|HOME||PUBLIC LIBRARY||ANOTHER PERSPECTIVE||INFOPERSPECTIVES||CONTACT|
At 3 p.m. on February 27 the Dow Jones Industrial Average, which was falling in sympathy with a meltdown on the Shanghai exchange, seemed to drop off a cliff. What actually happened is that the computer behind the average was falling behind, so Dow switched to a backup system, which caught up very quickly, bringing the day to a miserable end. The Dow incident is an example of the Roadrunner Effect, where the coyote runs past the edge of the cliff but doesn't fall until he looks down. A similar thing might be happening in the computer industry right now.
Did the fall of the Dow Jones average and other stock market indicators around the world portend tragedy? Nobody actually knows, of course. The only safe prediction at the time was that pundits would emerge like lice on a dipped sheep. Most would dismiss the downturn as a normal incident and predict with absolute certainty that better times ahead. Not one of them knew what they were talking about, because nobody can actually predict the way stock markets is headed.
It might turn out that the hopeful voices will be absolutely right and that the stock market's gyrations are merely an ordinary part of the financial system. Then again, it might be that the market tremors of spring 2007 are like the ones during the first part of 1929, portending a great crash that will wipe out fortunes in a trice.
Whatever you think, if you happen to manage computer operations somewhere this might be a good time to start contemplating Budget Plan B, which involves a lot of strategic change and maybe more than a little belt tightening, even as you try to work out the details for Budget Plan A, which involves business as usual.
We're making this suggestion because one of the very smartest and most successful people to ever come along in the computer business, Michael Dell, seems to have turned to Plan B himself. He is a great computer merchant and most assuredly not a panic merchant, so what he does ought to be carefully considered by anyone who is part of the computer racket, such as people who happen to run IBM.
Mr. Dell, if you have not noticed, is in the middle of a making changes at the top of the company he chairs that can be compared to those made by the Italian government. The main difference is that the changes at Dell won't lead to the collapse of the company. The Italians by definition have to fold their government whenever they want to make a big change. Still, Italy goes on, and of course it is a wonderful place. Dell will go on, too, but right now a lot of executives are probably practicing the safety maneuver known as the gluteus maximus chair lock.
IBM differs from Dell in many ways, not the least of which is the role of X86 servers in the two companies. For IBM, X86 machines are the least costly servers it sells, and very possibly the ones with the lowest profit margins. For Dell, X86 servers are the whole shebang and also products that have very nice margins compared to, say, desktop PCs.
What IBM and Dell have in common is that they both depend on sales to business. Dell sells plenty of equipment to individuals and small enterprises, but corporate sales are really the meat in its fiscal diet. Both companies are acutely aware that sometime last year, probably during the third quarter, the server business started looking a little bit like the housing market, which means kind of ugly. The weak numbers really showed up during the fourth quarter and might have continued their unhealthy trend during the current quarter, too.
When Dell first started feeling frustration during the third quarter of last year it might well have attributed some or even most of its discomfort to a big product transition, the forthcoming release of Windows Vista. But Vista has nothing to do with server sales and for now very little to do with corporate sales of desktop and laptop PCs. IBM must have noticed problems about the same time, but lots of IBM products have a long selling cycle and a long delay between orders and delivery compared to X86 servers you can get practically overnight. IBM's reported results lag by several weeks, and sometimes longer, the incoming orders the company sees. So, if you look at IBM's figures and try to match them to Dell's it is usually a safe bet that you can shift the numbers a month or two in time to see if they line up.
Dell's change in quarterly revenue on a year-to-year basis might be due for some revision because Dell is in the middle of some major updates to its accounting and reporting methods. Still, if you take a quick look at the figures for the past year you can see that the year ended in a downturn for PCs and closed out with flat revenue when it came to servers.
IBM's year-to-year pattern of server revenue varies by product line, as each server family has its own product cycle, but, aside form a nice uptick in mainframe sales during the third quarter, the pattern shows that revenue seems to have flattened out.
What makes the flat figures very unpleasant to look at is that they have come against a background of generally very good economic news. Corporate profits have been good and among companies that use a lot of computers, outfits like banks and other financial institutions, reported earnings have been stupendous. The volume of trading, which creates demand for servers, has been rising nicely for some time. Other segments of the US economy and most of the other big economies around the globe have been healthy, too, and that also ought to lead to steady increases in server sales. IBM does business everywhere and Dell has a significant presence in all the advanced economies. Both ought to be able to sell lots of servers to growing enterprises and to the governments that live off them. But this did not happen, not to the degree one would expect based on the headlines describing in often glowing terms the state of economies here and there.
Both companies are reluctant to throw out winning strategies and in the case of Dell, winning executives, too, in response to a little turbulence.
But Michael Dell didn't just sit tight for the past few months and wait for the world to return to some former state. He must have started looking for fresh talent months ago, tentatively at first (if he is capable of being tentative). He must have decided to make changes at the top of his company weeks ago. And now he is in motion, kicking off what will inevitably become a chain reaction reaching every corner of his enterprise.
IBM is probably in motion, too, behind its thick corporate curtains. It has already made a number of executive changes during the past few months, but it has done nothing that could be described as a major renovation. Maybe IBM is still recovering from all the great changes it made during the Gerstner era. Maybe IBM thought it had done all it had to when it disposed of its PC business. But we doubt IBM can afford to ignore the signals coming from the worldwide server market, indications that business not somehow different and maybe not as healthy.
What we expect is that, as is often the case, IBM will act a bit later than other companies might, but in the end implement some striking changes in its character and maybe even its structure.
IBM can't act as if it is in isolation. Sun, crippled during the dot com collapse, is back on its game and in some ways may be tougher and smarter than ever. Hewlett-Packard, despite its episodes of corporate scandal, seems to be concentrating very hard on every aspect of its diverse business and, in servers, building a broader base of customers. Even Dell, arguably the most troubled of the computer giants, remains a tough competitor in the portion of the server market where it can compete with IBM. Moreover, Dell is openly admitting, through deeds as well as words, that it must approach its opportunities with a fresh and open mind.
Yet the biggest factor in the server market might not be visible to anyone who concentrates on the vendors. The biggest factor may be on the other side of the table, among the buyers. If business conditions become more arduous, customers will be less abundant and much more picky. The server trade may become much more of a buyer's market, and one with fewer eager buyers than there were a year ago.
If business comes under pressure, a lot of customers, with the possible exception of Dow Jones, may find that they have more than enough computing capacity for the near term. And even if that's not the case, even if business in general needs a lot more server capacity, computing departments might have to learn that in uncertain times, they'll have to make do.
But things might not be all that bad. The tough patch in server sales might be short-lived, and so might the perturbations in the world's stock markets. If you want to guess how things will turn out, listen to the experts. If they all say that things are fine, quietly but determinedly make your way to the lifeboat.
— Hesh Wiener March 2007