I must admit it: I'm one of those miserable Londoners that, a few years ago, when he was starving himself in a box near the Thames, failed to take the opportunity to head down and roundly abuse David Blaine. It is indicative of a shameful lack of civic spirit on my part, and I've resolved to arm myself with a truck load of rotten tomatoes next time he is in town for one of his stunts.

Blaine's most recent effort, got me to thinking. He's a pretty polarising character at the best of times. But if you promise to do something – however unimpressive the sceptics may find it – it really is best if you stick to that promise, and don't dupe people, or argue over the semantic details of what it was that was promised.

I think the correct term for this is expectation management. It is a theme that I intend to come back to – from an IT perspective – over the coming months.

Blaine's escapades also highlighted another area I'm keen to explore: How do you know, for certain, what it is that your customers want? It may seem an obvious question, but without some form of feedback mechanism, that answer is invariably tainted by the questioner's own circumstances.

Specialise in elaborate, off-the-wall endurance feats? It's likely that the answer to what customers want will be "more elaborate, off-the-wall endurance feats".

Similarly, if your specialism is providing IT services at least some part of the answer is likely to involve a bit of technology. And indeed, unlike publicity seekers, the IT leader may console themselves with having some pretty concrete market research to guide their thinking, in the form of a business plan.

Nevertheless, while the principles of aligning IT strategy with business strategy, usually provide a sound footing for CIOs, it is not always the case. Business strategies need testing from time to time, circumstances change. And it is easy to slip in to the pattern of "we are doing things this way because it is how we've always done them." There may occassions where IT can show that the business strategy needs revising - giving the business want it really wants, without giving what it says it wants.

So when looking at the question of how we know what it is that people want, relying on that person to say so, isn't the whole answer either. Some of the other areas I hope to explore over the coming weeks and months, are techniques for digging beneath the surface a little, and really finding out what it is that the business wants – and how IT can help to deliver that.

IT has always had its fair share of detractors. Memories of bundles spent tackling the Y2K "menace", or the dot com "revolution" hardly help. But more corrosive has been a growing acceptance the IT has become commoditised, and no longer a source of competitive advantage. "All companies can invest in IT, and business applications are becoming homogenised, so IT has become just another operational cost" the thinking goes.

But not so, according to recent research reported in the Harvard Business Review, which looked at all publically-traded US companies.

"They found a clear correlation between levels of IT spending and a new competitive dynamic: Since the mid-1990s, when the rate of spending on IT began to rise sharply, the spread between the leaders and laggards in an industry has widened."

As the study points out, investment alone is not sufficient. And what that suggests is that it is the quality of the IT leadership that matters.

Instinctively, this appears to ring true. As IT enters an age of service orientation, old ideas about what is an application are breaking down. Instead, IT is becoming focussed on automating business processes, breaking down applications into component pieces that accurately reflect business process, and be rapidly altered to meet business needs.

This type of activity is a far cry from the notion that IT can simply be purchased as some sort of vanilla utility service, that provides as much competitive edge as electricity.