Guanxi (The Art of Relationships) Microsoft, China, and the Plan to Win the Road Ahead

1

Beast from the East

November 8-11, 2004

This is a new kind of manufacturing in China. Not just shoes, socks, baby strollers. Now we manufacture MIT students, papers, and software.

--HARRY SHUM, MANAGING DIRECTOR OF MICROSOFT RESEARCH ASIA

Half a world away from the calm beauty of Puget Sound, there's a lab where Bill Gates's software dreams come true. At Microsoft Research Asia, the drive to succeed is as intense as the traffic that roars by its front door in unbridled fury. If the software megagiant's other facilities around the globe seem idyllic, this one, in Beijing, is pure street. Microsoft's mantra here: work hard to get in the door; work harder to survive; then work even harder because the real work -- that of creating the global future of computing -- is just beginning.

If you find it hard to root for Microsoft, you've probably never met Harry Shum. The Beijing lab's managing director is hearty, engaging, and quick to make jokes. In his late thirties, he's also surprisingly young. "This is a new kind of manufacturing in China," he smiles, waiting outside his office. "Not just shoes, socks, baby strollers. Now we manufacture MIT students, papers, and software." His longtime colleague, HongJiang Zhang, walks by and concurs. Cultivating talent, he says, "is another level of 'Made in China.'" Zhang, who's a little older than Shum and initially comes across as more reserved, heads the Advanced Technology Center. An offshoot of the research lab housed in the same building, this first-of-a-kind division was created to accelerate the movement of the lab's technologies into Microsoft's product pipeline -- for China and the entire world.

Together, Shum and Zhang lead a nearly 500-strong organization that looks like a typical corporate lab butfeelslike a hungry start-up. Come in at almost any hour and you'll find scores of students -- in addition to their staffs, the two groups support some 300 interns at any time, most from Chinese universities -- tooling away on projects jointly supervised by their professors and Microsoft researchers. It's a place where 10,000 resumes arrive in a month and interns spend some nights on cots next to their cubicles. Add the buzz of Mandarin conversations, the window views of Beijing's sprawl, and the hint of cigarette smoke, and you are constantly reminded: this isn't corporate U.S.A. anymore.

Every week is busy here -- but one particular week in early November 2004 was special. The events packed into that whirlwind week spoke to every level of the company's strategy in the Middle Kingdom -- and to the all-out, breakneck pace of global innovation today. To celebrate the sixth anniversary of the Beijing lab's inception, Shum and Zhang entertained a host of distinguished visitors from around the world. The dignitaries included their superiors from Microsoft Research headquarters in Redmond, Washington: Dan Ling, vice president of research, and his supervisor, senior vice president Rick Rashid, one of a handful of Microsoft executives who report directly to Gates. Also on hand were notables from the lab's technical advisory board, which encompassed some of the biggest names in computer science, among them Chuck Thacker, winner of the 2004 Draper Prize, a $500,000 award considered by many to be engineering's top honor; Jitendra Malik, chair of the department of electrical engineering and computer science at the University of California, Berkeley; and Victor Zue, co-director of the Computer Science and Artificial Intelligence Laboratory at the Massachusetts Institute of Technology. In addition to a series of advisory board meetings, the frenetic week would include a Faculty Summit of 207 professors from throughout Asia, many of whom collaborated with the lab, and the Computing in the 21st Century conference that was co-sponsored by Microsoft and the National Natural Science Foundation of China.

On an overcast Wednesday in the midst of this hectic week, two large meeting rooms down the hall from Shum's office overflowed with animated conversations and dozens of research demos. Jetlagged vice presidents, technical advisors, professors, and other studious-looking visitors milled around perusing the demos while eating personal pan pizzas catered by a nearby Pizza Hut. Microsoft demos are legendary: mastering the art takes technical know-how, showmanship, and a clear understanding of why a project is important to the company. A good demo can erase months of frustration and get you noticed. That's why the Beijing researchers had been living and breathing this stuff for months. Eager to impress, some were nervous and struggled with their English, while others pulled it off without a hitch.

