Software players tied to hardware - Japanese software industry

SOFTWARE PLAYERS TIED TO HARDWARE

The Oct. 23 edition of Business Week said that the mantra, "the Japanese do not understand marketing, the computer industry moves too fast for the Japanese, the Japanese can't write software, cannot protect the U.S. computer industry much longer.

The Japanese computer industry is steadily moving up the technology food chain, ranging over memory chips and I/O devices, PCs and workstations to supercomputers. With the tremendous growth of Japanese financial power, evident more and more in the purchase of U.S. companies, Japan is now casting a dark shadow over the future of the U.S. software industry.

But can Japan really export software? Most people connected with the Japanese computer industry are more likely to believe that three-eyed aliens really did land in Russia. Americans knowledgeable about the condition of the industry in Japan are likely to be of the same opinion.

But Japanese-made software is already popping up, alien-like, in the U.S.--not in the office but in the home.

Nintendo of America, with sales of more than $1.5 billion in the U.S., is taking the overwhelming share of the computer game market. And half of these sales are in software.

Nintendo's U.S. operations began with arcade game machines in the early 1980s. Nintendo Entertainment Systems (NES) entered the picture in 1985, and it has since resuscitated the game machine market, which crashed when Atari faltered.

The company's success is explained by a marketing strategy that combines cheap, but high-quality, hardware with good software that is well-matched to American tastes, and strong service.

But Nintendo is not simply a maker of toys. Its success has been based mainly on its marketing and software development ability.

Despite long-held and sometimes erroneous beliefs about the Japanese software industry, it will not be long before the Japanese appear in the international market.

However, it will take time. In Nintendo's case, the right conditions were present in the domestic Japanese market--stiff competition necessitating products which were high in quality, low in price and original in software. But these conditions do not prevail in the software industry today.

THE CURRENT SITUATION

The Japanese hardware market is next in size to that of the U.S., and is growing rapidly. In 1987, 6,200 mainframes and 300,000 general-purpose computers were running, along with 400,000 small business computers and over 2 million 16-bit PCs. These figures represent an annual growth rate of 20%.

Against this hardware sales background, the software industry has been growing, but has not been able to meet burgeoning domestic demand. Industry sales in 1987 totalled $7.6 billion ($SUS1 = 145yen), according to market sources.

With annual growth rates in the industry exceeding 20%, sales have likely already passed the $10 billion mark. However, software products themselves accounted for no more than 13% of total sales, and only passed the $1 billion mark around 1987. The other 87% or so of sales come from subcontract software development and facility management.

The product sales breakdown is 50% originals, 40% imports and 10% products developed by other companies, according to published research.

WORKER SHORTAGE A "CRISIS"

If custom programs are to be developed to meet the demand for software, an increasing number of systems engineers and programmers will be needed. The biggest single problem in the industry is a shortage of workers.

According to the Ministry of International Trade and Industry (Miti) forecasts on the balance of supply and demand for software engineers, the shortfall will be 251,000 in 1990, 512,000 in 1995, and by the year 2000 will top 965,000 against a demand for some 2.14 million. In Japan, this is referred to as "the software crisis."

However, salary levels for systems engineers and programmers in Japan are high. It is difficult to get software engineers of a suitable standard at salaries of even $100,000 a year. Considering that the growth rate is low--2% or 3% on sales of $70,000 per person--it is clear that the software industry is not in good business shape.

A further problem is that programmers vary widely in ability. Already many companies have plans to import programmers from South Korea and Southeast Asia in the guise of "technical trainees," but even this is unlikely to fill the gap.

Many point out that the worker shortage is in fact caused by low productivity in the software industry, the inability of the industry to satisfy the packaged software market, and the closed nature of computer distribution channels in Japan.

DEVELOPMENT TIED TO HARDWARE

Software development, except in cases where users develop their own software, is normally closely tied with the manufacturer supplying the hardware, and is ordered direct by the manufacturer to the software houses.

Both the manufacturers and the users prefer customized products. Manufacturers lock the users into their hardware, while DP managers in client companies rest easy because their bosses are unable to compare their systems with those of the competitors.

The software industry has not really become an independent industry in Japan. Miti is reluctant to talk about the "information services industry," which includes data processing services and so forth. But PC software aside, there is simply no Japanese company whose principal business is packaged software.

Neither are there many companies which concentrate on subcontracted systems development. The bigger the company, the higher the ratio of sales deriving from facility management and hardware.

Software houses are virtually all affiliates of computer manufacturers, or if not, they are the sections of computing divisions of large users. Although they are nominally "independent," the majority reap most of their business from servicing the systems of the parent company.

This year's "information services industry's rankings has NTT Data Communications at the top with revenues of $1.629 billion, which compares to a company like EDs in the U.S. The company is an offshoot of NTT, which more or less monopolizes the communications market.

