We know that areas of the business have been out investing in external technology. All of their SaaS solutions are now facts of life. Who will now pay to integrate the resulting portfolio?

We’ve had a number of discussions in the past few weeks about this issue, and by and large the approach being taken in the enterprises we’ve talked to seems unproductive. The assumption throughout seems to be one of two types: either that this is yet another “swing of the pendulum” and that these will all just be taken over on the next centralization wave (in which case, do nothing to help them in the interim); or that the fault was in IT’s management of the relationship and lack of alignment, so “let’s fix that and then everything can be made right”. Neither of these positions, though, will truly solve the issue even if they are carried through successfully: all these disparate piece-parts and external contracts will still exist, and still require integration.

Corporate IT, done internally, was about applications and packages. Now it is about the infrastructure of interconnection. Bridge building is the order of the day.

What makes this a difficult task is that the shape and nature of these bridges isn’t always clear yet. For the next decade we’ll be in various stages of migration to “more of our IT done as a utility, using utility components and SaaS-type solutions” from the traditional in-house installation. Inhibitors to that transition include getting over package myopia, the package vendors recasting their products for a utility setting (which means, first and foremost, abandoning most of their current licensing practices), and a fair bit of information architecture. There, in fact, is the first bridge to build: a transition strategy to clean up the master data model and the instantiation of it.

Another issue making this difficult is that most organizations are in the window now (2008-2010) where the depreciation of their major systems has now completed. These are the client/server-based solutions installed in the late 1990s (as the Y2K deadline loomed). At the moment, the successors are almost always going to be “more of the same”. As architects, the second bridge is a financial one: how do we make it possible to move into a utility model, with its expected lower cost profile, relatively early in the cycle, when it still forms an advantage for a period of time? Mapping out the means by which monies are managed during this transition is as important as the inter-application decisions.

Naturally, as components are contracted for out-of-house, it will be necessary to build the connectors to them to integrate with the portfolio in house. There has been a recent wave of business area purchasing and contracting that must also, in many cases, be accommodated.

All of these efforts are unfunded at the present time — and are likely to remain so. IT budgets for the next two years are unlikely to do more than barely cover increases in labour, benefits and maintenance — and in many cases not even that. Furthermore, this period of retrenchment will further increase pent-up demand once the budget situation eases in the next decade. The business, in other words, won’t pay for this — it has its own set of priorities — and the IT budget is unlikely to stretch easily to cover everything that needs doing.

But it must be done. This brings us to the final architectural bridge: how do you eat into the “run what exists” portion of the portfolio? This will be the source of funds you will need to tap. Your success at making that a priority — and today’s cut-down environment is a good time to discuss this and act on it — will set the tenor for all the rest.