To Microsoft, what was in these rooms portended the future of computing -- and which competitors the software superpower was lining up in its sights. Target number one, which would loom ever larger in the crosshairs over the coming year: Google. An entire wall of demos highlighted smarter Internet search tools designed to take users right to where they need to go for answers instead of giving them a bewildering list of links, and to provide more highly targeted Web advertisements based on the exact nature of a person's search query. These efforts aimed to derail the Mountain View, California-based search company, whose faddish popularity once led Gates to quip, "There's companies that are just so cool that you just can't even deal with it."

Almost as urgent a competitor: Sony. The Beijing lab played a growing role in Microsoft's push for new kinds of graphics and interfaces to help it win the digital entertainment space over the consumer electronics giant. One of the more curious-looking demos employed a camera to track human faces, a key part of the next-generation interactive video games Microsoft envisioned. "PlayStation 2 already has something like this for motion, but not faces. We need something in Xbox," said researcher Dongmei Zhang, referring to Microsoft's video-game platform, which was slated for a major new release called Xbox 360 in late 2005.

The third company in Microsoft's sights: cell-phone maker Nokia. Software for mobile devices was still a relatively small business for Microsoft. At the time of the meeting, the company's Windows Mobile operating system had just surpassed Palm's in conventional PDA market share, but still trailed Nokia's system significantly when it came to cell phones and mobile e-mail devices. To this end, another set of Beijing demos showcased software that enables wireless videoconferencing and "seamless roaming" -- so that any cell phone or handheld organizer can provide voice, video, or data communication anywhere in the world, anytime, on any network.

Increasingly, the Beijing lab was where the action was in all these battles -- and the technical advisory board seemed impressed by the show. "They're doing really first-class research," said MIT's Zue, an advisor to the lab since its inception. After an hour of seeing demos and asking questions -- some superficial, but many complex and detailed -- the board retreated to a large room on the sixth floor for a private, closed-door meeting. Inside, the air was thick with anticipation. Researchers guzzled tea and coffee as they exchanged pleasantries. As technical assistants scrambled to get audiovisual systems up and running, Microsoft's top research brass, Rashid and Ling, took seats front-row center, flanked by other high-ranking staff and the advisory board. The room could seat seventy, but only half the chairs were filled. About 25 researchers sat in, some from Microsoft's main lab in Redmond, most from the Beijing lab. It was like a wedding, where the guests of the bride and groom occupy pews on different sides of the church: for the most part, Redmond and the States sat on the left side of the room, Beijing and China on the right.

The board had gathered, together with the lab's leaders, to discuss competitors and give more detailed feedback and criticism on key lab projects than was possible in the demo rooms. Harry Shum, a people person, kicked off the meeting by introducing all the visitors by name. He then called attention to a couple of new faces brought in from the States as assistant managing directors of the lab: speech expert Hsiao-Wuen Hon, who had done both research and product development for Microsoft in Redmond, and graphics guru Kurt Akeley, a co-founder of Silicon Graphics, who had recently finished his long-delayed Ph.D. at Stanford (he had taken a leave of absence in 1982). Both were heavyweight hires whose presence bolstered the lab's standing in China and also spoke to its status as a haven for top talent, not just from China, but from anywhere.

Shum then served up a few stats to demonstrate the lab's growing prowess on the global stage. In 2004 it was responsible for 7 papers out of 58 accepted for SIGIR, the world's largest and most prestigious conference on information retrieval (a key component of search), and 5 papers out of 80 presented at SIGGRAPH, the top graphics conference, where the next generation of gaming and entertainment wizardry is often unveiled. No other lab or department came close to these numbers, even those many times larger than the Beijing center. And it wasn't esoteric research, Shum reminded his visitors. To date, close to 100 technologies had been transferred to Microsoft products -- tops among the company's research arms outside of Redmond -- and the figure was growing fast.

After Shum's presentation came technical talks by key researchers on user interfaces, wireless networks, multimedia, graphics, and search -- the lab's five main areas of focus. The talks were all informative and spoke to Microsoft's competitive battles, but one stood out in particular -- and it came back to Google, Microsoft's archnemesis. Wei-Ying Ma, the smooth-talking, cherub-faced manager of the Web search and mining group, spoke with a light Mandarin accent, but his intensity and desire to help Microsoft compete came through. He explained how his team had solved a key problem in the search business. Today's search engines make much of their money by selling Web ads that pop up next to search results, but they typically rely on human workers to evaluate how relevant the ads are to any particular search query (a domain called "relevance verification"). This manual labor takes a lot of time and effort. Ma's team had found a way to equal human results automatically -- potentially saving many millions of dollars.

