The Agile Dance of Architectures

Viewpoint: February 4, 2004

John Seely Brown and I have focused for some time on the intersection of new strategic architectures and new IT architectures. We are convinced that this intersection will shape opportunities for significant strategic advantage. In fact, we believe that, far from diminishing potential for strategic advantage, new generations of IT architectures will actually increase the potential for strategic advantage. We ran into so much resistance when we ventured forth with this point of view that we decided to write a working paper to develop this perspective more fully.

In this working paper, “The Agile Dance of Architectures – Reframing IT Enabled Business Opportunities,” we discuss the opportunities created by a new strategic and a new generation of IT architecture. I have written about the new approach to strategy – which I call FAST strategy – in a previous web posting. I have also briefly discussed the importance of Service Oriented Architectures (SOA’s) – the new generation of IT architecture – in another web posting. Our working paper explores the interaction between these two developments since this shapes our view regarding the increasing potential for strategic advantage.

One architecture without the other is not sufficient to overcome organizational inertia. The two together, however, can create an engine to propel the organization forward. SOA’s will help to enhance organizational capability while FAST strategies will help to focus that capability and build momentum. Without the flexibility provided by these new technology architectures, this approach to strategy is very difficult, if not impossible, to implement. Conversely, migrating to these new IT architectures will be very challenging without the focusing discipline provided by FAST strategies. In short, it is the agile dance between these two architectures, in which each partner enables and shapes the other, that provides the power required to build significant strategic advantage.

FAST strategies focus simultaneously on two very different time horizons – a long-term horizon of five to ten years, creating a background for executive decision-making, and a much shorter-term horizon of six to twelve months, providing the foreground where operational and organizational initiatives play out. In this approach, the one to five year time horizon that has been the traditional focus of strategic planning processes consumes very little management time and attention – all the effort is focused on the near and distant margins of this traditional time horizon.

The key value of this strategic architecture is its focusing discipline. By defining a high-level view of the kind of business that must be created over a five to ten year period, it provides an effective framework for launching aggressive waves of near-term operational initiatives designed to accelerate movement towards this longer-term direction. Rather than dissipating resources across a broad portfolio of “experiments”, it mobilizes a critical mass of resources against a limited number of initiatives designed to rapidly improve operating performance (in contrast to financial performance, although obviously there is a relationship between the two). These initiatives are incremental in the sense that they are designed to yield near-term operational impact. They may or may not be incremental in the sense of supporting the previous trajectory of the business. When the business faces severe competitive challenges or rapid shifts in customer preferences, these near-term operating initiatives may represent significant departures from the previous trajectory of the business.

This focusing discipline and rapid waves of near-term operating initiatives mean little if the organization lacks flexibility. IT architectures (or lack thereof) today represent one the most significant barriers to flexibility. We are today locked in enterprise prisons, seeing opportunity (and threats) on the horizon, but shackled with architectures that traded flexibility for cost-savings. We have used IT to standardize and automate operations and information within the enterprise, but the discipline we welcomed in the name of efficiency had a price of its own. It gave us one-time operating savings, but then locked us into hard-wired platforms that made any major changes in business operations painful and expensive, if possible at all. IT architecture has become a choke point for operational and organizational initiatives.

Fortunately, service oriented architectures offer the potential to confront this choke point head on. At one level, service-oriented architectures have been on the horizon for quite some time now (the techies among you will immediately think of the promise offered by somewhat related component-based technologies like CORBA or COM). The name focuses on a fundamental shift in the view of software resources. Software has traditionally been viewed as functionality designed to support a specific business context and installed at the site where it will be used. In contrast, services are designed without knowing in advance the exact tasks they will be called upon to support. They are accessed when needed from wherever they reside.

Combine this concept of services with the notion of loosely coupled connections and you have the basis for a powerful, and much more flexible, IT architecture. Loose coupling is a design philosophy. It seeks to enhance flexibility by creating modules where all the information required to establish a connection with the module is described in the interface of the module. This approach delivers the full potential of the services concept. Think of each service as a module.With loose coupling, the services can be accessed and connections established without deep knowledge of the functionality embedded in the services. My earlier web posting on loose coupling went into more detail on the importance of this concept, not just in the technology world, but also in the business world.

There is one more piece to the puzzle required to make SOA’s really viable. We talked about information in the interfaces of loosely coupled services. If that information is presented in a proprietary language, it may be very expensive to translate, if it can be accessed at all. What we need are standardized ways of presenting this information so that it can become more broadly accessible. Web services technology provides this final piece to the puzzle. Using XML and a variety of derivative standards, Web services technology helps to make services broadly and much more inexpensively available. For a more in-depth discussion of the importance and business value of Web services technology, see my book Out of the Box: Strategies for Achieving Profits Today and Growth Tomorrow through Web Services

Of course, SOA’s are only beginning to emerge. But we are optimistic that they will rapidly spread given their attractive economics. These architectures effectively leverage large parts of the legacy IT platforms that are already in place.

What are the business benefits of SOA’s?
Well, we have already talked about enhanced flexibility. But that is only part of the story. The real benefit is strengthening the capacity for business innovation. Loose coupling in particular creates far more degrees of freedom for experimentation and innovation, both at the local level and higher levels of business activity. In particular, loose coupling expands options for reconfiguring business processes, not only the level of individual enterprises but across broad networks of enterprises.

Which brings us back to FAST strategies. Without the discipline of the FAST strategic architecture, enhanced flexibility and potential for business innovation can be rapidly undermined by lack of focus. It is the interplay of SOA’s and FAST strategies that offers the opportunity to overcome organizational inertia.

So, what does this mean for strategic advantage?
At one level, the basics of SOA’s and even FAST strategies are widely known and available. But, as we all know, there is a very long distance to be traveled between concepts and execution
. Few will make that journey successfully. Those who do venture down this road will soon discover something quite remarkable.

Strategic advantage will emerge from the waves of operational and organizational initiatives launched as part of a FAST strategy. Any individual initiative will be unlikely to provide significant strategic advantage, but the cumulative effect of multiple initiatives, all focused by a clear longer-term view of the destination and reinforced by a myriad of refinements in practice at the local level, will be much more significant.Advantage will come from path dependence. Companies will not be able to copy the end-state because its performance will depend on countless subtle practices – often that cannot even be articulated – shaped by the specific path taken.

To borrow another metaphor from our working paper, the continual interweaving of SOA’s and FAST strategies will produce a unique organizational DNA that will re-shape mindsets and organizational culture. These two architectures are ultimately learning architectures. Properly deployed, they provide capability for rapid learning at multiple levels. To use a phrase of John Seely Brown’s, they provide a context for “creative abrasion across specializations” and the mechanisms for rapid integration of the learning that results into the operational performance of the business. Companies that master these techniques will become moving targets that will be hard to track, much less copy.