Daring to Be Different -- Tokyo's GOL: An Idiosyncratic Success Story

Back to Contents of Issue: November 1999


Opportunities for foreigners and Japanese alike have never been greater than in today's Japan. But it wasn't always like this, and certain netpreneurs were here at the beginning. Roger Boisvert-founder of Global OnLine-is one of them.

by Tanya Clark

The GOL team Opportunities for foreigners and Japanese alike have never been greater than in today's Japan. But it wasn't always like this, and there are certain netpreneurs who were here at the beginning. Roger Boisvert -- founder of Global OnLine -- is one of them.

There aren't many individuals - in any country -- who can claim to have "been there" at the beginning. In Tokyo, in 1993, Boisvert was. His story, and the parallel story of his two Internet startups, ISPs Global OnLine Japan K.K. and Intercon International K.K., captures the magnitude of Japan's intensifying assault on the Internet world. Japan, of course, is a laggard. The U.S., Europe, Australia, and even some Asian nations grabbed on to the Net's potential long before Japan's first bureaucratic government clerk mastered basic HTML.

This reluctance has walked hand in hand with Japan's continued recession. The nation's unwillingness to face drastic change in the way it does business is often hailed as the main cause of Japan's failure to kick-start its economy. Nevertheless, behind the keiretsu, mega-corporations, and the dithering, scandal-ridden government and bureaucracy, individuals in Japan have struck out and built new businesses and new ideas on the back of the technological revolution. Masayoshi Son of Softbank Corp. and Kazuhiko Nishi of ASCII Corp. are two to have made their mark.

In Roger Boisvert's field, Japan now boasts more than 3,000 Internet service providers, up 50% over last year. The nation is getting wired. As of August 1 this year, Japan had nearly 82,150 connected domains, including over 64,000 commercial (.co.jp) sites, hundreds of ISPs, and (depending on whose statistics you use) somewhere between 8 and 18 million users. Many expect NTT's decision to introduce fixed access pricing is going to light the needed fire under Japan's Net. The market, analysts say, is set to explode.

Japan's commercial Internet is born
But in 1993, none of this existed. Headhunted from McKinsey & Co.'s Tokyo office by an American software guru to startup the commercial Internet in Japan, Roger Boisvert launched Intercon International K.K. (IIKK), and, says Boisvert, he switched on Japan's commercial Internet at 6:10 p.m. on Sept-ember 24 of that year. But Boisvert's path toward that moment-like the man himself-was complex and unconventional.

Kurt Baumann, the president of Intercon Co. in the U.S., approached Boisvert to start the commercial Internet in Japan. "The market should logically exist, but no one had found a way," related Boisvert. Baumann wanted him to "break through the government barriers . . . It seemed so hard to do I turned the offer down three times. Here I was coming out of straight-laced McKinsey in my white shirts and all, and here is a guy with hair down to here, six earrings, a beard down to there, and other rings that weren't so apparent," exclaims Boisvert, describing the hirsute Baumann. "It was an important decision, I spent a lot of time thinking about it." He asked his wife and future business partner, Yuriko Hiraguri, to meet Baumann. "My wife tends to be a pretty good judge of character," adds Boisvert. "I thought she would hate him because she is also kind of straight- laced-extremely so-but there was extreme intelligence in his eyes that you could feel." Boisvert bit the bullet the fourth time Baumann asked.

The problems began immediately. To start a company, a kabushiki kaisha, in Japan, you need 10 million ($92,600). Boisvert had half that. To get around the rules he borrowed the remainder, returning it after ten days. So now he had only 5 million to find office space, equipment, and personnel. In short, almost everything had to be scavenged. Computer parts were picked up anywhere they could be found, friends in the U.S. would cart bits and pieces over. Boisvert's first monitor had no screws. Office space was another story.

In this, Joichi Itoh, the owner of Digital Garage, another successful Japan Net venture, "comes out as a star," says Boisvert. "Jo was trying to help me out, he felt this was very important." Itoh said he knew someone with empty office space. Continues Boisvert: "Only problem is, this guy absolutely hates foreigners. He's like a right-wing nationalist type. So, we were walking over to his office, we made an appointment to meet the guy, and Joey is saying, let me do all the talking. You just explain this is good for Japan, Japan needs this.' I was pretty nervous walking up the stairs. We walked in and the guy opens the door, takes one look at me, looks at Joey and says, 'what the hell are you bringing this white trash for?' And Jo says listen to what he has to say.' I said we are trying to make this thing called the commercial Internet in Japan. It's really good for the country, the country needs it, has got to have it, if you don't have it, you know, Japan will backslide into the dark ages compared with the rest of the world." Within 15 minutes, IIKK had its first office, an "old karaoke bar on the sixth floor of a building in Myogadani," recalls Robert Seastrom, IIKK's engineer. "We literally didn't buy any equipment for the first office, for the first commercial Internet," says Boisvert. "It was all leftover parts."


