From 'The Economist'....
BUSINESS
A bad business
For the past 15 years the personal-computer industry has walked
tall. But now it is hobbling
MOST of the geeks shoving their way into New York's Javits convention
centre for last week's PC Expo saw only the usual cornucopia of computing
delight. As always, the message of the PC industry's annual extravaganza
was that everything is getting smarter, faster, cheaper. Never has the
strength of Moore's Law—that computing power will double every 18
months—seemed to be more gloriously affirmed.
Those prepared to look harder might, however, have seen the writing on
the wall for PC makers. Though they will sell more machines in the next few
years, 1999 may prove the high-water mark for revenues in the industry's
American heartland. As for profits, even in today's market they are hard to
come by—witness recent losses at Compaq, the word's biggest PC maker.
Although some of Compaq's difficulties are unique, stemming from its
acquisition of Digital in 1998, few of its competitors are thriving either. Both
IBM and Hewlett-Packard have struggled to make money from PCs. Even
Dell, the acknowledged pace-setter, has found its margins under pressure
since it began to sell cheap PCs for the first time earlier this year. As for
most of the rest, don't ask.
These low profits partly reflect overcapacity and competition for market
share. But PC makers are also worried that the relentless cycle of upgrades,
stoked by marketing and technology, is losing steam. The main cause is the
Internet, the very phenomenon that many believed would be their greatest
opportunity. Now, some Internet service providers (ISPs), such as America
Online, are even giving PCs away.
The shock is all the greater because the past 12 months have been
exceptional. Demand from businesses has been driven by the need to
replace older PCs that were vulnerable to the millennium bug. In the
consumer market, the Internet bandwagon is driving all before it, with
first-time buyers, now the main engine of sales growth, accounting for 27%
of all unit sales, according to Forrester Research, a high-tech consultancy.
At the same time, more households that already have one machine will buy
another to ease the competition for mouse-time between family members.
Forrester predicts that this year American consumers will buy 16.9m new
PCs—17% more than last year—raising household penetration to 52%.
However, both business users and consumers are increasingly happy to buy
cheap “good enough” PCs without the latest, fastest and most expensive
chips from Intel. Despite raging demand, average prices of both business
and consumer PCs will tumble by about 14% this year. As a result, total
revenues will be only 2-3% higher than in 1998.
That is an awful omen for next year. Few businesses will be replacing the
more-than-adequate machines they have just bought. And the recent flood
of consumer buyers will subside. The remaining first-timers are likely to be
more resistant to the complexity and expense of PCs, and to the cost of
using the Internet. Meanwhile, supply will overflow. Having ramped up
production over the past couple of years, Intel and its low-price
competitors such as AMD will not be able to scale it back at multibillion
dollar chip plants, and will swamp the market, cutting chip prices even
faster.
Forrester reckons that next year's PC-industry revenues in America could
be as much as $8 billion—or 14%—lower than they were in 1999. There
may be little recovery in later years, because falling prices will offset
modestly growing sales. IDC, another forecaster, thinks that the average
price of PCs will fall less dramatically than Forrester expects. Even so, it too
reckons the value of business and home desktop PC sales will be no higher
in 2003 than it is today.
The PC market has seen boom and bust before. The difference this time is
the absence of what was the main impetus for PC sales in the past: the need
for frequent upgrades to run the ever-more sophisticated products of the
software industry.
Users have little reason to change the PCs they have bought over the past
year. The latest release of Microsoft's Office, a suite that includes such
programs as Word, Excel and Powerpoint, demonstrates that software
developers cannot think of any useful new features that require more than
the computing power of current machines. Games run better and more
reliably on dedicated consoles such as PlayStation and Nintendo. Sony's
PlayStation 2, to be launched next year, will even rival the consumer PC by
adding a web-browser, e-mail and a digital video-disc-drive, which, among
other things, will enhance the machine's storage capacity.
But the real rebel against the tyranny that
says PC owners must always upgrade is the
Internet itself. Very little web content will
require PC owners to buy new hardware;
every new website will prolong the life of
existing hardware. Websites depend on
attracting as many visitors as they can.
Nobody would want to limit the appeal of
their sites by designing them to work
properly only on the fastest PCs.
If users have spare money to devote to
computing, they are more likely to spend it on increasing the bandwidth of
their connection or on staying online for longer, both of which will benefit
ISPs and cable operators rather than the PC industry. Although Dell says that
people typically upgrade their PCs six months after getting a broadband
connection, even today's cheaper PCs can cope with the fastest cable
modems. Forrester's Eric Schmitt says that users are likely to spend
whatever money they have left over on something that makes a difference,
such as a snazzy flat-panel screen or a rewritable CD-ROM drive.
By the time firms get around to replacing the PCs they have bought in the
past 12 months, corporate computing will have been transformed. As firms
rush to become e-businesses, client/server computing, which put the PC at
its core, is giving way to Internet computing, in which applications and data
run entirely on servers. When little is asked of the PC other than to gain
access to these servers by means of a browser, users will not be able to
spot much difference between old PCs and new ones.
What does all this mean for the PC industry? In a single word, pain. Internet
computing will lead to lots of portable “thin clients” that will be inexpensive
alternatives to laptops. Desktop PCs will remain competitive, but only
because of their amazing cheapness. Less mature markets in Europe and
Asia will absorb some of the overcapacity, but not for long. The same
market dynamics apply there too: after all, the Internet is nothing if not
global.
So the PC firms will have to learn new tricks to survive. Apple's success
with the novel design of its iMac shows one way to go. Super-efficient
direct sellers such as Dell and Gateway are starting to sell other electronic
devices and are even determined to become ISPs themselves. Last week
Michael Dell, Dell's founder, said that he means to start his own service.
Gateway, for its part, is in talks to buy EarthLink, a rival of America Online.
Meanwhile, technology heavyweights such as IBM and Hewlett-Packard
might question whether they should stay in such a commodity business at all.
IBM already makes more money selling its technology to firms like Dell than
it does flogging its own PCs.
Wintel's worries
For Microsoft and Intel, the two halves of the Wintel monopoly, life also
seems likely to grow harder. The $50 that Microsoft extracts for each copy
of the Windows operating system will look increasingly egregious. With the
Justice Department watching, Microsoft will be wary of exploiting its
monopoly to the full. Commercial versions of Linux, an increasingly popular
alternative operating system, also give PC makers a stick with which to beat
Microsoft—if they feel brave. As for Intel, it must now realise that merely
accelerating its product introductions is not working. Intel will doubtless
pack a lot of its chips into new Internet devices, so long as it prices them
cheaply. But it is vital for the Wintel pair's margins that firms rush to buy
servers runnning Microsoft's industrial-strength, but delayed, Windows
2000 and Intel's equally late 64-bit successor to the Pentium.
And what of the market leader? Poor Compaq falls halfway between an
IBM that sells technology and services and a Dell that is a slick
build-to-order assembler, a condition that selling its AltaVista search engine
has done little to remedy. The firm's chairman and acting chief executive,
Ben Rosen, may even feel that he did Eckhard Pfeiffer a favour when he
sacked him in April.
LINKS
Compaq, IBM, and Hewlett-Packard write about changes in the PC
market in their most recent annual reports. Summaries of recent research
by Forrester Research are available on the firm's website. |
