IBM chairman John Akers has been doing the rounds giving New Year interviews – Computerworld wasn’t the only beneficiary (CI No 1,089), Datamation got one too, and no doubt there are several others that haven’t come our way yet. The first striking thing is that an IBM chief executive is not only going out and talking to the press, making himself available for interview, but is larding his comments with mea culpas and impassioned declarations that IBM has got to try a lot harder, and as long as he is around running the company, will try harder. It may not all be very convincing to those who have got on the wrong side of the company in the past, but it does mean that users have a much bigger stick with which to beat the company when it doesn’t meet their expectations now than they had three or four years ago. And changes in the corporate culture did become very noticeable in 1988, making the time-hallowed task of IBM watching decidedly bemusing as for the second time in the 1980s, old hands at the game had to throw away the rule book. The first time was when IBM abandoned the not-invented-here attitude to all other companies that had successfully seen it right through from 1964, and started handing over critical aspects of its business to third parties, so that Microsoft Corp became the keeper of the soul of the Personal Computer, which by 1984 had become a $5,000m IBM business. IBM bought over 20% of Intel Corp, promised to limit itself to 30% of Rolm Corp and then bought the company outright, exchanged much of Satellite Business Systems for a 16% stake in MCI Communications Corp. Breakneck pace So suddenly, IBM watchers felt they couldn’t assume that IBM wouldn’t solve any other of its short term problems by going out and buying the company that seemed to have the answer. Yet that strategy, born only in 1981, is already very nearly dead. IBM is bankrolling Steve Chen’s efforts to take on Cray Research Inc in the supercomputer business, just bought 25% of optoelectronics specialist PCO Inc, has a 10% or so stake in Metaphor Computer Systems – but each of these is a tiny company where IBM can wield as much influence as it feels it must – there is no established corporate culture to battle as there was at Intel or Rolm, and none is critical to the heartland of IBM’s business as Rolm’s telephony technology was intended to be. While all those early investments were being made – and slowly turning sour, IBM was growing at breakneck pace. It’s hard to credit now that at the end of 1981, IBM was doing only $29,000m a year. It put on another $21,000m – a 1988 DEC plus a 1988 Unisys – in just the four years 1982 to 1985. And then its soaring growth was decisively derailed, and it is short odds against the company putting on as much as half that four-year growth over the years 1986 to the end of this year. During the years of rapid growth, profits seemed to follow effortlessly: since 1984 they have been hard won indeed, and as John Akers stresses that improving return on assets is the company’s number one priority now, much slower growth looks like being the shape of things to come for several more years. Hot breath During the 1980s, the environment in which IBM operates has also changed dramatically, with two serious number twos emerging in the shape of shape of Fujitsu and DEC. Neither of those companies shows any sign of being derailed in the way that the long-time champion of the 1970s, Honeywell, so emphatically was. What makes the feel of hot breath on its neck from DEC and Fujitsu so uncomfortable for IBM is the fact that each has got to its present size almost entirely by its own efforts. In the 1970s, Honeywell made more and more bizarre acquisitions, in the elusive quest for critical mass, as Unisys is doing now: there’s little to worry IBM there, because it can count on internal tensions and the problems of managing massive internal upheavals out of each successive acquisition to hamper Unisys’ ability to give it too many sleepless nights. But where DEC salesmen are still raiding IBM’s 4300 and System 36 bases as the h
appiest of happy hunting grounds, Unisys has to spend as much effort holding on to what it has at the top end from raids by IBM as it does making inroads into IBM’s still-vulnerable mid-range. One of the disconcerting things for John Akers must be the fact that the customers he says that IBM is now listening to so very attentively are not the ones who got saddled with a few of those embarrassing System 23s, those who grew up on the 4331 and now don’t like the 9370 route IBM has mapped out for them, they are the ones in IBM’s heartland that the company has always been able to depend on to enable it to meet the payroll at the end of the month. And it’s not that they are deserting IBM in droves Amdahl Corp of Comparex Computer Systeme GmbH may nibble around the edges, but the vast majority remain True Blue. It’s simply that with every enhancement that IBM introduces, fewer and fewer of them feel compelled to upgrade at the rate IBM needs if it is to prosper as it did in the 1970s or early 1980s. And so IBM has to listen as it’s never listened before in order to come up with the things that will persuade those users to put their hands into their pockets. And if IBM listens to 30 of its biggest users, it gets 30 different sets of demands, each of which is the answer to the prayers of only a small proportion of the base – and which together place an increasingly onerous burden on IBM’s battalians of software developers. And IBM is only able to meet the growing burden of software development by charging exorbitantly for that software when it does come to market. Over time, software and services will become a significant part of our revenue, Akers told Datamation. What are they now? They’re 20% or something. They’ll be – 30%, maybe 40%. That is not only IBM’s devout hope, it is a necessity. Chorus of grumbles But while System 38 users know what to expect, already the company is meeting a growing chorus of grumbles from System 36 users over the price burden imposed by the software they need if they migrate to the AS/400. System 38 users are locked in, and most are paying up cheerfully. 3090 users are equally locked in, but are paying up less and less cheerfully every year. And many are beginning to hope that help may be on the way. The only justification to Fujitsu – already set to be a $20,000m company, a third the size of IBM by the end of this year – for paying hundreds of millions of dollars to get to see MVS code legitimately is to use that code to sell its machines not just in Japan but all around the world. And the only way it will be able to compete will be by combining better functionality with significantly better prices than IBM. And, rushing to the rescue of discomfited System 36 users is the Unix fraternity, where competition on the software front is so fierce and so multifarious that users can get almost anything they want at whatever price they are prepared to pay if only they can find it. So that by 1992 or 1993, IBM’s monopoly of all that lucrative software is likely to be decisively broken. And that is why John Akers is listening to customers as IBM has never listened before: he knows that unless he takes on board everything they say, and satisfies the aspirations behind their expressions of discontent, IBM will not even be able to stand still, and the years of rapid growth will be a fond receding memory.