This automation, Ma said pointedly to the closed-door crowd, provided a key advantage over Google. A longer-term approach that could help Microsoft win the search war outright, he explained in expanding on some of the demos the group had seen earlier, centered on "mining" the expanse of data on the Internet for deeper patterns in the links between Web sites. Understanding those patterns could eventually yield much more accurate search results -- for instance, by allowing results to be grouped in easy-to-recognize categories so that users don't need to scroll far down the page to find the returns they are looking for. His team members, recruited for their expertise in the fundamentals of pattern recognition and information processing that underlie search algorithms, were in the process of honing the technology on thousands of users of Microsoft's MSN Web portal, which they had usurped as the company's personal testbed in the search war -- not just with Google, but also Yahoo and other competitors.

The presentations, which took up much of the day, painted a picture of an all-out assault on the state of the art in computer science. By the time it was HongJiang Zhang's turn to speak, the stage was set to drive home what this all meant for Microsoft's businesses. As the director of the Advanced Technology Center, the fastest-growing part of the Beijing outpost, Zhang was charged with accelerating lab research into products and spurring the company's innovation worldwide. In his talk, Zhang was all business himself. He detailed projects in the works for virtually all of Microsoft's business divisions. In fact, research was moving out of the lab so quickly that Zhang announced that the 100-strong technology center, itself only a year old, would double by the middle of 2005, surpassing its research-lab sister in size.

After his overview, Zhang turned the floor over to two of his top lieutenants, Baogang Yao, a lead developer of new Web-advertising technology for MSN, and Wei-Ying Ma, who again brought it home to Google and the search war. Ma spelled out the numbers behind one of the company's major growth markets, one where Google had totally cleaned its clock -- the paid ads tied to search. As of the third quarter of 2004, he noted, online advertising as a whole constituted an $800 million market -- double what it had been just a year earlier. "This is a new business, a new market to Microsoft," he summed up. A major goal of his group's work, Ma stressed, was "to make MSN stronger, to win in this online advertising space."

After the presentations there were some detailed questions from advisory board members. Rick Rashid, Dan Ling, and the rest of the visitors sounded enthusiastic and upbeat. They asked a few more general questions about resources and technical approaches. Then they sat back in their chairs and seemed to soak up the power of the Beijing atmosphere. Rashid was especially pleased by the first-year progress of the Advanced Technology Center. "There are just things [here] they can't even do in Redmond, they're just not capable of staffing," he enthused. "The whole point is really to do things that wouldn't have otherwise gotten done."

The seven-story building where more and more of Microsoft's global battles are fought sits a few blocks off Zhongguancun Road in Haidian, Beijing's high-tech district. The Sigma building, as it is called, has a large glass-front lobby with plenty of shiny metallic trim. A Microsoft Windows sign runs along the side of the roof and is easily visible from the street. In the mid-1990s, the Zhongguancun area was dubbed the "Silicon Valley of China." Today it is home to more than a thousand technology companies and is situated within a few miles of a dozen top universities and academic institutes, including two of the most prestigious schools in the country, Peking University and Tsinghua University.

To get to Zhongguancun, hop in a taxicab or hire a car and driver and head to the northwest part of town. Beijing doesn't blow you away at first -- the surrounding brown farmlands make the outskirts look more like rural Indiana than the capital of the world's next economic superpower -- but the city is relentless in getting your attention. From ubiquitous billboards to endless construction projects, the city exudes a gritty, raw energy as it gears up to host the 2008 Summer Olympics. In the residential areas, block after block of drab beige apartment buildings are crammed together. Crowds of elementary school kids hone their basketball skills on outdoor courts. Closer to the Microsoft lab, high-rises compete with smokestacks for skyline supremacy. Run-down buildings squeeze in next to bustling consumer electronics markets and the Beijing Satellite Manufacturing Factory, where China conducts its spaceflight research.

