We have all heard about the legal problems Microsoft has gotten
itself into, and about how the federal courts may break up Bill Gate’s
empire. There are many who say Microsoft’s problems are simply the result
of greed and arrogance. But, I think Microsoft’s problems go much deeper
than simple legal problems. There are more basic problems facing Microsoft.
For example, that their underlaying business strategy might be based on
an obsolete technological paradigm. Or, that powerful new hardware might
be a threat to their market position. Or, the consequences of software
immortality. Or, the reality of (at least local) user limits, and the resulting,
“This is more than enough to meet my needs” marketing environment. Or,
finally, the long term consequences of planned obsolescence through the
continual reinvention of the wheel. That is, an environment where software
that runs on Windows 3.1 but not on Windows 98; or software that runs on
Windows 95 but not Windows 2000. So, while it is very easy to talk about
the consequences of arrogance and monopoly, it is not enough. What are
the real problems that face Microsoft? I’ll start by discussing the user
limits that apply to software.
On Technology and Growth
Our culture tells us that technological growth is like an ever expanding
wave that sweeps all before it. But, this is not so. First, technology
is not a big amorphous monolith, instead it is a collection of different
interrelated fields, with various specialties. Second, technologies tend
to grow in interrelated spurts, kind of like a string of falling dominoes.
One technology has a growth spurt that influences another technology; that
influences another technology. For example, take the gasoline engine in
your car. Most of the technology that originally made it possible came
from steam engines and electrical telegraphy. And, once the technology
was created (or adapted), it changed slowly. We say the engines we have
today are a vast improvement on the engines of thirty years ago, yet mechanically
they are almost identical to the old engines. So, what changed? First,
the electronic controls were vastly improved because of the importation
of digital electronics. And, digital electronics made possible spinoffs
like electronic fuel injection. Second, the materials science is better
(longer wearing alloys, higher temperature oils, etc.). Understand, this
is how technology grows, one technology influences another.
Another thing, for a technology to grow someone must need it. Technological growth is not an inevitable wave. For example, the ancient Greeks could have developed an industrial technology, but they didn’t because they had slaves. For them, it was much cheaper to buy a slave then to invest in industrial technology. And their culture reflected the institution of slavery. In ancient Greece real men were warriors who did not get their hands dirty doing slave work. And, intelligent men (and occasionally women) were speculative philosophers who didn’t waste their time thinking about better ways for slaves to do menial toil. Today, this pattern hasn’t changed. Look under your car’s hood. The technological improvements came about mostly because environmental laws were passed. Performance wise, there is nothing wrong with a carburetor or a point breaker ignition system. If the pollution control laws had not been passed, the engines in our cars probably would not have changed much. But, as our culture came to value a clean environment, the regulations were more and more seen as necessary. Environmental laws were enacted, and that created a technological need. Understand, a technology grows because someone has a reason to invest in it. So, how does this apply to Microsoft? Please read on.
The User Limits
If we were to discuss a computer’s speed, you might say it has a 400
MHz (megahertz) CPU, or a 500 MHz CPU, or maybe give some speed benchmark
number. But, these numbers are relative. The real issue is how fast does
the computer run applications for the user. Note, human neural processing
speeds have not changed in 50,000 years. So, until we have cerebral implant
chips wired directly to the brain, the user is the most important speed
issue; at least in the desktop and laptop markets. This leads me to some
interesting conclusions. If a computer can run a set of applications at
a reasonable speed for a human user, with allowance for upgrades, then
it is fast enough for that user. And, if the users needs don’t change,
the computer will remain “fast enough” for as long as it is operable. So,
consider the needs of the user. If her computer is “fast enough” why should
she upgrade her software, or buy a new computer? Understand, “fast enough”
is relative. If you are doing word processing, an Intel 486 running Windows
95 might be enough. If you are doing video editing, this combo would be
much too slow. Fast enough depends on what you are doing. Now, consider
this. It may be that the computers that came off the line two years ago
are “fast enough” for most uses and users. If this is so, then the software
industry may have a huge problem; because software that is adequate on
an Intel 486 process, is more then adequate on a Pentium III. The implications
of this could be awful.
The Microsoft Angle
Back in the 1980’s and 90’s microprocessor capabilities grew enormously.
