Rather than simply carping, we do try occasionally to offer some constructive suggestions on how things that are clearly broke might be fixed, and this week’s news of IBM Corp’s planned radical restructuring immediately called to mind a piece that appeared in this space a couple of years ago, back in December 1989 (CI No 1,326). Headlined The JCN Solution That Would Enable IBM to Survive and Thrive in an Open World, it appeared at a time little different from now, noting that while John Akers seems to be endlessly and optimistically waiting for something to turn up when all the signs are that Europe, which has saved the company’s bacon up to now, is about to turn down at the same time as the US market goes into a more emphatic decline, he gave the impression there was nothing IBM management could do to improve the company’s terms of trade.

Grew fat

The problem, it went on, is easy enough to expound: IBM grew fat in the days of proprietary operating systems and vendor-specific hardware – so fat that it became by far the most successful computer company in the world, largely owns the proprietary mainframe business, and has an ample slice of the proprietary small systems business. But that world is now in terminal decline, the number of proprietary operating environments that will be a serious force in 1995 can be counted on the fingers of one hand. IBM can’t afford to throw its own proprietary environments away because they represent far too large a part of its business as well as the most profitable. And there is a vast – if very finite – army of customers that want exactly what IBM has always offered. It suggested IBM offered – and offers – superb service and its products are generally of a significantly higher quality and reliability than those made by most of its competitors. That costs. But it’s a price that the majority of IBM’s longstanding customers are prepared to pay. But those customers are no longer large enough in number to provide IBM with both its bread and butter and its jam, and the fundamental changes that are seeping up from the desktop to threaten even IBM’s hitherto rock-solid mainframe business. The fastest-growing sectors of the market simply won’t wait for every last IBM manager to sign off a new workstation or a new personal computer before the thing can be launched: we have seen over and over again that all that leads to is competitors being two steps ahead in terms of price-performance by the time IBM is satisfied that the thing really does meet its exacting manufacturing and reliability standards – and the vast majority of personal computer and workstation users don’t want products designed to be as good after 25 years as they were when they were first plugged in – certainly not if they have to pay for that extra quality. That was – and still remains – the core of IBM’s problem.

Drop the IBM name

Yet, the piece went on, IBM can’t just throw all its hard-won reputation for quality and service overboard overnight: what can the company do? The answer is quite simply to put mass-market personal computers and Unix workstations – everything that is non-proprietary – into a wholly-owned subsidiary under entrepreneurial managers brought in from outside and paid solely by results: do properly what IBM so nearly did when it formed Entry Systems Division to enter the personal computer market until it got cold feet and brought it back under the leaden hand of the Armonkeys. But the most crucial and revolutionary step has to be to drop the IBM name for the new subsidiary. Well IBM is now proposing not one subsidiary for open systems, but a whole string of proprietary and open systems subsidiaries, and the idea of bringing in managers from outside to run the open end of the company seems still to be a little too revolutionary – but as we reported yesterday, the company is mulling a plan to get personal computers made in Taiwan and offer the things under a new name that would bring IBM to mind. Our suggestion was to go for the next three letters in the alphabet and call the new open busines

s JCN – fit words to the letters later in the time-hallowed IBM practice of rewriting history when it comes to nomenclature, we suggested. JCN would have the best of both worlds: it would have access to all IBM’s plants and components around the world, yet would make no overt claims to offer IBM quality, reliability or service: indeed the last would have to be as cheap and as cheerful as possible. Yet everyone would know that it was IBM under another name, so that it would retain most of the benefits of the IBM name, and in the case of personal computers, its Micro Channel models would be more or less the same as PS/2s, although not built to such rigorous standards and of course it would be free to offer as many AT- or EISA-bus models as it chose if that was what it found the market wanted why throw that business away just because the semi-detached IBM wants to establish a new standard? IBM PS/2s would cost significantly more than JCN’s machines, but would be supplied with better service than IBM presently offers on personal computers, and instead of being schizophrenic about OS/2 versus MS-DOS, Unix versus OS/400 and ultimately MVS, everyone would know exactly where IBM stood, and true blue IBMers would be free to disparage the upstart sibling in a manner that would have made even a Data Processing Division man blench were those terms to be used about the old General Systems Division. Mix it

JCN would have the freedom to mix it in the gutter with the likes of DEC, Sun Microsystems and Hewlett-Packard, offering whatever it felt it needed in the way of communications to IBM mainframes and mid-range systems – and, crucially, the freedom to rejuvenate its product line every three months if it needed to in order to stay ahead of the competition. It’s an even bet that in a few years, JCN would be bigger than True Blue IBM, and margins should not be substantially worse because volume would be so much greater than it is now. Some might argue that IBM should then spin JCN off to its shareholders as a completely independent company, but as its ambition remains to own the computer industry one way or another, such a move would likely not appeal. Well, looks like the plan now runs very much along those lines, although there is a lingering worry that the schism between the different parts of the company will not be as radical as we outlined. Will it work? All that can be said is that it has to be an improvement on what has gone before.