Letters
To The Editor: More
On Waves Of Technology
Our
good friend Mark Stahlman, analyst, venture capitalist and
occasional contrarian, continues to look at the issue of the
length and cause of technology cycles.
He’s found several interesting issues to pursue.
Sometimes (as usual) we agree.
Amy:
Have
you been following the EDS (and CSC and IBM?) announced falloff in
new service contract revenues?
How
might this relate to the thought that without new platforms (and
thus new applications markets), the opportunity for sustained
growth in services for older platforms is limited?
Remember
that since the advent of the effects of Moore's Law in the 1960's
we have had a steady stream of new platforms . . . so how are we
to understand IT growth if there aren't going to be any more?
(Editor’s Note: Mark
Stahlman is alluding back to his notion of discrete-length cycles
or waves of technology, each enabling a platform.)
What
are the broader implications for the economy -- indeed, for
national security-- if we have reached the end of the Moore's Law
Era?
Mark
Mark,
good questions.
I
don't think we've reached the end of Moore's Law (yet) but I do
think we'll reach it eventually.
Of course, I may not care by then.
More to the point, we're likely to be scaling another,
different physics-based curve.
I
think it is true that services-based businesses are mainly about
buying the skills to implement new technologies that are not yet
very well understood by internal staff.
It's an effort to accelerate the exploitation of
technology. Therefore,
if we absorb technology faster than we build new technology we may
need less (maybe a lot less) of high-end services.
Or, we could figure out a way to rethink architectures, on
top of existing technology platforms (software on existing
hardware, if you will) that will make everything old new again.
That should remind you of web services, I think.
Amy
Mark Stahlman Disagrees
About Whether Managed Computing Will Make A New Market
Amy:
Try
this out – a kind of new theory on Waves and Theories.
Steven Milunovich of Merrill Lynch has just published his
latest installment in his series based what he calls his
"Wave Theory" -- State of the Industry: The Coming Era
of Managed Computing (10 September 2002).
In
outline, it closely resembles some work that I did on Wall Street
in the late 1980's and early 1990's (culminating in the HBR/Upside
article, "Why IBM Failed"), in which I predicted the
rise and fall of the "Network Computing" Wave (i.e.
workstations) and then went on to predict the rise and fall of the
Internet Wave (culminating with a "crash" around the
year 2000). Not bad
for 1992!
However,
in detail as well as overall conclusion, Steve does not follow my
research methods or findings.
Indeed, based on this report his current long-range
optimism appears to me to be over enthusiastic and, while
understandably driven by the requirements of his audience of
tech-investors, I find it misleading.
Let
me try to straighten some of this out and offer another point of
view.
Steve
is an old hand at following the "hardware" vendors and
was previously associated with Mary Meeker's Internet group at
Morgan Stanley. As you might recall, these are the folks that brought us the
wildly wishful curve of never-ending upward growth which we
discussed in your newsletter earlier this year.
At
Merrill, Steve has continued the task of trying to explain how the
industry works. After
huddling with the top brains at Sun, IBM and HP/Compaq (as our
hostess, Ms. Wohl frequently does as well), he has tried to
explain what they see as happening . . . in his own words, of
course.
Not
surprisingly (given the Wall Street game of near-term pessimism
and long-term optimism), in this report he has renamed what is
commonly referred to as "Web Services" as "Managed
Computing" (you name it you own it!) and pronounced that this
will be an engine of massive growth for the industry, once we get
over the temporarily over-bought calm we are now trapped within.
Sail on!
Weak
today, strong tomorrow. Buy
now, hold forever (or until the market tells you to
"rotate" sectors).
But, whatever you do, keep reading my firm’s reports and
maintaining our business relationship.
The
only problem is that is will not -- nay, let me say cannot -- work
out that way . . . and Milunovich actually says as much, stating,
"The technology matters less but managing technology matters
more." And
"Tech is maturing (but not mature)."
He
advocates “Integrate or Innovate, with larger companies
emphasizing integration."
What sort of a choice is this?
If
you remove the new technology, remove the innovation, remove the
breakthroughs and you emphasize the integration, management and
services aspects of the industry, you have significantly lower
growth prospects, single digits at best, probably shrinkage in a
"bad" year.