Everyone knows China is a waking giant. Its booming population of 1.3 billion represents the largest potential market in the world. It is already the undisputed world leader in everything from low-cost manufacturing to mobile-phone usage (400 million subscribers by the end of 2005). With more than 100 million people on the Web as of 2005 and 19 million PCs projected to be sold that year, it has roared past Japan to become the second-leading user of the Internet and second-leading purchaser of new personal computers in the world (after the United States in each case). And the penetration of wireless and mobile applications is well ahead of the rest of the globe. Computer gaming, text messaging, and social networking on mobile devices are exploding as a way of life. On Chinese New Year, typically more than a billion instant messages are exchanged nationwide.

To succeed here -- to tap into both this immense market and the incredible talent from the world's largest university system -- multinational companies like Microsoft have learned, often the hard way, that they must find a way for both themselves and their host to gain. China is a land of striking contradictions that must be navigated deftly and carefully. On the one hand, its leaders are outspoken about wanting the nation to become a technological powerhouse. On the other, its Internet censorship is among the most severe in the world and its intellectual property laws are weak and unenforced; piracy runs rampant. While China preaches education for its students, it restricts the very information available to help them learn. And although China is a communist nation, its level of entrepreneurship has never been higher: in spots you find a 24/7 drive for financial success that is unrivaled even in places like Silicon Valley.

For Microsoft, the Beijing lab has become pivotal to navigating this minefield of competing interests. For foreign companies, a big part of giving back to China -- what its leaders want to see -- involves training students and workers in cutting-edge research and management techniques. In China, government leaders control almost every aspect of their citizens' education and training, not to mention the mainstream media and consumer markets. Academia and government are intimately tied together, much more so than in the States. That's why establishing strong relationships -- the art ofguanxi-- with education officials and academe is so important. And that's why making research and teaching connections with Chinese universities is such a top priority.

This is what the busy week in November 2004 was all about. It began rather pointedly, on the Monday, with Microsoft hosting an international Faculty Summit less than a mile away from the lab. The gala took place at Beijing's aptly named Friendship Hotel. It's an older hotel with traditional Chinese décor -- lots of red paint and drapery, gold trim, pillars, wall murals, and sparkling chandeliers. In a spacious hall on the second floor, the 207 computer science and engineering faculty members the company had brought in from across the Asia-Pacific region gathered behind long rows of tables to hear the latest from Microsoft Research and some of its collaborators. Two-thirds of the professors hailed from Chinese universities; the rest came from Japan, Korea, Singapore, Malaysia, Thailand, and Australia.

At nine in the morning, Harry Shum took the stage to the beat of rock music blaring over the PA system. He made opening remarks under a small spotlight. "This meeting is about partnership with academia in Asia," he stated. "The important thing to the lab is quality of people." He proceeded to tout the lab's pride and joy -- the 1,500 student interns who had come through its doors over the past six years, most from top computer-science programs across China. "In fifteen years, I'm just saying, suppose Bill [Gates] runs out of money. We can still keep the lab running from our alumni donations," he joked. In closing, Shum mentioned that Microsoft Research Asia had recently been dubbed "the world's hottest computer lab" by an influential U.S. technology magazine: "I finally figured out why. It's because we don't have enough money to run air conditioning in the summer."

China's director general of higher education, Yaoxue Zhang, followed Shum at the podium. You couldn't ask for a better view of what China was looking for in its relationships with Western companies than Zhang. His remarks, in Mandarin, were translated in real time into halting English relayed through headphones connected to a wireless receiver. "We're faced with a very exciting, changing world due to the globalization of technology and the worldwide flow of talent," said Zhang. "Our students need the ability to compete internationally." Continuing on this theme, he cited the importance of foreign companies like Microsoft in bringing Chinese computer science up to speed. "Our science and technology level needs to be raised . . . We hope to attract more corporate investment in this area and, through international collaboration, improve our quality of education."

An elaborate buffet lunch of noodles, dumplings, meat dishes, and seafood followed the rest of the morning talks, which were focused and technical: posters and demo booths were set up around the food tables, so diners could continue to bask in the air of Microsoft and its collaborations. Above the din of conversation, the reactions from visiting faculty members were overwhelmingly positive. "Microsoft's academic outreach is focused on long term, which is really smart," said Helen Meng, an MIT-educated professor of systems engineering and engineering management at the Chinese University of Hong Kong. "Most other companies ask for turnkey solutions, but that's not what we're good at." Later that day, Roland Chin, vice president of research and development at Hong Kong University of Science and Technology, remarked, "Microsoft raises the standard in technology, and that will help Chinese students -- they want to work for multinational companies." All of this reflected a coming shift in the center of gravity of technology development, said Berkeley's Jitendra Malik: "The world in twenty years will be different. In terms of the U.S., instead of looking to Europe, it will mean a greater percentage of the action is in Asia. That's obvious."

