Commentary: Less IT is more

Gordon Brown plans to cut costs in the U.K.'s Civil Service. Apparently IT plays a role in his plans, and he talked in the budget of substantial investment in systems. But is more IT really the way to gain efficiency, or might less IT be the solution?

Back in the nineties we had an economic paradigm shift that brought a marked improvement in productivity through the deployment of IT. Then paradigm shifts went out of fashion and most of the dotcoms went bust, losing vast sums of money for hopeful investors. It transpired that much of the new economy related to low tech, often low paid jobs in service industries.

Just throwing out software would be to lose the baby with the bath water. Certainly, there are interesting examples of gains from technology. If the high street banks still needed the same ratio of staff to transactions as they did fifty years ago, the entire population would now be working for banks. That would not be good for the economy.

The trouble is that software suffers badly from feature creep. Vendors like to introduce new features so that they can sell upgrades to existing customers. Sometimes the customers want the new features. Often, though, many of the customers do not need the added features, but are forced into the upgrades to protect support services or to maintain technical compatibility.

Bloated software is the result. It is inefficient in its demands on hardware resources and its users may find that the time taken to understand the features far exceeds the eventual pay off. I know that I have struggled for hours with esoteric software features, not because of any great gain in my efficiency, but more because I hate to be defeated by a computer program.

Indeed NHS computing seems to be recognising the reality of software bloat. Part of the argument for price reductions on widely used software has been to turn the vendors' argument for valuable features on its head. The NHS is asserting that it wants to pay only for the features it uses, and the vendors can take out the rest if they choose.

Email seems a prime example of software run riot. The ability to send a simple message quickly and cheaply is beyond doubt. Why does anybody need to use HTML in email though? It makes messages far longer for a medium where content ought to prevail over presentation. Why for that matter are attachments needed? Surely, they would be far better handled by making files available on demand rather than pushing them willy-nilly to recipients. Without HTML and attachments, we would evade most of the billions of pounds allegedly incurred every time there is a big virus scare.

Moreover more than half of all email is now unwanted advertising. Probably half of the rest is from mailing lists that people follow with varying degrees of interest. And of the remainder, much is unnecessary copies by people seeking to cover themselves. Perhaps we need a system where the sender has to pay the receiver a penny for each five thousand bytes of message. Spam would be wiped out at a stroke, as would much fancy formatting and most of the boring legal disclaimers on the end of messages.

Another area of excessive complexity is the so-called personal computer. There were advantages in escaping from the rigidities of mainframe computing, but by now we seem to have thrown most of them away in favour of new constraints. If personal computing means making information accessible then it is to be welcomed. But if it means locking the individual to a cumbersome piece of hardware, then it is a severe handicap.

Many organisations are anxious to escape that link. Breaking it is a prerequisite for efficiency gains through hot-desking. Perhaps more significantly, it is also desirable to free up movement in order to encourage groupings of people to change and adapt to circumstances, unencumbered by technology constraints. And most IT departments waste much time and effort simply managing moves and changes.

It seems that Marks & Spencer has adopted some of these principles. It has cut down the number of applications available to its staff. And it is using Citrix thin-client technology to avoid many of the issues of desktop deployment. Unfortunately, the Citrix approach is still much more complex than one would like. To succeed with it takes considerable determination, especially while much software is written on the assumption of a dedicated personal computer.

There are still gains to be made from inventing and using new software. But if it is efficiency we are after, maybe the emphasis should really be on reducing reliance on IT and removing unnecessary complexity from the IT we decide to retain. After all, IT is all about abstractions and the best abstractions are the simplest.

Martin Brampton is founder of Black Sheep Research, an independent consultancy providing research, writing and speaking services on a wide range of business and technology issues. Martin was previously a director at Bloor Research, and has worked with IT as a user and analyst for over 20 years. He is a long-term contributor to silicon.com through videoed debates and his weekly column, which tackles a wide range of issues. He can be contacted through his website.