Tanya Clark, a Tokyo-based freelance writer, has been-like the managing editor of J@pan Inc.-a satisfied GOL user for several years. Contact her at tclark@gol.com. (Editor's Note: GOL provides free Web-hosting services to several Tokyo-area non-profit organizations, including Japan Webgrrls and International Women in Communications. Tanya Clark recently became president of IWIC.)

Meanwhile, Boisvert was meeting with the government ministries, whose approval he needed to start the Internet connection. The biggest task, says Boisvert, was educating the bureaucrats. "The government wasn't ready to give approval yet. They were interested. They were curious, but they didn't know what it was. They were asking intelligent questions: why should we do it? what is the impact on the country? What if we don't do it? Gee, the government was asking the questions they should be asking," says Boisvert. "They also wanted to do it the least risky way possible, and for someone who did not understand it, it was difficult to find a way to do so without risk." Boisvert was new to the Net himself. At the same time, "I was actually paying someone to educate me rapidly, someone in the U.S. was sending me notes every day," he says.

Authorization came by being thorough and persistent. The first task was to identify which government agency had the ultimate authority. In 1993, competition to control Japan's telecommunication policy had already begun among key ministries, including the Ministry of International Trade and Industry and the Construction Ministry. Nevertheless, Boisvert determined that the Ministry of Posts and Telecommunications (MPT) maintained licensing rights granting authority to companies to conduct telecommunications businesses. It was there that he targeted his forces-himself and a young Japanese speaker. "I hired a good-looking Japanese female, very polite, and brought her to the Ministry every day for a week, 95 basically." But, where companies had failed by going to the top brass of the Ministry, Boisvert tried a different tack. "I went to the bottom," he says, to a 25-year-old lower-rung bureaucrat. "The clerk has the time," explains Boisvert. "The clerk doesn't give you the solutions, but the clerk will answer your questions. So we were talking to the clerk every day. Can we do this? No. Can we do this? No. How about this? No. How about this? No, you can't do that either. How about this? No. No, you can't do this. So we went through four days of 'No,' to every question we asked. But although he answered 'No,' he didn't discourage us from asking more questions, he allowed it. It was his job to respond to the questions. The answers he didn't know, he found out."

The key, says Boisvert, was that the government didn't want head-to-head competition. They wanted co-operative competition. "I believe the phrase is co-opetition now," he adds wryly. IIKK got authority to go ahead on a trial basis because it was small and, says Boisvert, a foreign company that knows when to keep its mouth shut. Authorizing a trial run for a large Japanese company could have been riskier, concludes Boisvert. If the trial had failed and the government had tried to yank authorization, "the Japanese company is going to make life very uncomfortable for the Japanese bureaucrats," adds Boisvert.

Internet wars
"Once we turned up the system and announced we were the first fully functional ISP in Japan, there was a massive negative reaction from wannabes-people that wanted to be in the business but couldn't make it," says Boisvert. My first customer was TWICS, Japan's first dial-up access provider, he says. "I'm not going to get into details," Boisvert says, "let's just say that TWICS and IIKK suddenly got disconnected. For four days we couldn't get any response. My clients had no reason to take this. I had to respond." Seventy-five faxes went out to the world's media and within 40 minutes of the first one, claims Boisvert, we were reconnected. The issue abated after a few weeks, and Boisvert adds, "I rarely tell the story nowadays. But the uproar seemed to stoke the market. We were getting offers for everything, up to $50,000 for 5% equity in IIKK," he says. Within four months of its start, IIKK was sold to PSINet Japan, today one of the key players in Japan's ISP market.

The awkward age
Japan's commercial Internet market was born, but now Boisvert was out of a job-and he had the Internet bug. From January 1994, Boisvert spent six months trying to find some new way to crack Japan's Net market. "There was only one IIKK," he says, "so I had to find other ways to get access to the national capacity." In the end, he took a similar route as before, planning on creating a BBS he named Global OnLine-and this time the focus was to be on quality, the highest quality. Again, Boisvert scrounged office space, staff, and money. Once again, he needed the elusive 10 million to start a Japanese company. With a partner sharing half the load, it seemed the new company was ready to launch. But the partner was getting married and decided to bail. That's when Boisvert's wife, Yuriko Hiraguri, became involved in the business. "With five million yen of our money invested it was my problem," Boisvert says, "at ten million yen, it became OUR problem." In June 1994, Global OnLine Japan K.K. was born. And with GOL, Boisvert had found his life's mission.