What
Milunovich is describing is a rapidly maturing industry that has
run out of new ideas and can only consolidate, integrate and . . .
ultimately shrink.
(Editor’s Note:
Think of Coke and Pepsi, as mature an industry as you can
imagine, trading tiny bits of market share at the margins.)
This
isn’t really "Network Computing" because that wave
began in the 1980's and saturated its available market in 2000).
It seems to me that Steve is really just repeating what
he's heard from Sun, IBM and HP/Compaq.
Let's all talk about glorious growth . . . but plan for
saturated-market share-battles, commodity systems, mergers and
more layoffs.
(Editor’s Note:
A viewpoint I’d certainly share:
No guessing required.
During the week of September 16, first IBM and then Sun got
up at Press and Analyst briefings and said quite firmly that (a)
the market was flat to slow and (b) they were going to do better
than that by taking share away from the competition.
Both IBM and Sun, by the way, intend to do that by taking
advantage of HP/Compaq which they believe will be too busy
figuring out how to integrate itself to provide much attention or
competition for the next 12 to 18 months.)
In
contrast, Mark proposes an alternative.
Real
growth through real innovation!
Real breakthroughs leading to real changes!
In fact, I believe that there is a new Wave coming and it
will incorporate many of the advances from the past 20 years of
computer science into completely new computing platforms, which he
describes in greater detail on his own website (www.markstahlman.com).
Furthermore,
for economic as well as security and/or military purposes -- they
are closely related -- and even for cultural reasons, we must
begin to throw away our existing computer systems and get the
computer industry growing again or, else.
(Editor’s
Note: I’d call this
a technofantasy. Vendors
would love this, but the only way to get customers to buy new
systems is to offer systems with substantial incremental or better
yet revolutionary value. Mark
knows that, and you’ll see some more comments about this further
long, if you stick with us.)
Integrating
existing technologies will *not* solve what IBM calls a
"Crisis of Complexity" . . . it will only make things
worse. Web services
(by any name) can only promise meager overall industry growth and
if that's what the "big boys" are looking forward to,
then that's what they'll get.
Integrate
or innovate? Hopefully,
we'll do both.
Mark
Stahlman
Mark,
Of
course it would be better if we could simply move to the next
generation of slickly integrated, very innovative and vastly less
complex stuff. Unfortunately,
that’s not how it works. We
have to get there by continuing to use what we have (hardware and
software), while we test the new stuff and build the systems that
are going to exploit it. We
get there; it’s just not in one giant step.
You’re
right – we need to innovate as much as we can and then we have
to figure out how to integrate the new innovations into the
systems we’re already using and move forward, one step at a time
Amy
.
Have
Customers Changed Their Mind About Routinely Buying More Pc Speed?
Mark
went on to offer a few proof points you might find interesting, so
we’ll share them.
First, from John Markoff and Steve Lohr in the New York Times
Amy,
Perhaps
the two best reporters on the tech-beat have been producing some
very interesting articles lately --
John Markoff noted today, http://www.nytimes.com/2002/09/30/technology/30SPEE.html?pagewanted=print&position=top
that customers have discovered “faster may not matter.”
Customers no longer decide at predictable intervals that
they “must” turn in their old PC’s for new ones.
The purposes we mainly use PC’s for are pretty well
served by what we’ve already got.
The day before, September 29, 2002, in an article, http://www.nytimes.com/2002/09/29/technology/circuits/29CHIP.html?pagewanted=print&position=top
by Markoff and Steve Lohr, the technology writers noted that
although you’d expect the giant Google search engine to be a
natural customer for Intel’s fast Itanium chips, in fact Google
has declined the investment.
In
these articles, John and Steve raise some very basic questions --
-
Is
the presumption of endless PC demand reasonable?
-
What
(if anything) is being done to introduce new breakthrough
products?
-
Who
has a grip on the current situation?
While
their focus on PC's is understandable, these questions can (and
should) be extended to the rest of computing . . . as well as to
networking -- since the new demand for networks comes from new
computers, not from new people.
