Subscribe / Unsubscribe Enewsletters | Login | Register

Pencil Banner

IT initiatives from recent years need to be made integration-ready

Rob Hailstone | Jan. 25, 2010
Integration processes have become as important as the functionality of the systems being integrated.

The recession has hardly slowed the flood of innovations that will impact IT. Consumer-driven innovations stemming from Web 2.0 vie with leading-edge research-driven technologies such as in-stream information processing. While each could be justified in a standalone sense, the overall impact will be to increase the complexity and cost of IT, and further dilute IT budgets. To deliver meaningful benefits, new solutions have to be integrated into the broader IT landscape.

We live in an IT world that is defined by integration processes

Integration processes have become as important as the functionality of the systems being integrated. For several decades the key integrator has been the human operator, but, although adaptable, humans are error-prone, expensive and slow.

Addressing the problem with smart middleware was an important positive step in increasing the predictability and performance of the integration process, but it suffered from complexity due to the number of endpoints and incompatible technologies. It is only in the last few years that the concept of build to integrate has become widely adopted, with the advent of XML-based standards providing a technology-neutral way of addressing system interoperability. However, there is a risk to this progress created by the introduction of further IT initiatives that are outside of this integration envelope.

We shouldnt expect or wish the stream of innovations to stop, but the next half decade is more likely to be marked as a period of consolidation, when we will learn how to transform all of the pieces into interoperable components of a full architecture that jointly supports new business models.

There is no such thing as standalone

Many integration problems result from the fallacy of standalone systems. The truth is that for a system to be standalone it has to be addressing a requirement that has no part in any business process and hence is irrelevant and unnecessary. For example, the Web was treated as a standalone marketing issue for the early years of its growth. Only later did we recognise the need to integrate it with front- and back-office systems, with the ensuing additional integration work.

As we take on board further IT initiatives, some of them may appear initially to be standalone. In particular, some of the Web 2.0 collaboration technologies are regarded as side issues to the major business processes. In fact, they enable the critical processes of building and enhancing relationships within and between organisations and their ecosystem of partners and customers. It should be obvious how to integrate the intelligence captured by blogs and wikis into mainstream processes but it isnt. As they have not been designed to be integrated, the information trapped in these new silos is lost to the organisation, while it should be driving dynamically optimised processes.


1  2  Next Page 

Sign up for Computerworld eNewsletters.