Partners and competitors of IBM Corp that have suffered at the company’s hands will feel a sense of schadenfreude that the company is brought so low, and will reflect on how right they were when they said back in the late 1970s and early 1980s that if the US Justice Department split the company up […]
Partners and competitors of IBM Corp that have suffered at the company’s hands will feel a sense of schadenfreude that the company is brought so low, and will reflect on how right they were when they said back in the late 1970s and early 1980s that if the US Justice Department split the company up as it was then threatening to do, they would have four or six IBMs to compete with rather than just one. The best thing that could have happened to IBM was to be broken up a decade or more ago.
Mischief
The wretched state to which the company has been brought is nevertheless a tragedy, a tragedy for all the towns in America and across the world where IBM was until recently the biggest employer, a tragedy for all its surviving employees, who now have to operate in an environment pervaded with gloom and a sense of failure, a tragedy for all the data processing managers that made it their business to know all that could be known about the company and its products and culture, who now find their hard-earned knowledge is a rapidly wasting asset. A tragedy for all those whose pension funds and insurance companies bet too heavily on the company’s shares. A tragedy for the customers of all the banks, mostly French these days, that have substantial funds tied up financing IBM mainframes on the assumption of residual values that are emphatically no longer justified, and therefore don’t now have that money available to finance new and promising businesses. But don’t let anyone get away with telling you that IBM’s problems are all the fault of the world economy: if Hewlett-Packard Co can grow at 20% in the current climate, if ICL Plc can remain in the black, if that former corporate basket case Unisys Corp, which compared with IBM started out under its present management with no advantages at all, can return to profit, IBM’s woes have to be acknowledged to be primarily the fault of the company itself, and only exacerbated by the economy. And the failure of IBM management is that it didn’t understand – and still doesn’t understand – the extent of the mischief the genie of open systems it let out of the bottle with the original open IBM Personal Computer could wreak. How’re yuh gonna keep’em down on the farm, after they’ve seen Paree, they used to sing back in the 1920s. How’re yuh gonna keep’em loyal to proprietary systems after they’ve seen generic PCs is today’s equivalent. It was inevitable – and should have been obvious to the kinds of forecasting brains that IBM could afford – that the open systems movement would spread up from the desktop to infect first the minicomputer world and then the mainframe. And that means that even the AS/400 is potentially at risk, and can only continue to be a winner if IBM builds in a big discount to compensate for the fact that it is proprietary – yet instead of recognising this vulnerability, with its incredible but consistent short-termism, the company is squeezing AS/400 users until they squeal with its software pricing on the machine. The AS/400 business is widely hailed as one of IBM’s few current successes: in fact it’s a failure – it should have won every System/36 user by now, and have won enough formerly non-IBM customers to be running at between $25,000m and $30,000m a year, not the $14,000m it is currently doing. And incredibly, users at the top end are still crying out for more power because IBM failed to keep its promise of increasing performance at a rate that would keep it ahead of customer demand.
Moribund
But oh no, IBM couldn’t do that – it might lose 5% of its moribund mainframe sales to the AS/400 – well here’s news for IBM – at the present rate, it will have lost 100% of its mainframe sales in five years, almost all of them not to competing IBM products but to competing companies. Why is Hewlett-Packard eating IBM’s breakfast, lunch, dinner and supper? After all, it had a very vulnerable HP 3000 base to protect. The answer is that it saw the way the wind was blowing and instead of desperately tripping its users up as they tried to move to open systems, it embraced them wholehea
rtedly, giving users the option of either staying on the slow track with an MPE V operating system that over time would converge with Unix, or taking the fast track and jumping across to the HP 9000. Nor can Hewlett be dismissed as having been lucky and got it right first time: it didn’t. The early Precision Architecture HP 3000 machines worked fine – until you started hanging terminals off them. Then they ground quickly to a halt, because the company hadn’t understood what a drain input-output would be on the functioning of the machine. Instead of offloading underperforming RISC machines on its customers – thousands of which were crying out for more power – it rushed to boost the power of its obsolete 16-bit processor by every tweak possible, and virtually gave the things away. And the unhappy customer base stayed loyal in enormous numbers so that the company is now reaping the benefit.
Dreads and aspirations
But it’s decades since IBM last saw beyond its customers’ cheque books and recognised vulnerable human beings with hopes and fears, dreads and aspirations. IBM turned its first Unix system, the RT, into a child of the ghetto and told 90% of its customers that they didn’t want it, it wasn’t for them. And its customers took the hint and bought their Unix systems from Sun Microsystems Inc or aforesaid Hewlett-Packard. For fear of losing one or two sales of its obsolete mainframes, IBM designed the RT so badly that it had to junk the machine completely and start again from scratch to create the (incompatible) RS/6000. As a result, the RS/6000 may be a $2,200m business this year while Hewlett-Packard and Sun are running at between $5,000m and $6,000m in Unix-related sales. There is always another side to the story of course, and IBM still has its apologists: Tim Mead, who was lately in charge at Datamation magazine – another former icon now down on its luck, wrote an outraged letter to the Wall Street Journal after its recent critical piece. He points out that under John Akers, IBM has built its desktop computing business to $8,500m in worldwide sales in 1991, giving it 19% market share; remained the worldwide leader in information technology revenies, at $62,800m last year up nearly 30% from $48,500m in 1985 when Akers took over; and avoided the worst crisis that could have befallen it – which would have been to sacrifice investments in research and development for the sake of a fast buck – it spent $6,600m last year. Unfortunately, this week, IBM has announced that it is lopping $1,000m of its research and development budget, and indicated that turnover will slip back below $60,000m…