It was a productive luncheon, full of ideas; lots of business cards were exchanged. But the man in charge of Microsoft's outreach didn't have time to eat. While most guests were still savoring the last of their dumplings, Shum was off to continue his academic campaign. He headed downtown in the backseat of a private car to the ultramodern Grand Hyatt Beijing hotel, whose lobby boasts impressively tall windows that look out over an elaborate fountain. That afternoon, Microsoft Research Asia would sign a historic agreement with China's Ministry of Education to create collaborative computer-science labs at four of the country's top academic institutions: Hong Kong University of Science and Technology, Zhejiang University, Harbin Institute of Technology, and Tsinghua University. Microsoft saw this as a way of showing that its commitment to Chinese education and training was serious, and that the company was investing in China for the long haul.

On Tuesday, the day after the Faculty Summit, Microsoft'sguanxi-building kicked into even higher gear. It was a cold and rainy Beijing morning, but the sprawling campus of Tsinghua University teemed with life. Packs of undergraduate students navigated the wide, tree-lined paths that run along rows of old buildings. Some chatted noisily with their friends; others walked quickly with heads down. Many students and professors hurried to class on rusty old bicycles, weaving to and fro to avoid the pedestrians and the large puddles forming in the streets.

Tsinghua University is widely regarded as the top engineering school in China. Established in 1911, it is the alma mater of the country's current top three leaders, one-quarter of the members of the Chinese Academy of Sciences, and one-fifth of the members of the Chinese Academy of Engineering. It's no coincidence, then, that Microsoft Research Asia picked the campus as its home base for academic collaboration and recruiting. The lab drew more students from Tsinghua and supported more collaborations here than at any other school in China -- and it was a centerpiece of the Ministry of Education agreement that Shum had signed the day before. So its importance made the school a natural choice to serve as the first leg of the annual conference on Computing in the 21st Century. Each year this conference and "tour" had opened in Beijing and, beginning in 2000, had then moved to a different second city -- providing a showcase of top experts and the latest advances in computer science that had wowed and inspired thousands of Chinese students.

On the east side of the Tsinghua campus, attendees were greeted by a strange sight that you would probably never see on an American campus. Seven-foot-tall placards with the faces of Rick Rashid, Dan Ling, Harry Shum, Chuck Thacker, and other notable speakers lined the steps leading up to the university's sports stadium. Rashid's poster smile beckoned as if to say, "Enter here, top students, and we will educate you." In the distance, framed by the placards, an enormous smokestack belched clouds of white steam into the sky. It was a vivid snapshot of today's China, as it steps forward into the information economy of the future with one foot mired in its industrial past.

Inside the gymnasium, the Microsoft tour received a celebrity's welcome from 3,400 Tsinghua students, faculty, and assorted dignitaries. A stately brass melody played over the speakers as Shum once again took the stage to introduce the day's events. Spotlights careened around the stadium's interior, beaming Microsoft's logo to the upper decks. Flashing streams of digital bits were projected onto the sides of midlevel balconies. Many students and guests wore headphones to hear translations in English or Chinese. The president of Tsinghua University, Binglin Gu, followed Shum to provide the official welcome. "This is a great opportunity to learn from Microsoft's cutting-edge computing techniques," he remarked. "It is very inspirational for us. I hope in the future we will have more world-class scientists born out of the students here today."

Rick Rashid, the next to take the podium, echoed the sentiment, saying to the audience, "The next ten years will be the time for you to create new applications" in information technology. He appeared to be doing well for having stepped into a 16-hour time difference two nights before. The students in the stands and on the gym floor seemed to hang on his every word, many eschewing the Chinese translation available on their headphones. On the left side of the stage, his slides were projected in English; on the right side, a Chinese version appeared. Rashid pushed for the "democratizing of information" -- to put raw materials out there so that people can eventually be connected to everyone and everything in the world using a wide array of intelligent devices.