Global OnLine consumes Roger, says Laurence Hewitt, manager of GOL's Web design division, NVision. "It consumes him, all of his energy, all of his thoughts and dreams." But it was still only 1994, and very few people in Japan even knew of e-mail or the Internet. Boisvert began recruiting customers among friends, offering free dial-up e-mail accounts. "We were providing services to about 300 people before we even opened an office." Boisvert and Hiraguri installed eight phone lines in the study of their typically small Japanese apartment, the phone wires coming through the window. Some wanted to start paying, says Boisvert, "but we still couldn't charge for it. It wasn't good enough. I was bleeding cash badly." Commitment to quality is, Boisvert hopes, Global OnLine Japan K.K.'s hallmark.

From the end of January 1995, Boisvert was confident enough to start charging. By then, 1,000 customers had signed up. But financing was still a challenge. This was an independent startup, without any kind of corporate backing. That made financing nearly impossible, as startup venture capital in Japan was virtually unknown. "We pursued one firm," says Boisvert's wife, who was by now full-time at the company, having quit her job teaching at an English-language school for children, "but the VC firm wouldn't invest. Instead, our contact there put his own money into GOL on the condition that other individuals followed suit." They did, and one of the other investors later became GOL's chairman, Yuji Suzuki. The magic of the Internet lured others. "I don't know exactly when, but a lot of people started helping," says Hiraguri. Friends would bring in friends, bring their own computers, printers, even furniture. "That was a lot of fun. Lots of people were helping us for nothing. Without their help, I don't think GOL could exist today," Hiraguri says.

A day in the life
Day-to-day operations are managed by COO Christopher Phelan, while Boisvert, as president, is left to pursue new opportunities. Hiraguri makes it her main job to ensure the staff is happy, while keeping an administrative eye over everyone. (The most difficult staff task, says Hiraguri, is ensuring that the telephone support staff aren't sent over the edge, calming the jangled nerves of customers with oddly behaving computers and phone connections.)

Working for an Internet company in Japan creates opportunities for people who can deal with change, says Christopher Phelan. Most of GOL's staff and customers are Japanese nationals, although it is still hard to know whether to describe GOL as a Japanese company or a foreign company in Japan. Boisvert prefers to describe GOL as a Japanese company run by a Canadian. In fact, Boisvert is a tad touchy on the subject, from the days when the company was called Gaijin Online (gaijin is Japanese slang for foreigner). Many of the first customers were the "gaijin" friends of Roger and the staff. But this emphasis did help GOL forge a strong niche in an increasingly competitive market.

To house the now 70, and growing, staff serving the 20,000-plus customer base, GOL recently leased four floors in a West Shinjuku high-rise. Stone blue carpets and white walls dominate. Work desks are clustered together, somewhat swamped by open space. When Boisvert moved to these offices, he was determined to ensure there was more than enough space to expand. Notably, the sixth floor is awash in more empty space than equipment. "There is plenty of room for telcos to co-locate," explains Boisvert, touting a business possibility he has been pursuing for some time-if he can find partners with the right fit.

The yen starts to flow
From day one, Boisvert has recycled all of GOL's earnings back into the company. For much of the early period, he drew the lowest salary in the firm, he says. But reinvesting meant the balance sheet rarely looked stellar, a deterrent to new corporate investors. So GOL further forged its private, independent brand, attracting a wide selection of individual shareholders among friends and stockbrokers. The sun finally broke through this past July, when GOL signed with the Connecticut-based venture capital firm, J.H. Whitney & Company (variously described as worth $3 to $10 million). Boisvert hopes the shareholding taken by J.H. Whitney will finally provide GOL the financial boost it needs to grow explosively.

This is certainly a start, but the type of growth GOL may need to pursue will require much higher amounts, according to many analysts. The question is, will this be enough? "Either you are a niche player or right at the top," says Mahendra Negi, first vice president at Merrill Lynch's Tokyo office and their software and Internet industries analyst. "Anyone in between will find it difficult." GOL, argues Negi, could survive in the strong niche it found, providing bilingual and high-quality service to the foreign community and some Japanese users in Tokyo. But GOL has moved into the corporate market, which places it in direct competition with entrenched players. "It is not clear how they will do," Negi concludes. GOL "can use $100 million in cash [and] a quick listing on the market." GOL needs brand building, to buy market share, he says. In fact, the company is talking about an IPO on the American market (NASDAQ), possibly in 2000 or 2001. Negi argues any listing would have to be before the end of 2000 to take advantage of the opportunities about to be unleashed by fixed pricing.

Fixed pricing for Japan Net access
"Fixed pricing is coming," says Tom Satoh, an independent consultant who helped set up AOL Japan and is now a partner at Heidrick & Struggles Tokyo. Satoh is convinced, as are many other analysts, that Japan's Internet market is set to explode now that NTT is removing the final barrier to growth. High local call costs (10-about 9.3 cents U.S.-for 3 minutes online) have meant Internet surfers log off as soon as they can. Fixed pricing, even at the flat rate, 10,000 ($93) per month unlimited connection time proposed by NTT, will leave the market wide open. There is a "huge Internet potential energy" in Japan, says Toby Rodes, telecom analyst at Dresdner Kleinwort Benson Asia. "Moving to this flat rate price will allow it to go kinetic."