(Editor’s
Note: I’d argue that generally speaking it’s people’s need to
solve problems that drives the need to buy the computers; the
computers aren’t ordering themselves just yet.)
In fact, there have been only two surges in PC demand . . . ever.
There was an original "standalone" PC growth-wave in the
1977-1985 time period and another for "networked" PC’s
in 1992-2000. Each of
these waves represented the introduction of a new
"platform" -- addressing different applications and
fueled by different core buyers -- spreadsheets/desktop-publishing
for corporate buyers for the earlier "PC wave" and
browsers/email for homebuyers in the "Internet wave."
Is there another wave in the future for PC makers? No, probably
not.
Is there anything more to do with this sort of a machine -- after
deploying them first as standalone "productivity tools"
in the 1980's and networked "connectivity" tools in the
1990's -- perhaps other than throwing them away?
Despite the massive ad-spending to convince people to use them as
photo-albums and moviola-film-editors, none of these
"rich-media" applications qualify to give new life to
the PC, creating another broad, high-volume new
"platform."
Is the PC dead? Yes, probably so . . . and it will never again
enjoy the sustained double-digit growth that it has (remarkably)
enjoyed twice already.
The implication of merely asking these questions on the front page
of the NYTimes business section is that the economic impact of all
this is already being discounted.
We are -- in this sense -- already becoming "Japanese."
Mark
Mark,
Have
you noticed how IBM has all but removed themselves from the
desktop PC business, corporate and consumer? How Microsoft
is looking for non-PC markets for its operating systems?
There
are people who are yet to be reached by the PC market in both the
U.S. and in countries, particularly less developed countries,
around the world, that still represent considerable volume
markets, but they may be reached by companies with lower overheads
and different pricing and business models than the lords of the
Silicon Valley.
Amy
Can
The Industry Return To The Good Old Days?
Then
Mark noticed that Craig Barrett from Intel was claiming it was the
economy, stupid. That
is, as soon as the economic picture brightened, corporations would
start buying more new stuff and the PC business would go back to
where it belonged.
Speaking
in Dubai, Intel Chief Executive Craig Barrett said on September 29
he did not expect a real turnaround in computer sales until
companies returned to profitability and started spending on
technology.
"I expect some uptake in (personal) computer buying during
the holidays but I don't know how strong," Barrett told a
news conference in Dubai emirate in the United Arab Emirates.
"(But) a turnaround will happen when there is corporate
purchases of IT infrastructure and that cannot happen until there
is corporate profitability in the United States, Japan and Western
Europe...It is very difficult to predict when that will
happen."
Barrett said he expected Intel, the world's largest semiconductor
maker, to emerge stronger from what he described as the most
dramatic recession for the IT industry in 10 years.
"Absolutely yes. We expect to emerge stronger than we went
in. We are still profitable and we are still investing in research
and development and in capital investment," he said.
Meanwhile,
Craig Mundie of Microsoft was proclaiming that XML and various new
“toys” (Mark’s term, not Craig’s, I think) would fund a
rebound in sales. We
don’t think so.
Mark
believes, and I agree, that structural and philosophical changes
in corporate thinking about technology may have changed more than
that. Companies have
figured out that buying faster desktops every two or three years
isn’t necessarily an investment in worker productivity – but
investments in collaborative software and systems management might
be. So it’s not
that they won’t spend – some of them, by the way, are spending
through the recession to SAVE money – it’s that they’re
spending and will spend differently and smarter and this
advantages and disadvantages different market sectors.
Of course there are always a few lucky companies that can
either see the shift coming or move fast or who are very well
diversified and can simply change emphasis.
I’d nominate Microsoft and IBM for gold stars here.
In
any case, expecting things to return to the “good old days” is
rarely a reliable prediction.
Too many changes in the meantime – new technology
opportunities, new business attitudes, and new problems to solve.
And good technology vendors will always arise to meet those
challenges. Which
ones – old friends or new – is the question.
(back to top)
Comments or Questions: Send Email to
opinions@wohl.com
Home/
Search
/ 2005 Articles /
Issue Archive / Free
Newsletter
Entire contents © 2001 by Amy D. Wohl. All rights reserved. Reproduction of this publication in any form without prior written permission is forbidden.