As Intel created ever more powerful CPUs, Microsoft introduced a series
of ever more powerful versions of its operating systems, DOS and later
Windows. Eventually, Microsoft’s became the dominant writer of computer
operating systems. Also, Microsoft created software applications, like
Excel, run on it’s operating systems. With each new generation of hardware,
more powerful software became possible. So powerful, that older software
versions were totally obsolete in comparison. And, Microsoft was well positioned
to be the first with new features because they also wrote the OS. This
put Microsoft in the “virtually” position of perpetually selling new products
to a fresh market. Soon, Microsoft became an incredibly profitable monopoly.
But, there was always a limit, the capabilities of the human brain. We
humans can only accept so many inputs, along only a limited number of pathways.
So, software features that exceed those human limits are just bells and
whistles. Understand, speed is relative to the user and the task. Once
the computer is fast enough to run the users applications, any extra speed
is useless; unless it can be used to run other programs in the background,
like music from a CD player (actually, the background is a software illusion,
a computer actually runs software in slices, one after the other very fast).
But, how many background programs are really needed? Eventually, any extra
CPU speed is wasted. So, once user limits are passed, hardware upgrades
and new software versions become optional. And, the demand for new software
(and profits) must decline because fewer people are out shopping. Eventually,
we may reach the point where software companies go bankrupt because of
the research and development costs of their new, but unmarketable, software.
Microsoft could be in a terrible bind. Microsoft’s corporate culture is
based on the assumption that “new” software is the path to incredible profits.
Fifteen years ago, when Microsoft first introduced Word, their word processor,
they quickly sold millions of copies. When they introduced the new versions
with new features, they sold millions more. But, as time passed, the number
of new good new features declined. Today, most of the new features amount
to bells and whistles. The power of the new CPUs is not really needed for
word processing, except for dictation and speech recognition (it would
be nice if the software was smart enough to get the punctuation right).
So, why does anyone with an older version of Word need to upgrade? Think
about it, the next version of Windows could bankrupt Microsoft.
The Threat of Powerful Hardware
Another problem facing Microsoft is the speed and power of the latest
CPUs. Back in the days of the XT and AT, about fifteen years ago, 64 KB
(kilobyte) of memory cost $500 dollars, a 30 MB (megabyte) Hard drive was
huge, and 5 MHz was fast processor. Back then, OS options were limited
by the technology. Most users had to choose only one OS to stuff into the
hard drive. And, the same things that limited OS choice, limited software
choices. Back then, you could only stuff one word processor, and one spreadsheet,
and one drawing/painting program, and one browser on your hard drive. So,
once Microsoft became the dominant OS provider, it became natural to use
their other software as well. Today, things are different. For example,
if you own a Macintosh, you can run Mac OS, Unix, Linux, or BeOS directly,
and you can run Windows in emulation. If you have a large enough hard drive,
you can run all four OSs on different partitions. If you own a PC, there
is an even wider choice of OSs. And, there is enough space on the hard
drive for an interesting range of applications. So, you are not limited
to just one anything. Emulation also leads to interesting possibilities.
For example, on a fairly recent Mac most Windows applications run at a
reasonable speed, except for high end games and video. Understand, the
possibilities of this kind of speed and power are becoming available to
all computer users, regardless of operating system or CPU type. For example,
Java, a recent the computer language, is based on an emulated virtual machine,
so software can be written to run on any machine (if an emulator is available
for the hardware). Unfortunately, Microsoft built its monopoly, in part,
on the limits of the old hardware. The new hardware is a threat to Microsoft
because of its power. Because, computer users are no longer tied down to
one OS or one set of applications. Because, it makes Microsoft products
more optional.
The Threat of the CD-ROM
The next technological threat facing Microsoft is the CD-ROM. Understand,
software is as immortal as its medium. And, once an application passes
a few user limits, obsolescence becomes a relative thing (except for bug
fixes in the newer versions). Also, in an emulated world, there will always
be machines capable of running that software. A few years ago, the executives
at Microsoft probably thought that rapid technological growth would make
their products obsolete in a few months. And, back then they were right.
But, passing user limits tends to makes obsolete a relative term. I think
many people look at it this way, “If it’s good enough, and it works for
me, why should I upgrade?” For Microsoft, this is horrible. Most of the
stuff Microsoft put out on CDs, is still out there. Most of the software
Microsoft put out over the last ten years is only relatively obsolete,
in terms of user needs. And, many of the new features in the Microsoft’s
latest software are simply more bells and whistles. Unfortunately, the
CD ROM forces software makers to compete with their own older products.
I once bought a copy of Corel Office Professional for Windows 3.1/95 for
$25 dollars from a bookstore that was having a clearance sale. I have seen
people trying to sell used Windows 95 CDs for anywhere from thirty to fifty
dollars. Do you see the threat?
