Minimal IT - In defence of complexity

23 October 2007

In defence of complexity - part 2

The strongest arguments against simplifying IT architecture involve the role of IT in the organisation. We have to decide whether the arguments for simplification or the arguments for complexity will win.

Continuing from last week, there are subtle arguments against the simplification of IT architecture into independent systems that are aligned to organisational responsibilities:

Business is not rational. Any attempt to implement a simple architecture will fail because business structures are not clearly defined. (The pro-simplification response is that irrational business does not necessarily require complicated architecture. A "creative simplification" of IT into independent systems is a good model even if business is not that rational.)
Although most business managers are intellectually capable of managing IT, the intricacies of IT mean that there is a value to specialisation. Specialists will inevitably build architectures that make most sense from their point of view.
IT is a demonstration of political power and capability, not just a rational response to the opportunities of automating the storage, processing and communication of information. All organisations have to show off, to impress customers, investors and employees. IT is a good candidate for this exhibitionism, and major cross-divisional systems have political benefits that outweigh their practical disadvantages.
IT delivers value precisely because it does cut across existing business structures. Misaligned, shared IT provides opportunities for creative dissonance. Rational, simple IT might be more efficient in the short term, but in the long term chaotic and overlapping IT leads to a more vibrant and successful business.

We have now come to the end of my heretical journey. I have described the arguments in favour of a radical simplification of IT architecture, and the arguments in favour of retaining current complexity. We need to ask whether the arguments for complexity will continue to hold sway, or the arguments for simplification will overcome.

This is not a silly question about predicting the future. It is a serious question about what direction we want to take.

We can stick with the status quo, and continue with IT architectures designed for the convenience of those who provide IT. We can continue to use IT as a political tool for driving business change. We can continue with our current organisation and management of IT. Our big IT management problems will stay the same: we will battle to deliver major IT projects, fight legacy systems and struggle with skill shortages.

Alternatively, we can radically simplify our IT architectures to create truly independent, decoupled systems that are aligned to organisational responsibilities, and which do not share technology layers with other systems. This will keep IT focused on simple information automation, and out of organisational politics. It will, in time, dismantle our current IT organisation and management, and merge IT into general business management. It will reduce major project failures, the decline into legacy, and skill shortages.

I have shown you my vision of where we could, and should, go with IT. Do you share this vision? If not, what is your vision for the future of IT, and what are you doing to make that future a reality?