In the number two position with sales of $587 million is the Nomura Research Institute, a member of the Nomura Securities group, the largest securities trader in the world. Number three, NBC with results worth $585 million, is a Hitachi subsidiary.

Only at number four do we find the biggest of the independents, CSK, with sales of $412 million. However, software development and marketing account for only 54% of its figure. It started out in the facilities management business, and the segment still nets 20% of its sales.

With the current state of affairs in Japan, the thought of exporting software can no longer be a dream. High salaries, low productivity, a poor development environment, and ordering procedures typical of the closed Japanese distribution system--and, on top of that, the language barrier--make it imperative.

But what conditions would allow Japan to take its place in the world software market?

First is the presence of competitive domestic industries which dearly need advanced data processing. Japan presents a unique marketing opportunity to the software industry because of its powerful manufacturing sector and growing service sector.

Countries with small domestic markets are unlikely to become exporters, but this is hardly the case with Japan. The exporters of programmer labor are not exporting products and services.

The second condition is a situation wherein productivity in the software industry grows to equal, current productivity levels in manufacturing industries. The introduction of Case has provided the opportunity.

Few observers of the Japanese software scene believe that anything great will come from the Miti's Software Industrialized Generator and Maintenance Aids (Sigma) project itself.

However, everyone involved with computers in Japan recognizes the challenge to improve software development productivity, and the importance of Unix as a standard tool. The influence of the Sigma campaign on users has been profound.

In the third condition, the U.S. software industry would have to fall into stagnation. If the U.S. industry loses its leadership position in the innovation of software production--a position attained by its superior productivity--by using low-cost labor, then the Japanese can surely overtake the U.S. quickly.

In sum, two things could change to create the conditions needed for the Japanese software industry to rise to prominence. The Japanese software industry could introduce factory production methods to achieve high product quality, high productivity and lower costs. The U.S. might then turn to offshore programming to take advantage of even lower-cost labor.

But if the U.S. does this, it will be unable to bring in innovative new systems to further improve productivity, and U.S. companies will be unable to offer interesting positions to creative engineers.

Does this sound familiar? It will be the semiconductor industry all over again.

Until now, the U.S. software industry has looked like the stratosphere from the Japanese perspective. But the Japanese have little respect for those who refuse to teach them.

The U.S. computer industry was dealt a grievous blow by the political setting of semiconductor prices by the U.S. and Japanese governments.

In Japan, semiconductors are termed "the rice of industry." They are respected in the way that the people of Japan respect rice as their staple food.

Unfortunately, semiconductors became the object of politically motivated protection in the U.S., in the same way that Japanese rice is politically protected. But what U.S. industry has to pay for semiconductors is a heavier burden than the cost of rice to the Japanese consumer.

CHALLENGE BEGINS IN FACTORIES

The Japanese challenge to world software markets will begin in factory systems, design systems, and will then spread to financial systems and will ultimately turn to office systems.

Japanese manufacturers are now acting to commercialize products on the basis of their own development programs. In the case of office systems, the current situation is not encouraging (Japan's detractors say that there are no office systems in Japan), and it will probably take 10 years for packaged software to be as important in the marketplace as it is in the U.S. today.

MEETING FUTURE NEEDS

Masahiko Kobayashi, major project development team leader in Miti's Agency of Industrial Science and Technology, says that if Japan proceeds as planned to improve software productivity in the short term, "After 10 years, a Japanese style of software will emerge, which will meet fully the needs of the domestic market."

Kobayashi adds, however, "I have no idea if it will become popular in foreign markets."

Certainly, Japanese manufacturers are good at upgrading product quality by constantly making minor improvements. This approach has been successful with consumer products. But will it work with software?

Software development is a field in which originality is needed; success for the Japanese software industry is a decade away. The Japanese educational system is too rigid to nurture creative talent. Further, the Japanese have become less diligent than in the past. Programmer standards in other parts of Asia are consistently better.

Thus, given Japan's strength and weakness, the Japanese software industry will probably appear on the world stage as a combination of Japanese production control systems, Asia's best programmers and the creativity of U.S. engineers.

The response of the U.S. software industry will undoubtedly be to move into Japanese markets rapidly. The Japanese have five years to become aware of these latent forces and act.

What might happen in the United States five years hence is a problem. But whatever does happen, it seems certain that the software industry of the '90s and after will be both an international and multinational industry.

Kamata is president of Soken Planning, Co., Ltd. of Tokyo, and publisher/editor of "Corporate Publishing Report: The Newsletter of Corporate Electronic Publishing Systems."

COPYRIGHT 1989 Wiesner Publications, Inc.
COPYRIGHT 2004 Gale Group