It was an ambitious and long-term vision of global technology that might make some in China uncomfortable. But the growth of Chinese information technology and the country's hunger for knowledge and training in computer science was already insatiable. In a backstage VIP room, Ya-Qin Zhang, a soft-spoken, China-born Microsoft vice president who had been the managing director of the Beijing lab for nearly four years before Shum, relaxed with a cup of coffee. Zhang, a major figure in Microsoft's China story, would also speak on the tour. He said he felt proud to see that his company was continuing to attract and nurture computer-science talent in his native land. "This is truly, truly a phenomenon," he gushed.

Maybe the last two days had gone too smoothly. After the main presentations, Microsoft put on a press conference for about twenty Chinese journalists. Severe-looking guards in green uniforms manned the doors to the room, locking hands with each other to block outsiders from entering. Up at the interview table were most of the day's keynote speakers, including Shum, Rashid, Ling, Thacker, and Tsinghua professor Chen Ning Yang, a Nobel laureate in physics who had taught in the United States for decades and probably ranks as China's most famous scientist. Yang had spoken on the history of physics and computing -- and had stolen the show. Almost all the questions from the local journalists were directed to him: When will Chinese universities be considered world class? (Research-wise, within ten years, he said; but at the undergraduate level, his students at Tsinghua already exceeded their counterparts at Harvard University.) Should the government spend more on high-energy physics research? (No, was his surprising answer.) At one point, Microsoft's media representative, Sheila Shang, reminded the reporters that this was a Microsoft event and that questions should be directed to all the speakers. But that did little to change the nature of the queries, which kept going to Yang. The unintended message to Microsoft -- it's not always about you -- was a sign that the company still had a lot of work to do in China.

The press conference was a glitch, but Shang would still count scores of articles touting Microsoft and its vision of computing. The next day, Wednesday, brought the technical advisory board meeting -- a midweek interlude of private discussions, and a bit of a respite from the public hustle and bustle of the rest of the week. But as that day came to a close, a private bus waited outside the lab and the pace picked up again. The second leg of Microsoft's conference tour was about to begin. Next stop: Chengdu, a rising high-tech metropolis in the southwestern province of Sichuan, a two-and-a-half-hour flight from Beijing. There, on the following day, many of the "21st Century" speakers would repeat the same demos and speeches, as if they were on a concert tour, playing the same sets in different cities.

On the ride to the Beijing airport, Rashid, Ling, Shum, Ya-Qin Zhang, Hsiao-Wuen Hon, and a handful of others gathered their thoughts and reflected on the week's events to that point. It was dinnertime and already pitch-dark outside. As the bus hit the highway, the Microsoft team wolfed down a distinctly un-Chinese meal of hamburgers and french fries delivered to the lab from a nearby McDonald's restaurant.

The air buzzed with excitement as they talked -- much of the conversation centering on the wonder of China. Rashid noted that cell-phone reception to call his Seattle-area home was better in Beijing than it was in his own office in Redmond. None of this came as a surprise to Zhang, who had long appreciated the exploding mobile market in China and had launched the Beijing lab's research focus on wireless networking. Back in Redmond, Zhang ran the engineering side of Microsoft's mobile-device software division, which aimed to provide the dominant platform for handheld devices -- in other words, to do for cell phones and PDAs what Windows had done for PCs.

The mood was light, and the Microsoft team was starting to loosen up amid a stressful week. Rashid was in a particularly good mood, waxing eloquent as the talk turned to competition with Finland-based Nokia in the wireless arena. "When Ya-Qin crushes Nokia, he'll get the Nobel Prize," the senior vice president proclaimed. The others burst out laughing while Rashid continued in fine form with his tongue-in-cheek cultural analysis. "I'm serious. See, the Swedes hate the Finns. If Ya-Qin puts Nokia out of business, the Swedes will give him a Nobel Prize."

Rashidwasjoking. But as the executives headed west to the next stop on their tour, it was just another reminder that Microsoft was here in China for one reason: to win.

