight of the big picture. In his comparison of the entity model where there exists a hierarchy between purpose, brain, and brawn, and the factory model where there exists a hierarchy between owner, manager, and worker, Graves explains that it is "easy to become distracted by the fact that within each cell, there will be many subsidiary services, and often many instances of a particular type of subsidiary service, yet just one nucleus, one 'brain'". 

En route to the chapters on practice, the author provides discussions on services and service relationships, viable systems, delivery services, management services (that comprise policy, strategy, and direction), coordination services, pervasive services, and properties and patterns. While some of the content in these chapters exists in other texts, what this reviewer enjoyed was seeing commonly covered topics written with a refreshing perspective combined with new subject matter not seen elsewhere. For example, in his discussion on service completeness, Graves mentions that "given the natural recursion of services, we know that enterprise viability will depend on the presence of all the support-services, and on the effectiveness of the interfaces between them. But to re-use services – that key concept of loose-coupling – we also need to be able to split them apart, to make services temporarily 'incomplete'. Viability depends on completeness; re-use depends on incompleteness." As is the case throughout this book, the author provides abundant, relevant diagrams throughout, and this reviewer especially appreciated diagrams such as the one in the chapter on service purpose that visually depicts a modified Business Motivation Model (BMM) comprised of relationships between vision, role, mission, and goal.

After presenting the Functional Business Model that describes the enterprise in terms of business functions, and providing industry examples from manufacturing and government within the social services sector, Graves outlines application themes such as identifying business systems, identifying single source of truth, mapping valuations to the Functional Business Model, identifying viable relations, identifying viable communications, and determining process choreography. Well done text written in plain English that is recommended to anyone new to enterprise architecture or seeking to reexamine their approach to enterprise architecture engagements, especially when working with clients new to this space.