But where does this leave local ISPs-especially smaller ones like Global OnLine that are shifting focus away from their niches and onto the larger market? Many analysts and Internet market players in Japan have a soft spot for Global OnLine. Outright criticism of the company is rare. Yet, there are some obvious strategy questions. For instance, what chances does a company with a subscriber base of around 20,000-including up to 3,000 corporate customers-stand in a market set to be swamped by well-backed corporate behemoths, like AOL Japan and Sony Corp.'s So-Net? If the market follows the same patterns of development as in the U.S. and Europe (likely, but perhaps at higher speed), is GOL agile enough, flexible enough, and adequately financed to survive? Will Japan Internet connection costs fall to almost zero, as some analysts predict, leaving many ISPs up the creek with no paddle? How will ISPs need to evolve to survive in this new market? Is GOL ready? Roger Boisvert would vehemently challenge any critical judgments of his life's work, but the company does face severe challenges.

"GOL's people must be terribly smart to play this game straight out in the next few years in Japan," says Heidrick & Struggles' Satoh. Satoh believes only the big will survive the shakeout, saying GOL's task is to increase its subscription base to 200,000 in the short term-and one million soon after that. "They don't have much time," Satoh says, "the market is really catching up." Satoh argues that the difference this time around, in Japan, will be the pace of technological development. In the U.S., and to some extent Europe, a lot of the necessary technology-for Web surfing, e-commerce, networking, and the more sophisticated websites-was still in development during the Internet's early years. That, he argues, helped hold back the pace of development. "Technology-wise, there was still a bit of room." Today, the only thing holding Japan back is fixed pricing, and once that barrier is released "everything will go boom," Satoh says, arguing that what took six to seven years to develop in the U.S. may only take three years in Japan. To survive, Internet companies will need to be very agile, have a very aggressive growth plan, and the ability to sustain it. And that will take money. GOL argues that they have their business plan in place. Boisvert calls it the five legs: network integration, website development, e-commerce, Internet access, and the telco plan. Again, the question, is this enough?

Surviving the Net jungle
"To survive, entities will have to have an appealing brand name and maybe create a portal to enclose their customers and weave to important websites," argues Hiro Nobu-sawake, telecom and media analyst at ING Baring Japan. Tom Satoh agrees. "Small ISPs will be hard pressed to create their own resources to attract and retain customers," he says. The ISP business is very easy to get into, Satoh says, pointing out all you need is the networking equipment and some off-the-shelf software. "But when the market expands to millions of users-[who] want everything at a low price-the more subscribers you have, the greater your buying power. You can almost give away network access." Especially if content' becomes the key to success-as many analysts argue it will. "Some people manage virtual networks better than others, some service people better, some can provide seamless access and reduce down time, but you just can't count on that as your advantage," argues Satoh. "Network gears are getting more reliable. In the end it doesn't really matter who constructs the network. What will differentiate [the winners and losers] is what the end users really see." "Most providers have no future," says Haruhiko Ichikawa, general manager of Nihon Net. Ichikawa argues the competition in Japan is too big, too intense. Everyone, he says, from the electricity companies, the cable companies, and the private railways to large corporations like Sony Corp., is building connection networks. Even today, most of Japan's 3,000-plus ISPs are losing money. Ichikawa also concludes GOL's biggest drawback is lack of content.

But GOL has heard all this before-at every step of its relatively short history, someone has been predicting the demise of the small ISP-and GOL is still here. Boisvert's goal is to stick to the game plan, increase the focus on corporate customers, and rely on his team. And the company's "gaijin" connections may provide another, hidden, advantage. Providing content in exchange for free access requires companies-Japanese companies-to understand the concept of providing a loss-leading product. "I refuse to believe that there is a Japanese company, aside from the media conglomerates, that understands content and advertising adequately enough to even come close to that kind of business model," says Dresdner's Rodes. GOL's foreign management is not bound by the same type of rigid thinking that could be the cause of Rodes' frustration with Japanese commercial creativity. It may also provide GOL, and other firms, with the breathing space for flexible development that the U.S. and Europe found through technological development delays.

Boisvert has built up a surprisingly loyal following for a company so small. GOL provides a lot of services-including e-mail accounts and websites for no charge to community groups throughout Japan. This kind of goodwill is another big advantage for Boisvert as the firm moves from the scrag-end beginnings of Japan's commercial Internet to creating his vision of success. When asked about the future, Boisvert laughs and adds, "Just watch us-you'll enjoy it when it happens." Eventually, Boisvert says, he can see the name "Global OnLine" on billboards throughout Asia.

Note: The function "email this page" is currently not supported for this page.