The Problem with the Intel Pentium CPU
Then there is the problem posed by the Intel Pentium CPU. The Pentium
chip design is based on CISC (Complex Instruction Set Computer) architecture.
Back in the age of the 8 and 16 bit CPUs (the early 80’s), CISC was a reasonable
way of doing business. Back then, Intel and Microsoft created a business
standard based on the x86 series chips (8086, 286, 386, 486, and the Pentiums).
Today, we have 32 and 64 bit RISC (Reduced Instruction Set Complexity)
architecture CPUs that can process data about 40 percent faster than comparable
CISC chips running at the same clock speed. And, when Intel moved to 32
and 64 bit architectures, they had to keep the old 16 bit (8086, 8088,
and 286) instructions to maintain compatibility with older hardware. Much
of the underlying software in Windows 98 is still based on code that was
originally written for the old 286 and 386 CPU’s. For Intel and Microsoft,
moving to a new RISC design would be very expensive (also for just about
everyone else except Motorola), because it would mean starting over from
scratch. The change would require new hardware standards, totally new software,
multiple emulators, and so on, and so on. Meanwhile, the Intel cloners
would have an inside track on the new RISC market. Remember, Apple almost
went bankrupt, in part, because of the switch to RISC CPUs (the Power PC
Macs). But today, Apple’s products have a huge technological edge in the
market because of the switch. In the end, I don’t think Intel can afford
to switch to a new CPU architecture. And, a change in CPUs could be disastrous
for Microsoft because they would have to rewrite, not just modify, their
software. But, this may be a long term problem for Microsoft because so
much of their core software is beyond the user limits, and so can at worst
be only relatively obsolescent (except for graphic artists and virtual-reality
gamers).
Holding on to the Successful Strategies of Arrogant Youth
The next problem at Microsoft is the software between it’s executives
ears. Microsoft got into trouble because of it’s core business strategy.
Times and circumstances change. When a company is small it has to be a
shark; a tough aggressive competitor in the economic sea. When a company
becomes big, it has the resources to make money on the margins. Compound
interest and huge amounts of capital can work wonders. For a big company,
sometimes the best strategy is to be the great tree that all the little
birds come to for shelter. Extreme competitiveness just creates enemies.
Being in too many businesses (what monopolies are about) can cause internal
conflicts and inefficiency, while spreading resources too thinly. So, while,
an aggressive graduate of the business school might see lateral expansion
as a good way to maximize profits and markets; in actual fact it can decrease
profits by diluting the company’s overall efforts. The result is a large
number of mediocre products, rather then a smaller number of well-done
products. Microsoft, by going into things like electronic encyclopedias,
and games, diluted their core businesses, computer OS’s and high end office
tools. I have heard many people tell stories of Windows bugs (and I have
run into a few myself). I’ve heard that Windows 98 was released with a
serious memory bug. And, Windows fell 2000 way behind schedule. On the
other hand, there is the example of Hewlett Packard. When it’s management
realized that the company was in too many businesses, they chose to split
the company into two entities. This made good business sense because two
smaller firms would together be more profitable than one big company with
internal conflicts. Perhaps, this is an example of how Microsoft’s management
could have done things. Instead, Microsoft’s aggressive business practices
just made bitter enemies, and reduced future profits. If Microsoft’s executives
thought there was a good reason to be in the gaming market, or in CD ROM
encyclopedias, then the company should have found partners. If they had
wanted to be in the web browser market, they should have marketed Internet
Explorer non-aggressively, instead of trying to kill Netscape. Their attempt
to kill Netscape forced them to make Internet Explorer a free product,
instead of a profitable one. Understand, size and reputation can be powerful
marketing tools if used right. So, why be overly aggressive when you have
size (with the perception of stability) and well known name working for
you? On the other hand, a negative reputation can be destructive to any
business. I think Microsoft’s executives should have changed business strategies
years ago.
The Problem with Management Philosophy
The last thing threatening Microsoft is their philosophy. The managers
at Microsoft ignored the fundamental principle of free market economics.
