To take a view of how successful Louis Gerstner has been in turning IBM Corp around in the two and a half years he has been at the helm, one has first to agree on where the company was heading immediately before he took over, and that is tricky. John Akers, who carried the can, and his predecessors, who have at least as much complicity in the events that led to its downfall, certainly made mistakes that would have sunk a smaller company, but in his final moves at the company to cantonize IBM, we would argue that Akers had at last put the company on course for its best hope for future prosperity, and that by reversing that strategy, Gerstner has made things look better in the short term at the expense of a much less promising medium term. Let it be said straight away that IBM is not Digital Equipment Corp or Unisys Corp: it will unquestionably be around in some substantial form for the foreseeable future, and will come to be regarded as a solid part of corporate America – as successful as Xerox Corp, rather less successful than General Electric Co Inc, rather more successful than underachiever, Eastman Kodak Co.
Evil hour
But there is still so much that is dying within IBM that without a dazzling acquisition – an Electronic Data Systems Corp rather than a Computer Sciences Corp – it has little chance of remaining the world’s biggest computer company much beyond the Millennium – a little over a decade ago, it was still 10 times the size in revenue terms of the next biggest computer company, today Hewlett-Packard Co and Fujitsu Ltd are together bigger than IBM. But surely, with the share price flirting with $130 and Wall Street to a man and woman having the stock back in its good books, there can’t be too much wrong with the Louis Gerstner model of IBM? After all, the mainframe business is flourishing again now that the CMOS strategy has been successfully executed. Well yes, but that successful strategy has only postponed the evil hour when the base becomes unsustainably shrunken to justify the investment needed to keep it current. IBM’s problem is that the mainframe base still generates almost all th e company’s profits, yet no new applications are being written for System/390, there is a Year 2000 panic just over the horizon that will see many users rushing to cut their losses by moving to new packaged software on much newer hardware, and the more IBM insists that the mainframe is just another server, the more users are going to realize that the thing is just not on the same price-performance curve as Unix or Windows NT servers, and screw IBM down until it is. At that point, the cost of m aintaining a proprietary OS/390 becomes one of rapidly diminishing returns. So then what does Louis have mapped out? He could always do a DEC and simply give the business away to Microsoft Corp and Intel Corp by making the escape route NT on Pentiu m Pro. The least bad option for IBM with its mainframe users is probably to seek to migrate them to the RS/6000 and its SP variant – but the company’s ever-qualified commitment to RS/6000 has ensured that it has not been nearly as successful as it c ould have been or IBM in 1990 and 1991 insisted it was going to be: the Unix server race is still seen as a two-horse one between Hewlett and Sun Microsystems Inc, with IBM back there in the pack with the likes of DEC while Silicon Graphics Inc bites at its heels by showing how serious and persistent you have to be if you want to differentiate your machines from the pack.
Meanness
Moving the AS/400 to the PowerPC might have been the right strategy if the rest of the PowerPC strategy had worked, but IBM’s indecision, coupled with that perennial self-defeating meanness that insists that no hardware must ever be given away, however great the potential prize to be secured thereby (a meanness that has cost the company most of the System/36 base, but which has oddly never prevented the company from giving OS/2 away like a drunken sailor at the bar), has ensured that IBM has s ecured just about the worst of all possible worlds with the PowerPC – whereas had the Microelectronics business been spun off as Akers wanted, there is little doubt that with IBM’s continuing commitment, PowerPC would be much more successful and Microelectronics a lot healthier. The same applies to the disk business: the PowerPC strategy for AS/400 might genuinely have succeeded had the AS/400 base been as big as it could have been had IBM not constantly crippled the machine and its sales people for fear of self-impact on the benighted mainframe – after all, the architecture of the AS/400 is much more modern than that of Pentium Pro, which still carries with it much of the same baggage as OS/360 – sorry, OS/390. But instead, IBM is hidin g attrition in the AS/400 base by totting up processor revenues to include internal disk and comparing them with those of a couple of years ago that did not include disk – and will soon be doing the same with some of the CMOS mainframes. And once again, as with the decision to junk Ramac in favor of Storage Technology Corp’s Iceberg, the storage systems business is made the whipping boy for others’ failings. How fervently the AdStar folks in San Jose must wish they had the same freedom of action as the people running Lexmark International Group Inc so that they could light out after Seagate Technology Corp instead of being bogged down in internal IBM politics! Then there’s the personal computer business, going into gentle decline through the inability to move at anything like the speed the market demands. It ceded leadership to Compaq Computer Corp three or four years ago, and sees Dell Computer Corp looming large in its rear-view mirrors – but at least the shrunken personal computer business isn’t costing IBM a lot of money now.
Future strategies
So what of IBM’s strategies for the future? Networks with everything is rapidly turning into a systems integration business as IBM buys in most of its networking gear – that way, so that it makes money for its suppliers while taking all the kicks from the customer – and again sends out the message that software is worthless by giving Notes – which cost its shareholders $2.5bn – away. And then there is the facilities management and implementation business. Anyone that is aware of the 20-year US air traffic control fiasco or the German Bildschirmtext contract must wonder uneasily how it is that, having so abjectly failed to cure any of the company’s other defects, Gerstner should miraculously have cured its inability to bring in a big project on time and on budget. We won’t know until the chickens start coming home to roost in a year or three, but those that believe he has will put the mortgage money into IBM shares. The rest of us will suspend judgement, while continuing to wonder why, when what IBM desperately seemed to need at the helm was a dictator who was intimately familiar with every aspect of the computer industry, its board chose to appoint a cost-cutting committee man with little understanding of the industry.
