The news that IBM Corp has a budget of $500m to promote OS/2 has produced a stunned response from everyone we’ve talked to – does the company really have that kind of money to throw away? Few people contest that OS/2 today is a very good, solid operating system, even fewer doubt that it will ever be anything more than an also-ran, a CP/M-86, a CTOS. CP/M-86 was incontrovertibly years ahead of MS-DOS in 1981, and Microsoft Corp struggled mightily over many years and several releases to bring the underlying code of MS-DOS up to where CP/M-86 was at launch. A strong case can be made for CTOS being a better networking operating system than NetWare – its loyal users love it, but then the loyal users that didn’t need deficient or non-existent handholding from Burroughs Corp loved Master Control Program for the 48-bit mainframes – ever met anyone in the whole world who loved MVS? In the computer industry, you can’t snatch victory from the jaws of failure by throwing money at a product perceived in the market as an also-ran.
Manifestly dreadful
Microsoft never touted its single most successful product, Windows, as the answer to every man and maiden’s prayer and in its first two incarnations it was manifestly dreadful. With Windows 3, it burned slowly for a few months and then took off like a rocket: the time was right and the product was just about good enough, despite many underlying weaknesses, high on the list of which was a distressing tendency to crash and burn not once a week but several times a day. But only when Windows started taking off did Microsoft start bringing out the big promotion guns, and very soon, those guns were diverted to promote the next big trick – OS/2. For about a year, Microsoft went along with the IBM line that Windows was for the little guy, OS/2 was for the serious user, until it found the serious user was saying oh yeah? and resolutely sticking to Windows. Microsoft’s other big failure so far has been its furious efforts to knock Novell Inc’s NetWare off its networking perch. Over and over again the big guns were wheeled out in the late 1980s to assure the world that Microsoft was mortified that it had only 10% of the network operating system market and that it was determined to build that to 30% in two years. NetWare is today even more strongly entrenched than ever, and Microsoft is still on the launchpad. The lesson here is clearly that success comes by getting it right first, letting it build quietly, and then bringing the big promotional guns to bear, so that your flackery will be pushing at an opening door.
Remember PCjr
IBM’s latest pitch for OS/2 is to go for the reasonably experienced home user – but such users are probably old enough to remember PCjr… One argument for IBM spending $500m promoting OS/2 is that it would be worth it to the company to spend that kind of money if in the process it simply killed off Windows NT, because the margins that NT servers command would put margins IBM needs on the AS/400 completely out of reach. But does IBM need to spend that kind of money to that end – it seems to us that Microsoft is doing a pretty good job of killing off NT without any outside help. There can be few more graphic examples of a counsel of despair than Microsoft’s decision to hand over support of all NT users in Europe to ICL Plc. The move makes perfect sense in the short term because ICL has 30 years of experience of supporting mainframe users and understands their wants and needs, where Microsoft has none. But is that really the right answer to solving the problem of one’s own ignorance? Shouldn’t the company be building up its own support capability for large corporate server customers, if necessary by buying an experienced mainframe software support house – and there must be plenty of those around that are hurting mightily today as the mainframe wanes – and retraining the entire staff on NT? If NT customers are no longer Microsoft customers when it comes to support, how is the company ever going to learn what they really need in future releases of NT? Getting that informati
on in precis form from ICL is just no substitute for getting it direct from individual customers. If Microsoft is not prepared to invest its own money in building up a support capability for NT, why should users be prepared to bet their businesses on the operating system? Unix may not be the best operating system in the world, but at least there are a zillion companies out there selling and supporting users on the operating system and a whole industry dedicated to coaxing and cajoling it into meeting the needs of disaffected mainframe users – and doing so at open systems prices. No-one we have spoken to can find a precedent for a major computer product initially perceived as an abject failure, as OS/2, despite all its merits still is, that has been turned into more than a very modest success by persistence and promotion. IBM’s own history is littered with names that ought to make the company wince, led by 8100 and Series/1.
Cynical and suspicious
And as every year passes, users become more cynical and suspicious and the promotional sleight of hand that worked five years ago just won’t wash today. There are those that believe that IBM’s determination to press on with OS/2 against all odds is simply a matter of pride – but $500m of new money is a helluva high price to pay to assuage wounded feelings of a few executives. Is there no-one left at IBM that remembers the company’s past disappointments and failures, nobody that knows any computer industry history? Over time, IBM will sell millions of copies of OS/2 if you include the ones that it is giving away – the CD-ROM on the cover of the current issue of PC Week for example, which will only bring in its wake a stream of users who got into the game for nothing pleading for support that they are not willing to pay a realistic price for because they got the software for nothing. But when the competition is selling tens of millions, those millions of copies of OS/2 simply become a growing drain on scarce cash resources – OS/2 becomes a cost centre rather than a profit centre. Hesh Wiener of Technology News of America points out that $500m is nearly a dollar a share. Many IBM shareholders originally bought their stock for the handsome dividend it no longer pays. If IBM really can’t think of any better way of spending $500m than using it to promote OS/2, we are sure that IBM’s shareholders would be able to put it to much more productive use.
