At the recent Effective IT conference in London, industry sceptic Nicholas Carr kicked off proceedings with a modified version of the attack on IT in his book, IT does not matter. Jyoti Banerjee assesses whether the new version stacks up in the light of current IT experience.
(Right at the outset, let me just say thanks to Brian for inviting me to contribute to his blog. We've known each other for what seems like hundreds of years, but I doubt either of us would let a happenstance like that affect our opinions of each other's views. Our contrasting experiences and backgrounds - US versus European bases, Texan versus Indian backgrounds, etc - should help give this site a different twist on your screens).
Back to the matter at hand....
It was the Nobel Prize-winning economist from MIT, Robert Solow, who quipped that you could see computers everywhere except in the productivity statistics. In writing his book, Nicholas Carr was simply following in Solow's distinguished footsteps by attacking the value business has got from the massive investments it has made in IT. His argument was that IT does not provide competitive advantage to a business as there is little or no differentiation in the bulk of the technologies used by modern businesses.
Those for and against the argument have had a field day in the IT press, with both sides claiming victory through case studies and stats that prove their point and disprove the opposition. With Carr given the opportunity to rehearse his arguments in front of an audience of IT professionals in London, it was clear that the heat has not gone out of the debate, though one wishes for a little more light all around.
To differentiate or not
Let's explore the debate a little further. One could ask whether it matters that 70-80% of all IT is undifferentiated. After all, the implication could well be that at least 20-30% of all IT is strongly differentiated. Although the two statements are two sides of the same coin, they actually present quite different arguments.
Take the example of the car industry. Its products are largely undifferentiated (they have four wheels, four brakes, a steering wheel, a roof, an engine, etc) but the industry has created huge brand differentiation by focusing on the bits that are different (front-wheel versus rear-wheel versus all wheel drive, for example, or 50mpg versus 35 mpg, and so on). Although 70-80% of two cars may be undifferentiated in terms of the raw materials, there is no question that the buyer can distinguish a Ford from a BMW, a Hyundai from a Honda. is there really a difference between those cars that would justify their different brand strengths or company profitability? How is it that a company like BMW can go to the same suppliers as everybody else and yet deliver unmistakeably different brand performance in the marketplace? Maybe the differentiation is all done in the few percent of components that are actually different from one car to the next.
To me that would make sense in computing terms, as well. We might all use the same computers (undifferentiated) but the few that do smart things with their computers, or build smart processes around them (differentiation), could perform vastly better than the others. So to me, the case of differentiated or undifferentiated infrastructure, as presented by Carr, is not one that makes me say anything but "so what."
Hagel and Brown in their book The Only Sustainable Edge offer an interesting argument that the secret to competitive advantage is the relentless building of distinctive capability, within an organisation, plus in the networks it operates in. The implication of offering distinctive internal capabilities is that organisations have to choose what they are going to excel in, as it is not possible for any single organisation to excel in everything. By choosing to focus on what the organisation does best, it becomes mandatory for the organisation to then seek external partners who can provide world-class capability in those areas the organisation is not distinctive in. Clearly, the outsourcing impetus comes from those organisations that audit what they do and decide that in a number of areas they cannot compete with those that offer world-class capability. Instead, they join together in networks with those who can plug their gaps.
In effect, the Hagel / Brown argument would support Carr's position that a company should focus on what it is good at, and leave the rest to others who are better equipped to deal with those issues. Outsourcing started with infrastructural issues, such as IT and facilities. It has since progressed to cover horizontal activities such as accounting and payroll. Today, companies outsource things that a few years ago would have been regarded as core to their operating processes. Why should IT be any different? Why should IT professionals hang on to processes or skills within the organisation when their own competences are best employed elsewhere? In this sense, Carr is absolutely right: why should IT see itself as something special when it probably isn't, and should be handed over to a partner more competent in delivery.
Where I hesitate to hang my hat on the Carr coat-rack is in the area of utility computing. In becoming an advocate of utility computing, Carr is making it difficult for others to buy into his argument. The reality is that utility computing is still too new and too immature to be the mechanism by which enterprises can exploit quality world-class IT infrastructure. While many of the products already exist to enable utility computing, the two big gaps right now are in hardware and in process management.
Hardware
There are just not enough server farms around right now to allow utility computing to fly, despite huge attempts by all and sundry to build them fast. This is obviously a big enough gap that Microsoft has decided to spend a large part of its $35 billion cash pile on server farms in every location around the world they can find a big enough source of electricity. As these server farms come online, the hardware argument against utility computing will go away. Till then, there are no enough utility computing providers who have world-class capability to deliver the sort of infrastructure that tens of thousands of enterprises will need.
Also, all the server infrastructure cannot protect you from calamity when the evid day comes. Witness the outage suffered by MySpace due to record-breaking heat in Los Angeles where its data servers are held.
Processes
Of course, any new product can introduce change, even revolutionary change. More importantly, it takes time to deliver widespread change. It takes time to configure processes, enterprises, and now networks of enterprises together in such a way that the resultant meld of processes delivers competitive advantage of the sort that shows up in an economist’s productivity statistics. It took about thirty years before the impact of electricity could be measured across an economy because it took that long to figure out the best way to re-configure businesses in such a way that they could exploit electric machines, electric processes, etc.
We are seeing the same kind of reconfiguration taking place around digital processes. Eventually, that reconfiguration may well encompass utility computing as well. Till then, I can live with largely undifferentiated IT infrastructure if it allows us just a tiny room for innovation. Because that little margin of differentiation is often enough for people, smart people, to build competitive advantage.
That’s what the entire discussion about IT and its impact on business boils down to: People. Preferably, smart people. Now there’s an idea with legs….
