Businessweek Archives

Fujitsu Gets Wired (Int'l Edition)


International -- Cover Story

FUJITSU GETS WIRED (int'l edition)

The company is staking its future on the still elusive frontiers of cyberspace

Nothing could be more "California" than a wedding in cyberspace. Last Valentine's Day, groom Joseph Perling sat on Venice Beach before his laptop. His bride, Victoria Vaughn, logged on from nearby Hollywood. The Reverend R. John Perling, the groom's father, presided from his Beverly Hills church. As friends tuned in from around the country, the couple assumed online personas and exchanged lawful vows in a virtual chapel in CompuServe Inc.'s Worlds-Away community.

WorldsAway, a hip cyberhangout with some virtual-reality-like features, is among the 20 most popular forums on CompuServe. And guess what? It's owned and operated by Japan's biggest computer company, Fujitsu Ltd.

Once a flat-footed follower of more with-it American companies, Fujitsu is now staking its future on the shimmering, still elusive frontiers of cyberspace. By 1998, the $34 billion company, the world's second-biggest computer maker after IBM, aims to get as much as 30% of its revenues from Internet-related services, software, and hardware. To get there, Fujitsu's tough 64-year-old President Tadashi Sekizawa is plowing nearly half the company's $3.6 billion research-and-development budget into Internet and multimedia-related products and services.

Of all Japan's high-tech powerhouses, Fujitsu seems best equipped to cross swords with America's Internet warriors. Long Japan's most aggressive computer company, it has now embarked on a cultural transformation to break out of the gray-suit syndrome. Already, scattered clusters of researchers and engineers choose their own hours, specific interests, and apparel, resembling nothing so much as free-wheeling Silicon Valley garage-shop denizens. To compete more effectively with lean rivals overseas, the company that once balked at building systems using other companies' electronic parts now imports close to $2 billion a year in low-priced Asian components.

If Fujitsu can pull off this transformation, it will challenge conventional wisdom in the Valley, and elsewhere, that the Japanese are destined to miss the hottest action of the Networked Age. Having watched the Japanese flub most other forays into the U.S. computer-systems market, veteran technology-watchers wonder how they can be players on the fast-moving Internet.

LEAPFROGGING RIVALS. Fujitsu's youthful staff bristles at such glib assessments. Fujitsu can win, they maintain, because it has the necessary strengths in three key areas: semiconductors, computers, and telecom. Outside Japan, computers are still a weak link for Fujitsu. But nobody takes the company's design and manufacturing skills lightly. A world-class communications giant, Fujitsu has, for example, begun test-driving next-generation digital switches in North Carolina's statewide Information Superhighway. Even IBM, so far, has failed to integrate this triad of digital skills.

Recent aggressive moves in the Japanese market have helped repair Fujitsu's tarnished image in personal computers. And its research labs boast a stunning array of next-generation components. Leapfrogging its competitors, Fujitsu is getting set to roll out 42-inch-wide flat-screen plasma displays that could evolve into wall-hanging TVs. Someday, these could be bright windows onto the Infobahn. Nor is software immune to Fujitsu's ambitions. As part of a new agreement with Computer Associates International, America's third-largest software company, Fujitsu is charging into database software aimed at the World Wide Web, going up against the likes of Steven P. Jobs and Oracle Corp.'s Lawrence J. Ellison (table).

In Japan, Fujitsu is readying some dazzling, network-based software products--from virtual-reality games to business applications for the World Wide Web--that are as exotic as anything reached by a click on a Netscape directory. Says Sekizawa: "We have to discover a whole new path." Agrees board director Tatsuzumi Furukawa, Fujitsu's point man on cyberspace and a likely successor to Sekizawa: "We need to overhaul the company."

Fujitsu does face some crucial challenges. One of its most vital American assets--mainframe maker Amdahl Corp., in which Fujitsu holds a 43.5% stake--is steadily losing market share to archrival IBM. Likewise, Fujitsu controls ICL of Britain, whose performance has been disappointing. Partly because of these frustrations, 70% of Fujitsu's sales and most of its profits are still earned mainly in the home market.

But Fujitsu's strong position in Japan is what is bankrolling its cyberspace effort. Now, thanks to a massive, nationwide splurge on information technology, profits are flooding Fujitsu's coffers. Morgan Stanley & Co. estimates Japan's spending on information and telecommunications hardware and services will expand from $294 billion in 1993 to $619 billion in 2000 to $1 trillion in 2010, and Fujitsu stands to benefit handsomely.

Fujitsu never bled red ink the way IBM did. But like Big Blue, Fujitsu is resigned to the fact that its mainframe computer business will gradually shrink as customers opt for much cheaper networks of servers and PCs running off-the-shelf software. But this trend, known as "clientserver" computing, has taken shape more slowly in Japan than in America, giving Fujitsu a chance to adjust.

A crucial source of profits at home is Fujitsu's 16,000-strong army of system engineers--the largest such battalion in Japan. As Japanese investment by banks, insurance companies, and manufacturers starts to crest, this army is busy setting up and running networks that link mainframes, workstations, and PCs. If some of the hardware comes from various vendors, that's fine with Fujitsu, since the margins on making such diverse gear work smoothly together exceeds those of selling hardware alone.

