To print: Click here or Select File and then Print from your browser's menu

This story was printed from silicon.com, located at http://www.silicon.com/

Story URL: http://comment.silicon.com/0,39024711,11032703,00.htm

Devil's Advocate: Buy anything but the market leader
Radical advice?

By Martin Brampton

Published: Tuesday 16 April 2002

No one ever got fired for buying the proven, market leading product, right? But did any company ever really make much progress by doing just that? Columnist Martin Brampton gets all contrary...

When I recently considered how to spend the government's IT budget (http://www.silicon.com/a52351 ), one reader made a number of positive suggestions, including an emphasis on open systems. But before we think again what to buy, let's consider how we get into such a mess over the buying of technology.

Maybe each generation has to learn afresh. It is odd, though, that in this age of graduates who have studied information technology, we should have lost the experience of the past. Many of those who lived through IBM's dominance in the seventies and eighties came to regret the view that "nobody got fired for buying IBM".

Nowadays, they continue to look to IBM as a valuable source of technology and services but one question is always prominent. If we commit to this way of doing things, which other suppliers could we turn to for an easily substituted alternative? IBM has become a more valued corporate citizen as a consequence of the renewed scepticism of its best customers.

Yet a new generation seems to have bought into Microsoft up to and beyond the point where market dominance shows its downside. If there are now sectors of the market where competition is hard to discern, buyers have only themselves to blame.

It is commonplace to suppose that our economic system epitomises the free market. But companies, in their nature, do their utmost to eliminate competition. Everything to do with brands, pricing and technical standards is designed to reduce the threat from rivals. What could buyers do to invigorate competition?

Purchasers of unit trusts are notorious for choosing trusts that have performed well in the recent past. The most spectacular case was the mountain of money invested in technology trust ISA contracts just before the collapse of most technology shares. It is well known that a better strategy is to buy into trusts that have been performing poorly, since over a long enough period, the performance of individual trusts does not diverge a great deal.

Spending money on IT, the corresponding strategy is to do exactly the opposite of what most buyers do today. Instead of always trying to buy the market leader, one must always aim to buy anything but the market leader. This will not always be feasible, either because there is no viable alternative or because there has to be some consistency in technology purchasing.

The strategy can also go wrong for another reason. Just as you have built up an investment in an innovative company that is clearly not the market leader, you find that it has been taken over by the market leader. Go back to square one!

Can open source rescue us here? It certainly has some very interesting characteristics that can be usefully compared with the internet. The roots of the internet were, as you probably remember, the ARPAnet, a network designed to be resilient in the presence of multiple failures. Successful open source projects are similar. There is a network of enthusiastic developers, none of whom are irreplaceable.

Competition in open source is primarily for the achievement of technical excellence, and when it succeeds, new standards are established. And they can be immensely valuable for buyers. The old open systems movement was based on the idea that Unix would eventually be standardised. It never happened.

Linux has actually brought into being the standardisation that years of negotiations failed to achieve. As all the major Unix vendors move to make their systems capable of running Linux applications, so buyers have a new option. Choose the Linux version of applications, then run it on whichever Unix system is most appropriate.

The amount of resource that needs to be spent on new software development is a relatively minute part of the whole IT business. But it has a quite disproportionate effect on standards and competition. If buyers look all the time to promote competitors, to leave themselves viable alternatives and to support movements that are independent of the big vendors, they are likely to reap valuable rewards. Forget about market leaders.