If Beijing is the hard-driving East Coast, then Chengdu is the Wild West. The frontier city of 9 million is rawer and hazier than Beijing. It's darker at night, the air is smokier, the airport more desolate. But Chengdu is also known for its pleasant teahouses, restaurants, gardens, and relatively relaxed pace of life. It has a rich 2,000-year history and was the capital of several ancient kingdoms. By most accounts, it was the birthplace of printed paper money and a pioneer in manufacturing silk brocades. These days, it is the closest big city to Tibet, which lies 250 miles to the west -- making it a popular gateway for travelers to the land of the Dalai Lama.

None of this, of course, was what brought the Microsoft team here for what would be a short, focused visit. Late at night, as they touched down at Shuangliu Airport and walked briskly through the terminal, flashbulbs popped all around the visitors; unknown photographers, presumably from local newspapers or government organizations, snapped pictures without asking any questions. The group was treated like an entourage of celebrities. A private bus awaited to whisk them to their hotel, the relatively posh Sheraton Lido near the center of Chengdu, about thirty minutes away.

The bus came equipped with its own tour guide, a cheerful young woman who sat at the front facing the passengers and doubled as a standup comic. She seemed to delight in practicing her English but was extremely nervous -- she mentioned this several times and apologized, citing what an honor it was to entertain this group. Sichuan, the birthplace of the country's reform leader Deng Xiaoping, was the fourth-largest province in China, she pointed out. It was home to 85 million people and well known for its spicy cuisine, Yangtze River cruises, and giant pandas -- its dense bamboo forests harbored 80 percent of the world's panda population. The jittery guide proceeded to tell a series of awkward jokes about pandas -- how it's impossible for them to appear in color photographs (only black and white), how they look like they never take their sunglasses off, and so forth. (Something was lost in translation, perhaps.)

The Microsoft group was a fairly tough crowd, but they let her continue, reveling in the corniness of jokes that were so bad they were good. People from Sichuan, she went on, are known for being easygoing, hardworking, and having a good sense of humor. At the mention of these traits, the Beijing crew immediately poked fun at Baining Guo, the manager of the lab's graphics group who was along to run the demos during various presentations. Guo was originally from a small town about an hour's drive from Chengdu. When the guide said Sichuanese people are hardworking, the others mocked him: "Hardworking, or heartbreaking?" they laughed. Guo took the good-natured ribbing in stride, as if he'd heard it all before.

In the center of town, the Chengdu version of Boston's "Big Dig" construction project was under way to build a huge underground traffic tunnel. Overlooking the scene, an eerie statue of Chairman Mao stood in a large square with one arm extended, dimly lit by nearby streetlights. According to the tour guide, the statue stood 12.26 meters high, to commemorate Mao's birthday (December 26), and 7.1 meters wide, to mark the anniversary of the founding of the Communist Party in China (July 1, 1921). "Very thoughtful design," muttered one China-born member of the Beijing crew.

Along the nearby Tai Sheng East Road, storefronts were packed tightly into each block, seemingly braced for the crowds that would appear at morning's light. During the day, these sprawling street markets full of computers, cell phones, and other electronic gadgets sold whatever you needed -- much of it secondhand or counterfeit, especially in the back alleys off the main avenues. That brought to mind Microsoft's long-awaited next version of its Windows PC operating-system software, Vista, which at the time was codenamed "Longhorn" and slated for release in mid-2006. "They probably already have Longhorn here," quipped Ya-Qin Zhang, as he gazed down one of the streets.

The next morning, Thursday, it was time to meet with the mayor of Chengdu, Honglin Ge, and then kick off the "21st Century" conference at the nearby University of Electronic Science and Technology. Founded in 1956, the school is considered one of the cradles of the Chinese electronics industry. The scene outside the main auditorium on campus was similar to that at Tsinghua University two days earlier, with placards of the distinguished speakers, rows of buses running down the street, and students lined up to get in. Here, where Microsoft Research had never been before, the conference was an even hotter draw, and there were rumors of a black market for tickets. Just outside the doors, a young woman, apparently an undergraduate, approached one of the authors and asked if he was Rick Rashid -- and could he let her into the proceedings, because she didn't have a ticket.