Since Fujitsu is a trusted supplier to giants such as Toyota Motor, Mazda Motor, and Nippon Telegraph & Telephone, it has a leg up on rival suppliers when these customers start experimenting with new computing concepts, such as "intranets"--the use of an internal Web site as a companywide information nexus. The North American market for intranets is already fiercely competitive, dominated by the likes of Sun Microsystems Inc., and Netscape Communications Corp.

EASY MIX. Can Fujitsu join their league in the U.S.? Possibly, if it can leverage its huge core strengths at home. Ten years ago, for example, the company opened the nation's premier online service, called NIFTY-Serve. Some 1.6 million Japanese now log on regularly, and the number could hit 2.3 million by yearend. That service relies on a high-speed network called FENICS, which Fujitsu developed for mainframe clients. Now, the company wants to expand the FENICS network directly to North America and Southeast Asia. Fujitsu already runs CompuServe's online services in Australia, Korea, and Taiwan.

Alliances are now easing Fujitsu's entry into the U.S. software market as well. Its new Jasmine program relies on reusable chunks of code known as "objects." This approach reduces tedious programming for the customer and makes it easier to mix images and text in databases constructed for the graphics-rich World Wide Web.

But it wasn't dreamed up in a boutique. Instead, the database software grew out of a collaboration with carmaker Mazda Motor Corp., one of Fujitsu's mainframe customers. "It's wrong to think that the mainframe world is simply being eaten away by PCs," Furukawa argues. "Rather, the question is how PC users can get the most from assets developed for mainframes."

To get the product to American desktops, Fujitsu is depending on Computer Associates, which has added a snazzy new user interface and has assumed responsibility for U.S. marketing--an area where Fujitsu is weak. "We can't compete on a global basis in software without making use of top technical people in the U.S.," says Furukawa.

Collaboration is also the linchpin in Fujitsu's U.S. telecom strategy. And its early partnerships in this area could soon start blossoming into major contracts for large, fiber-based digital switches and transmission gear that will become the backbone for America's Information Superhighway. For a decade, Fujitsu, NEC, and Hitachi have been funneling billions of dollars in R&D funds into a high-speed switching technology called "asynchronous transfer mode," which handles images, sound, and text with equal ease. Three years ago, North Carolina began installing Fujitsu switches as part of the new infrastructure linking schools, government offices, hospitals, and prisons.

The highway's architects, mainly engineers at Bell South, Sprint, and GTE, picked Fujitsu over AT&T, Northern Telecom, and others, in part because the price was "extremely competitive," according to one source. Today, the highway links 74 schools and hundreds of other sites. Students in remote areas of the state, for example, can attend classes via video-equipped computer terminals at the state's top universities. Best of all for Fujitsu, the project has become a technology showcase for other states--and other countries. "We're at the forefront of fiber-optic technology," says John Friedrick, executive director of the North Carolina School of Science & Mathematics in Durham. "We've had visitors from South America, Israel, Australia, and South Korea."

Ultimately, Sekizawa's greatest hurdles may be cultural. He argues that fundamental changes are needed to free up creativity and entrepreneurship, and that's prompting the most drastic cultural overhaul in the company's history. At a Tokyo development center of Shuzo Morita, director of Fujitsu's thrust into new computer interfaces, long-haired Japanese hackers and animators rub shoulders with gray-suited engineers to create games and entirely new 3-D software domains. Some are inhabited by exotic creatures that evolve and interact with humans across the computer screen. Young software developers here have formed a "virtual work group" with colleagues scattered across Japan and in the U.S. And English is the lingua franca of this internal, networked community.

TASKMASTER. President Sekizawa could be the man to force Fujitsu into the future. A telecommunications engineer who graduated from prestigious Tokyo University, Sekizawa set out in June, 1990, when he became president, to make the company more profitable and agile, slimming down an organization that had grown too fat and comfortable during the Japanese bubble economy of the late 1980s. A dour taskmaster who seldom expresses satisfaction to the troops, he has cut nearly 10% of the parent company's workforce since 1993. He also launched a flatter managerial structure that links pay more closely to performance. Some software developers, for example, participate in a program called "spirit," in which overtime pay is replaced by productivity-linked pay. And staff with good ideas are encouraged to consider spinning off venture businesses.

One of Sekizawa's most crucial decisions was to promote an "open systems" strategy, which lets customers integrate hardware and software from different vendors. He also stepped up overseas sourcing and assembly in Southeast Asia. The result: a less arrogant, more market-oriented company better able to respond to tougher U.S. rivals. "The company has gotten leaner, and now they're addressing the areas of growth," says Barclays de Zoete Wedd analyst David Benda in Tokyo. "They've got a very good chance of making it."

