In the next layer down comes the business requirement to trade in differentiated ways to suit different markets, and access a bigger overall share of the potential overall market. You can think of it as how Volkswagen Audi Group, VAG, uses multiple brands: Audi, Volkswagen, Skoda, SEAT, to get more coverage of the overall market for cars and increase its share. In real life it’s likely to be even smaller granularisation such as positioning Volkswagen in a different way in Europe to South America, but you get the principle. Now think how the people and devices in the personalisation layer have to be flexibly connected to the differentiation layer? Extend that to think about this being all about the orchestration of clusters of ‘services’ rather than a handful of monolithic applications.

That’s going to be tough to manage and provide governance in this continually changing environment where change and flexibility is the key criteria for business success. It also has to connect in a reliable and policy managed manner to the third layer where the corporate processes that comprise the core competences of the enterprise. In the case of Volkswagen-Audi this is how they use common processes to design, build, ship, finance and service their products, in fact in many cases this is so well done that the same car platform supports multiple brands with a degree of differentiation over laying the common shared platform.

There needs to be a strong degree of stability here for an enterprise to operate effectively by optimising its core strengths. A point I covered in a previous post in which I emphasised the need for say twenty percent of an enterprise to be continually changing and adapting to the external market drivers in order to maximise competitive advantage, over eighty percent that should operate in a more stable manner that allows optimisation of resources and capabilities. Finally in the lowest fourth layer we have classic IT functionalities to record enterprise transactions and provide corporate compliance.

Okay now reflect on the provisioning for the two upper layers, a place of continual change based on people, and devices using services and not applications. Now reflect on why you would provision in the same way as for the classic IT around computers, and applications in a stable environment. It’s pretty clear that the only reasonable answer is that you wouldn’t do it the same way! Take some time to take a close look at the answer that the VCE coalition has come up with in its new V-Block technology. You can get behind some of the thinking by looking at two of their key technology leads views at Chuck Hollis’s blog and at Chad Sakac's Blog.

Does it strikes you, as it struck me, that we really need to ask if provisioning with the current infrastructure optimised for applications, plus using some virtualisation, is going to be enough to handle a pure services environment with all that this means. V-Block might just be what we are looking for, and as a side affect it might answer a few questions about orchestration, and SOA issues too!