Japan's Computer Industry:Detroit Revisited?
In this first installment of a two-part series, Scott Nash explains why
Japan's previously successful approach to industrial production will not
carry over to the computing world. Part two (scheduled for next month)
will cover the remedial steps that can help Japanese companies to avoid
outright ruin.
by Scott Nash
Thirty years ago, competition was stiff among the Detroit automobile
makers. Every year, they competed to see which could turn out the biggest
cars with the most add-ons (and thus earn the highest profits). The crazy
thing was that this was the acknowledged path to success in the American
automobile market.
This "competition" came to an end in the mid-1970s, though - the rules of
the game changed with the "oil shocks," and Japanese makers came to market
with cars that were smaller, cheaper, and a better fit to the public's new
needs. It took the tradition-bound American manufacturers nearly two
decades to recover, and they are only now gaining lost ground.
A similar scenario has unfolded in the Japanese computer industry. During
the '80s, Japanese computer companies fought to deliver the biggest
roadsters on the block - proprietary mainframes, with huge margins and
subcontractors dedicated to programming them. This lasted until about two
years ago, when the "Compaq shock" served notice that there were new
players in the game. American companies - honed by years of bruising
competition in their home market - came into Japan offering machines that
were smaller, cheaper, and a better fit to the Japanese computing public's
needs.
Betting on the average
For most consumers, Japan is synonymous with electronics, and superiority
in computers logically follows - or so goes the common wisdom. Japan has
been a success in industrial production, after all, and software
production is just another market. What this scenario ignores, however, is
that the process of software development is a lithe and agile beast, very
different from the elephantine course of industrial production. If Japan
attempts to attack the computer market with the methods that were
successful for industrial production, failure is preordained.
When I joined my Japanese computer consulting firm, I did so as a normal
company entrant, which means that I underwent the normal first-year
training. This training, and my continuing work experiences, have made one
thing clear: Japan, flush from success in the industrial markets, is
attempting to apply the same educational and business approaches to the
computing market. Yet the methods that made Japan an industrial power are
totally ill-suited to success in the computer industry. Few people
(Japanese or foreigners) can see this, though, and fewer still realize
that Japan, as we currently know it, will not survive the computer age.
There are two fundamental problems: education about computers and
business practices in computer companies. The Japanese educational system
has succeeded quite well in bringing the Japanese people to the highest
level of education in the world, and it has transformed Japan into an
industrial power. However, there are three flaws in the Japanese
educational system as it relates to computers: everyone is taught at the
same level, regardless of ability; students are taught to memorize,
instead of asking "Why?"; and creativity is discouraged.
A centerpoint of the Japanese methodology is that Japanese teachers teach
to the average. Education is based on the group principle: everyone learns
at the same rate. In the school classroom, everyone moves to the next
grade together, even the "slow learners." This method ensures that all
students are brought up (or down) to the same level of ability, and it has
supplied Japan with, on average, highly educated industrial workers.
When this system is applied to writing software, though, it means that
the potentially most talented programmers are kept from reaching their
full potential. During training in my company, for example, our first
programming assignment was: draw a flowchart, get it checked, write out
all of the code by hand (anyone who has ever worked with a verbose
language like COBOL knows what fun this is), have the instructor tell you
which lines weren't indented the way he liked, rewrite the code, then
finally enter the code into the computer. This process was the same for
all members of the class, from rank beginner to expert. The funny (or sad)
thing was that the Japanese expert programmers, of whom there were three
or four in the class, didn't complain, even though they could have been
spending their time learning more about the language instead of wasting it
making sure indentations matched up on handwritten code.
This system, which enforces the average, hurts programming abilities.
Egalitarian ideal or not, some people are more skilled at programming than
others, and if Japanese educators do not challenge and encourage these
talented people, Japanese computer firms will remain "average." (Which, in
the computer world, means they will fall further and further behind.)
Learning everything, knowing nothing
Also troubling is the inability of most Japanese to ask "Why?" The
Japanese educational system emphasizes rote memorization, not
consideration of cause and effect, which strengthens analytical abilities.
Furthermore, in Japan, questioning the old methods is deemed impolite, and
rarely done. Improvement and advancement in computers comes from
constantly questioning, though - throwing out those old methods that do
not work. Relying on old methods soon leads to stagnation.
Nowhere is the Japanese attitude toward learning more evident than in the
Information Management exam given every spring and fall. Most of our
training was directed toward getting us to pass this exam, the standard in
Japan. You might imagine that with this high regard, the test must truly
be awesome, the paragon of information technology. In fact, it was revised
just this year to keep up with the cutting edge in technology - by adding
cursory questions about DOS and UNIX.