It goes like this, “If I scratch your back, will you scratch mine?” This
principle is so simple, yet so many people miss it. Look at the world,
the most successful economies are those that are best at rewarding good
back scratchers; at having limited competitions where the best producers
win. This is why an economy should be a game with rules, with a referee
(usually the government). On the other hand, the economies that repeatedly
fail are based on either on the notion of unlimited competition (El Salvador,
Somalia, etc.), or on pious notions of the virtues of noncompetition (the
Soviet Union, Yugoslavia, etc.). An economy, or a business, based on the
principle of, “Lets get together and knife that fellow in the back and
share the loot (only you’ll have no share in the loot because there will
be no one to protect you from me),” cannot produce. Yet, I think generations
of business school students have had their heads filled with this kind
of nonsense. A successful business does not operate on the principle of,“
Stab the other guy in the back (as long as the government watchdog doesn’t
notice).” But, this is how many corporate managers behave. Understand,
a business works because people come together to reach a mutually profitable
goal; not by conspiring against each other. Managers behaving like little
toughs on the school play ground (while teacher isn’t looking) do not serve
anyone’s best interests, especially the company’s stockholders (I wish
the people who run the business schools would realize this). For example,
why did Microsoft try to kill Apple Computer? They would have done much
better finding ways to support Apple. Think about it. First, if Microsoft
had viable competition, the antitrust lawsuit would never have been launched.
Second, helping Apple would have given Microsoft a “friendly” competitor.
Not a bad way to do business. Microsoft would have been better off if it’s
managers had seen things this way.
On Time and Technology
Microsoft built it’s monopoly on the computer technology of the 1980’s.
Back in 1980 a powerful desktop computer ran at 2 MHz and had 64 KB of
memory. By 1985 the “oomph” factor was up to about 5 MHz and 256 KB of
memory, and no one thought computers would ever need more then 1 MB of
memory. In 1981, Microsoft entered the OS market with MS DOS 1.0. As time
passed, and computers grew more powerful, Microsoft created new generations
of ever more capable operating systems to run in them. The basic idea behind
Microsoft’s marketing strategy was that each new generation of more powerful
computers would quickly make the old software obsolete. But, as the machines
became more powerful, user limits were passed, so people began hanging
on to their software longer because the upgrades were expensive and the
old stuff met their needs. So, Microsoft (and much of the rest of the software
business) tried a new marketing paradigm: planned obsolescence through
the continual reinvention of the wheel. That is, Microsoft tried to ensure
that Windows 3.1 software, would not run under Windows 98; and that Windows
95 software would not run under Windows 2000; and so on. But now, this
strategy is beginning to run aground too, because many businesses have
huge investments in Windows 95 software. Microsoft cannot abandon it’s
business customers outright, and so cannot totally abandon Windows 95.
Also, Microsoft cannot abandon it’s old customers because the power of
the new CPU’s makes multiple OSs and emulation possible. Understand, software
written for Unix, or one of it’s flavors like Linux, does not become obsolete
overnight because the OS has so many developers. Unix is not owned by one
big corporation, So, Big Corp cannot decree a migration to the new version.
And, planned obsolescence just makes Unix a more viable competitor for
Windows. Microsoft’s planned obsolescence strategy for Windows is becoming
a threat to Microsoft; not just to the people who have an investment in
the Windows operating system. But, Microsoft’s managers may not realize
this, or what their real marketing options are. It may be that they are
thinking in terms of marketing consumer products, like soap. Or, maybe
they are trying to maintain the profit levels Microsoft had years ago.
In the end, markets change with time, and successful managers adapt their
strategies. Can Microsoft adapt to changing market realities?
Finally
I think the dynamics of the Software Business is similar to a thunderstorm.
A thunderstorm can grow to enormous size and power on a summer afternoon,
and then collapse in the evening. Similarly, a software company can grow
to earth shaking size in a few years, and then collapse just as fast. Totally
new, not merely improved, technology is the growth driver, because new
hardware requires new software. But, software is intangible, it’s value
is based on need. Software that cost $200 dollars ten years ago, might
today be bought for a dollar. Fifteen years ago, there was a lot of revolutionary
technology coming down the pipeline. Today, I don’t see as much of this
kind technology in the pipeline. But, I do hear allot of hype about things
that are merely improved, not new. So, if Microsoft is depending on growth
based on growing technology, they may not grow as much as they expect.
If Microsoft is basing it’s strategy on planned obsolescence, they could
have a disaster. If Microsoft’s managers think that aggressive business
practices are the way to increase profits, they are mistaken. In the end,
I think the solutions for Microsoft’s problems are right in front of it’s
managers noses. It starts with an understanding of the fundamental principles
of economics. It ends with an understanding of business ethics. If they
do not see that, then Microsoft deserves whatever the federal courts dish
out.
I hope you enjoyed reading this.
December 3, 1999 - July 27, 2000