Fujitsu's restructuring, moreover, suggests that Japanese companies are learning to make a virtue of social constraints against massive layoffs and radical restructuring. While Fujitsu's staff cutbacks are aggressive by Japanese standards, they pale in comparison with American competitors. And while U.S. companies such as AT&T have split themselves into smaller, more focused units, Fujitsu and other Japanese giants remain vertically integrated. This gives Fujitsu a broad array of technologies that no American competitor can match in-house.

Japanese computer makers haven't been as nimble as some of their more niche-oriented U.S. competitors, but they've retained strong financial positions that let them cherry-pick technologies that are created elsewhere. WorldsAway was developed in 1986 by Lucasfilm Ltd. of the U.S. Fujitsu acquired the idea and then hired away its creators. The operation is now run and maintained by Fujitsu's subsidiary in San Jose, Calif. More than 15,000 subscribers, intrigued by this extension of "chat," log on via CompuServe in 147 countries around the world. "I have a lot of respect for what Fujitsu has been doing in this area," says Linda Stone, director of Microsoft's Virtual Worlds Group, which is promoting a similar product.

Fujitsu is also looking for ways to compensate for its lack of entertainment content. Nerdy even by Japanese standards, Fujitsu "doesn't even have employees who can talk about movies," concedes Furukawa. But the company recently invested $50 million in a major Japanese film studio, Shochiku Co. And Kazuto Kojima, director in charge of the digital media group, has been nosing about the world's great content makers. He's visited 20th Century-Fox Film Corp., Paramount Communications Inc., and other Hollywood studios, as well as European and Japanese filmmakers. "I'm not interested in investing big money. I'm just seeing whether there's any way in which telecommunications, computer, and film companies can work together in the future," Kojima says. He hopes to increase Fujitsu's content business to $100 million by 1998.

Fujitsu's most audacious gamble is its bid for PC market share--something most Fujitsu managers believe is critical to the company's success in cyberspace. Its first stab at a multimedia PC, called FM Towns, was brilliant technically but never took off. So in 1993, Fujitsu joined other Japanese PC makers selling machines compatible with DOS V, a Japanese-language version of DOS that runs Microsoft Windows.

PRICE WAR. Fujitsu's real assault came a year later. In a rebellion against its own paternalistic traditions, Fujitsu dumped many of its long-time Japanese suppliers for American and Southeast Asian rivals selling components for 20% to 30% less. The company also began building peripherals in Thailand and assembling finished PCs in Taiwan. Next, it will start assembling PCs in Hillsboro, Ore. Non-Japanese parts in many of its PCs already exceed 90%.

The overhaul brought costs down. But to build a big share in Japan, Fujitsu still had to undercut the competition. So it invested hundreds of millions of dollars in advertising and distribution. Today, analysts reckon Fujitsu is losing $300 on each $2,000 machine it sells--or as much as $1 million each day. The result: Last year Fujitsu doubled its Japanese market share to 18.2%, becoming the No.2 vendor after NEC Corp., according to Dataquest Inc. Fujitsu aims to sell 2.5 million PCs in Japan this year and 5 million units by 1998. That would hike its share to about 30% by some predictions..

Competitors who doubt the company has the stamina for such price wars may get a rude surprise. Lowball bids for government contracts, for instance, helped Fujitsu surpass IBM Japan in domestic mainframe sales in 1980. It once shocked even Japanese rivals by submitting a bid of just 1 yen to win a local computer software contract. Now, Fujitsu appears determined to boost its share in PCs, whatever the cost. The reason, says Furukawa: PCs are the key to "network services, the Internet, and software sales."

Fujitsu can afford to do that because of the way it cross-subsidizes product lines. It may continue to rake in cash, for example, on its next-generation peripherals. Take wall-hanging televisions. Fujitsu is not alone in pursuing this dream: Matsushita Electric Industrial, Sony, and Sharp also want to replace the world's supply of cathode-ray tubes.

But Fujitsu's approach is considered among the best. The company has invested about $200 million to build a plant in southern Japan that will begin monthly shipments of 10,000 42-inch units priced at $5,000 this October. Fujitsu foresees a $2 billion business by the end of the decade, which would still represent replacement of only a small portion of the 100 million TV tubes sold each year. Plasma displays "could match Fujitsu's computer business if they do it right," says Chuck Goto, managing director of Smith Barney Inc. in Tokyo.

Ultimately, of course, not all of Fujitsu's gambles will pay off. Outside Japan, Fujitsu executives have yet to articulate how they plan to turn around their floundering mainframe operations or take on systems-integration giants such as IBM and Electronic Data Systems Corp. But even if just part of Fujitsu's strategy clicks into place, the continued flow of cash, expertise, and ideas will allow it to cruise deeper into network technology. No other Japanese computer maker has as clear a view of the target. That's why a rewired Fujitsu could pack a few more surprises in cyberspace.BY STEVEN V. BRULL IN TOKYO, WITH ROBERT D. HOF IN SAN FRANCISCO, JULIA FLYNN IN LONDON, AND NEIL GROSS IN NEW YORKReturn to top


Later, Baby
LIMITED-TIME OFFER SUBSCRIBE NOW

(enter your email)
(enter up to 5 email addresses, separated by commas)

Max 250 characters

 
blog comments powered by Disqus