Japanese beginners are being forced to study technology that is soon to
be (if not already) obsolete. And the Japanese still study diligently,
without thinking to ask why they need to. What better way to ensure that
Japanese firms will never be able to compete with the American computer
firms?
The problem is that Japan is single-mindedly applying the method that
made it successful in manufacturing: ensuring that all members of the team
are interchangeable because they all have knowledge of all areas. However,
whereas in manufacturing, the laws of physics and chemistry never change,
computer technology is changing constantly. For computers, the American
system of a person being a specialist instead of a generalist is the
necessary solution.
The Japanese approach is akin to teaching a class and spending the first
50 minutes explaining not only where the pencil sharpener is, but also the
detailed physics behind how the rotors inside work. At the end of this
time, you realize you only have 10 minutes left to teach the students what
they can do with the pencil. If this style of teaching is not revised,
Japanese companies are going to have a very hard time adapting to changes
in the marketplace because there will be no experts. People will have
spent so much time learning "everything" that they will understand only
the surface aspects of each subject.
The Japanese have long been criticized for a lack of creativity. After
looking at many of the shareware offerings written by Japanese
programmers, however, I would argue that good Japanese programmers are
perfectly capable of producing creative solutions. It is the corporate
organization that is cramping the style, not a lack of individual
creativity.
Risking the future by relying on the past
In the workplace, the problems stemming from educational practices are
magnified by the hallowed corporate traditions: the seniority system, the
sempai-kohai ("senior-junior") relationships, and a lack of visionaries.
Much has been written about the Japanese business system - most of it
laudatory. And the system did work well when Japan was primarily a
manufacturing power and expanding at an 8% clip each year. But it no
longer applies to the new business realities; like a roulette player who
has been fantastically lucky, Japan's businesses are betting the farm on a
short-term aberration.
One basic problem in the Japanese office is the traditional system of
advancement and promotion. Advancement in Japanese companies is based
primarily on age, not on ability. In manufacturing firms, this creates a
well-rounded executive who understands all aspects of the business through
long experience. But, in the computer business, when a fifty year-old
executive who has worked only with mainframes must make strategic plans
for the new downsized era, chaos soon ensues.
The roots of the problem also lie in the sempai-kohai relationship of
workers, another frequently lauded aspect of Japanese business. Since
their days in high school sports clubs, lowly freshmen have cleaned the
senior's bats and picked up after the games, in gratitude for receiving
guidance from the seniors. What this means in the computer field is that
new hires are typically assigned to routine code maintenance for prior
systems, even those who are skilled at writing programs. It's the
"seniors" who do the heavy programming. By the time those freshmen become
seniors, their creativity and vitality has left them after years of non-
creative grunt work.
These two influences combine to cause the most serious problem that
Japanese computer companies face: a lack of vision. Lack of vision may not
have been a handicap to build next year's new car model - it's still going
to have a steering wheel and four tires. But the electronics and computer
industries are different. You can already see the effects in the chip
market, where after the initial (copycat) Japanese onslaught, Intel has
regained the number one spot in the world by taking large risks and
concentrating on costly CPU design and production. Having vision gives the
conviction to take "make or break" risks. Detroit learned this lesson
well twenty years ago.
Japan at the crossroads
Not all things about Japanese business methods are bad; there are some
things that we can learn from Japan about designing software. Japanese
closeness with their customers and long-term relationships mean that they
can really understand customer needs fully - and thus, design solutions
that are the better suited in a shorter time. Furthermore, the concept of
"kaizen," or incremental improvements, has produced some great shareware
enhancements to both PC and Mac systems. On the balance, however, the
combination of educational factors (a lack of creativity in business
programming, teaching to the lowest common denominator, and little
emphasis on analytical skills) combined with business factors (the
seniority system, sempai-kohai arrangement, and lack of vision) will
thwart Japan's attempts to turn its computer companies into world-class
powers.
On the current course, a change in Japan's circumstances will only come
in two ways, both difficult choices. The hard (and less likely) way will
be for Japan to realize it must make fundamental changes to its culture
and way of thinking to adapt to a changing market. The harder way will be
for today's large companies to slowly fade to obsolescence while newer
startups succeed, thereby causing societal and corporate upheaval. In next
month's conclusion, I'll discuss the changes that you can implement within
your company to help avoid the disastrous effects of these gruesome
scenarios.
