What is it that the U.S. is doing right and Japan doing wrong in the information technology business?
The diverging fortunes of Oki Electric and EMC Corp. illustrate the differences crisply.
A fable of two companies
By Neil Weinberg
THE NEW PRESIDENT of Japan's Oki Electric Industry, Katsumasa Shinozuka, speaks with evangelical zeal about computer-telephony integration. Swapping data between computer and phone systems
for mail order, home banking and other service providers
is an ideal growth business for a computer and telecom equipment maker like Oki, he says.
"It will be our strongest business," boasts the 58-year-old Shinozuka. "The company that takes concrete action the fastest is going to win, and we're already marketing products."
There's only one thing wrong with this picture: Computer-telephony integration is nothing new. In the U.S., vendors like Genesys Telecommunications Laboratories have been pushing it for several years. In Japan, where even voice mail is still rare, the business is just getting off the ground.
This pattern is all too familiar: An information technology percolates out of the U.S.; Japanese firms pile in late as resellers or with their own me-too versions and kill themselves in a price war while U.S. entrepreneurs move on to new technologies.
This is just what happened in memory chips. Giants Hitachi and Toshiba posted their first losses in a half-century in the six months through September. Smaller Oki, with less than $6 billion in revenues, expects to lose a record $375 million in the year through next March.
The problem runs deeper than Asia's recession. Oki and its compatriots are hardware-centric generalists in an era when wealth is being created by turning out new, narrowly focused technology—not me-too electronic boxes.
Consider Intel in microprocessors, Cisco Systems in networking equipment and EMC Corp. in corporate computer data storage. In particular, why does EMC do so well, with a 35% market share in its line of business and a 19% net profit margin? Because the technology is changing rapidly enough to demand ingenuity and imagination, and because there is a software component to it.
EMC's sales grew an average of 50% a year from 1990 to $2.9 billion in 1997. EMC's market capitalization, meanwhile, has risen 227-fold to $34 billion, or 12 times last year's sales. In the first three quarters of 1998 EMC netted $537 million on revenues of $2.8 billion.
Contrast Oki. It's like a lot of other Japanese electronics firms, a sort of department store of technology. Oki produces computers, telecom equipment, semiconductors and dozens of other products. It leads in just one market: automated teller machines. Japanese banks aren't great growth prospects these days, though Oki sees some potential in convenience stores and other locations.
Indeed, Oki's $5.8 billion fiscal 1997 sales have barely budged since 1990. Its finances are a mess. In addition to the expected record loss, debt is 2.1 times equity, versus 38% for America's top 25 electronics makers. Oki's stock is down 80% this decade, and its $1.3 billion market cap is just 0.2 times sales.
It's not that Oki lacks smart engineers. It was founded in 1881 as Japan's first telecom equipment maker and turned out a phone just five years after Alexander Graham Bell invented the device. As a leading Nippon Telegraph & Telephone supplier, it attracts top engineers and rates high for quality.
Nor is management oblivious of Oki's ailments. In 1992, when the firm was also in the red, then-president Jun Jinguji admitted publicly that Oki was unlikely to ever make money in semiconductors, competing with far larger rivals. Six years on, Oki still makes chips. In 1995 President Shiko Sawamura said Oki must become like a "specialty shop" amid giant department store rivals. It hasn't quit one major business since.
Time and again management has placated employees rather than cut moneylosing businesses. Little wonder. Top executives own almost no shares, and foreigners just 2.9% of the firm. Oki's top seven shareholders are Japanese financial institutions
hardly angels of wealth creation.
------------------------------------------------------------------------
"Under [Japanese-style] consensus management we wouldn't have taken on the high risk of open systems."
------------------------------------------------------------------------
When Oki dropped its $375 million red ink bombshell in late September, Shinozuka said he would cut his payroll by 10%, or 2,700 people, and abandon plans to mass-produce next-generation chips.
There's less to the plan than meets the eye. Hundreds of the 2,700 will be shifted from Oki's parent company to subsidiaries. Oki will keep its memory-chip operation in business to supply system LSIs
chips packing several functions on one silicon wafer. Guess what? Hitachi, Toshiba, NEC and Fujitsu are hot and heavy on system LSIs, too.
Shinozuka calls his revival plan "Phoenix 21" and has set up a new Silicon Valley office to tap into industry buzz. Strip away the slogans and protecting turf is still priority one. Shinozuka diplomatically predicts that Oki's computer, telecom and chip units will account for the same shares of sales in three years as they do today. So much for computer-telephony integration coming to the rescue.
What about Oki's perennial loss-makers? "If a unit isn't making money, we'll transfer it to a group company, and if it still loses money, we'll close it," says Shinozuka. "We have to manage in a proper Japanese way or we won't be appreciated by society."
As Oki flounders in a proper way, EMC is enjoying textbook growth. Richard Egan, an early Andy Grove protégé at Intel, and former college buddy Roger Marino set up EMC in 1979 to build add-on memory for minicomputers. Rather than stick with a dying business, they moved into mainframe data storage in 1989, devising a technology far faster than what IBM was then using to control 80% of the market. The technology, called RAID (redundant arrays of independent disks), involves clever algorithms for dealing out data among an assortment of hard-disk platters. Within a few years EMC had won away the top share.
As corporations began deploying open client-server systems in recent years, many attached storage devices to their local servers. Pundits predicted the death of the mainframe. Turns out scattering databases among local servers leads to lots of problems managing and accessing information.
EMC recognized early on that open systems were here to stay but that mainframes and centralized data storage still had big advantages. It invested heavily in technology that offers more compatibility than anyone else's among open systems, mainframes and other computer systems.
Like Microsoft, Intel and other industry leaders, EMC has used acquisitions as a strategic weapon. Epic Systems, a data-backup specialist it bought in 1993, helped EMC add the world's fastest open-system backup technology to its arsenal. Oki Electric, by contrast, hews to the Japanese model of relying largely on organic growth or none at all.
------------------------------------------------------------------------
"We have to manage in a proper Japanese way to be appreciated by society."
------------------------------------------------------------------------
In 1997, EMC's third year in open-system data storage, sales hit $1.5 billion. Stock options and quarterly cash bonuses give EMC executives and salespeople plenty of incentive to hit their targets. EMC is again pushing the boundaries of its industry with a new technology—fibre channel. It leads in supplying data storage devices that connect to host computers via optical fibre networks to transmit masses of data over long distances without clogging up other operations.
"We're in the storage business, but have redefined it from mainframes to Unix and NT systems, and increasingly to software," says EMC Chief Executive Michael Ruettgers, 55. "Under [Japanese-style] consensus management we probably wouldn't have taken on the high risk of open systems, given that we already had a comfortable mainframe business."
Ruettgers plans to stick to his knitting. Why not? He sees EMC's data-storage revenues tripling to $10 billion by 2001. Oki, meanwhile, may or may not get some growth out of its new telephone gear and remains as unfocused as it ever was.
The Japanese are "going to run away with the world computer market. It's going to be another TV industry." That's what Clyde Prestowitz, former Commerce Department counselor for Japanese affairs, told Business Week in October 1989. How wrong he was. |