Inside the auditorium, classic rock music from the Eagles ("Heartache Tonight") blared through the speakers. Most students were in their seats thirty minutes before the start of the event. The smell of cigarette smoke in the hallways was much stronger than at Tsinghua University, but that wasn't the worst of it. In the back "VIP" room, the bathroom had only a foul-smelling squat toilet -- basically a hole in the tiled floor -- and no ventilation. Even distinguished speakers had to come through to use it. (This probably wasn't what Chuck Thacker had in mind when he said he was "astonished by the progress that's been made" in the eight years since he last visited China.)

The lights went down. The president of the university and the mayor of Chengdu started the show with opening remarks praising the conference and Microsoft's role in bringing world-class computer-science researchers to campus. Then it was time for the keynote lectures from Shum, Rashid, Zhang, and Thacker. (Nobel Laureate Yang didn't make the trip, so no one worried he would steal the show this time.) After the talks, the keynote speakers took part in a panel discussion. The best question came from a student who approached the microphone and spoke forceful if broken English: "Good afternoon, super scientist stars! What is Microsoft doing on privacy and security, with increase in personal information on network?" It seemed that Chinese students everywhere were well aware of the hurdles to Microsoft's vision of democratizing information flow in China.

Perhaps the most telling scene, though, came after the event finished. On their way out, students lined up in droves to pick up a free gift being handed out by conference volunteers: blue and yellow bags bearing the Microsoft logo that contained fleece sweaters embroidered with "Microsoft Research Asia." As they carried the shiny gifts back to buses waiting to take them to their dorms and living quarters, the party favors seemed well worth their cost to the company of three U.S. dollars each.

That night, in a grand finale to the nonstop week, the Microsoft team indulged in a lavish 21-course banquet in a downtown Chengdu restaurant. With all the events and meetings finished, they could finally relax and let their hair down. They filed into a large private room decorated in the style of ancient Sichuan and sat down around three circular tables with pristine white tablecloths. At the main table, Shum, Zhang, and Hon sat to either side of the ranking executives Rashid and Ling. Chuck Thacker and his wife joined them, as did Ling's mother, who had flown in from Shanghai. The exquisite meal featured such regional delicacies as mushroom soup, chicken feet, Sichuan-style spicy beef, noodle soup, cured tofu, and custard. As they ate, the group rehashed the week's events and what they meant for the company -- and took time out to poke fun at one another whenever possible.

Dinner was accompanied by professional performances of traditional Chinese music, comedy skits, and Sichuanese opera. The latter came complete with mysteriousbian lian, or "face changing" performers wearing elaborate, colorful masks and costumes. Face changing originated as an entertainment form in the 1700s during the Qing Dynasty, supposedly in a performance about a prisoner who startled his captors by changing his appearance -- allowing him to escape. As a performer dances and flutters his arms around in the air (women are not face changers, traditionally), a mask is removed faster than the eye can follow. The technique, kept secret for generations like a prized magic trick, allows one actor to play up to four or five different roles at a time.

After dinner, the face changers danced and mingled around the Microsoft tables, shaking hands with the patrons and magically changing their masks in front of the startled and amused diners. That led Rashid to point out that face changing would be a useful skill to have in certain sporting situations, as when a hitter in baseball squares off against a pitcher in a tense showdown. At the end of the night, Shum did his imitation of face changing as he shook hands with people -- turning away and suddenly turning back with a funny, contorted expression.

The whole scene symbolized an issue crucial to Microsoft's future in China and the rest of the world. In the coming year, as it had begun to do in the past decade, Microsoft (or any other multinational concern) needed to keep up with the changing face of innovation. To do that, the company would need to evolve its own practices, improve its own relationships, and change its own face around the globe. Over the past six years, the rise of Microsoft's Beijing research lab had proved a major coup in the struggle to improve the company's image in the Middle Kingdom. But what was truly at stake was not image, but the company's very lifeline to innovation and, ultimately, survival: recruiting top foreign talent and creating the novel products that propel future growth.

That, as Chairman Bill might say, was the way to win the road ahead.Copyright ©2006 by Robert Buderi and Gregory T. Huang

---

Excerpted from Guanxi (the Art of Relationships): Microsoft, China, and Bill Gates's Plan to Win the Road Ahead by Robert Buderi, Gregory T. Huang
All rights reserved by the original copyright owners. Excerpts are provided for display purposes only and may not be reproduced, reprinted or distributed without the written permission of the publisher.