Our host, the information processing czar of a financial services firm, explained that the mainframe complex we were touring housed every conceivable datum pertaining to his company’s business… and the engines to manage it. The pod of mainframes and their enormous disk farms, he expanded, kept track of every client’s account, every employee, every item the firm bought, sold or owned. By consolidating all that information in one central location, the company’s executives were able to achieve absolute control over their domain. Any other computing scheme would fall short, we were assured; the company had studied various alternatives. This commitment to mainframes was central to his organisation’s success, our host continued, for it enabled executives to get the greatest productivity from their desktop computers, which were all linked to the central system. And just how many personal computers, we asked, does your company have? Actually, replied our host, with a perfectly straight face, we don’t know.
* * *
To do justice to our host, whose candour we found refreshing, it should be pointed out that his company’s lack of information about its personal computers had, as far as we could tell, absolutely no bearing on the soundness of the business. If anything, the little knowledge gap over which we stumbled may have in a roundabout way reinforced our impression that this company’s information processing department knew very well how to carry out its mission. Not knowing what it didn’t really have to know may have helped it focus all its talent (and computer power) on useful work. This isn’t usually the case, however. IBM’s decision vastly to increase its plant capacity during the 1980s offers dramatic evidence that ignorance can be costly. When IBM embarked on its construction spree, its reasoning was simple, clear and, as it turned out, daffy. It seems to have gone like this: IBM was losing business because it couldn’t respond to exploding demand for its large systems and peripherals. More output would require more factory space. So the company built vast plants and thereby created the capacity the high efficiency capacity – the future demanded. There were only a couple of errors.
By Hesh Wiener
First, the relationship between plant size and unit capacity was affected more dramatically (and in the opposite direction) by miniaturisation than by growing volume. Second, IBM believed it could continue to sell ever more large systems just because it had done so in the past. If IBM’s big enchiladas had all the facts at their disposal – and the wits to use them – they might have run their company differently, by which we mean not into the ground. By the time IBM tallies the rest of 1993, its losses for the past three years will be $16,000m, give or take a couple hundred megabucks. We do not wish to be unduly harsh, but we we must confess that we think the company could have done a bit better. For the brief period when more manufacturing space was needed, IBM might have rented facilities on relatively short, renewable leases. Where IBM identified dramatically changing components, it might have subcontracted the assembly of old technology while it figured out how to build the new. And if it had due respect for other participants in the industry – its partners as well as its competitors – it might have noticed a few developments that were attracting the attention of customers. Personal computers were expanding in power and versatility faster than any other class of processor. And the companies winning the desktops were selling their very fast, very capable machines for compellingly low prices and, by doing so, growing at a breakneck pace. Meanwhile, IBM’s large systems business, not uncoincidentally, was shrinking at just about the same rate. Plenty of people in IBM must have seen this, but they apparently didn’t include anyone who was in charge, or who could be heard by the bosses. Perhaps more importantly, if IBM used all the information at its disposal, it might still have embarked on its erroneous course… but corre
cted itself far earlier. The cost of its mistakes would have been monumentally less. Confronted by all the information that could have been distilled from its vast files, IBM’s executives would have been forced to act more wisely.
John Akers and his closest colleagues might have been embarrassed by their missteps, but they also might have held onto their prestigious jobs. And so, too, might some of the rank and file employees whose only present contribution to IBM’s welfare is an indirect one: they have added to the information processing requirements at the IBM shops run by government unemployment compensation departments around the globe. Yes, if they only used their computers wisely, IBM’s managers could have found out in a timely fashion where all the personal computers were, not only within IBM and the company whose mainframe centre we visited, but elsewhere, too. Of course it might have been possible for IBM to find all those personal computers anyway, even without its mighty databases, had it been a bit more observant. We don’t have the resources of IBM, but we found them anyway. They are standing where they have been for years: directly between IBM and its survival.
Copyright (C) 1993 Technology News of America Co Inc All rights reserved