by Duncan Clark
Masayoshi Son publicly endorsed Alibaba’s strategy. Four years after SoftBank’s initial investment in Alibaba, he declared himself “extremely pleased” and predicted that “Alibaba has the potential to become another extraordinary success like Yahoo.” Meanwhile, Alibaba disclosed that revenues from its Alibaba.com B2B site had grown threefold in the previous year, propelling that business at last to profitability.
Despite the fresh backing for Taobao, eBay itself remained oblivious to the rising threat, considering itself far superior to this quirky, local rival. When asked by BusinessWeek in the spring of 2004 about rivals in China, eBay’s senior vice president Bill Cobb mentioned only one: 1Pai, a joint venture between Yahoo and Sina.
Jack reveled in being ignored. “During the first year, eBay didn’t consider us their rival. They didn’t even think that we could be their rival. They thought, We haven’t even heard about Alibaba. Such a strange name. Chinese all know what tao bao means, foreigners don’t.”
eBay was confident that its global network and experience would ensure EachNet pulled well clear of any competitors. But corporate bureaucracy, worsened by the extended and dysfunctional reporting lines all the way to San Jose, were to smother whatever embers of entrepreneurialism still burned within EachNet in Shanghai. eBay’s China adventure, lasting from 2003 to 2006, is today a case study in how not to go about managing a business in a distant market.
eBay’s first big mistake was to tell the market that China was already an ace in the hole. Meg Whitman deserves credit for recognizing China’s potential as an Internet market before that was popular. Her early interest in business opportunities in China was sparked by a family connection to the country: “In the 1970s my mom was invited to be part of group of women, led by actress Shirley MacLaine, to visit China. There were lots of reasons not to go: China was then an undeveloped country that had been closed to outsiders for many years. And my mom had just ten days to get ready. But instead of worrying about her safety, she seized the chance to have an adventure. Over four weeks, the group covered two thousand miles in China, mostly by train, visiting schools, farms, and villages.”
Whitman recalled that the trip “changed my mother’s life and, indirectly, mine, too. My mom learned Mandarin in subsequent years and returned to China eighty times. And after her first trip, she told my sister and me, ‘I’ve seen women doing all sorts of marvelous things—so realize you have the opportunity to do and be anything you want.’”
Mary Meeker, then an Internet analyst at Morgan Stanley, where she was dubbed the “Queen of the Internet,”21 was one of Meg Whitman’s lead cheerleaders. The dot-com bust had dented the reputation of nearly the entire Wall Street tech research community, but China played a big role in Meeker’s redemption. In April 2004, Morgan Stanley released a 217-page report under her name that profiled the Chinese Internet sector. It would be reprinted more than twenty-five thousand times. Meeker had a reputation for being a contrarian: “One of the greatest investments of our lifetime has been New York City real estate,” she said, “and investors made the highest returns when they bought stuff during the 1970s and 1980s when people were getting mugged. . . . The lesson is that you make the most money when you buy stuff that’s out of consensus.”
Whoever Wins China, Wins the World
Meeker now saw China as the Next Big Thing. Picking the right Chinese company to bet on wasn’t easy, Meeker said, so she recommended instead that investors leverage Silicon Valley companies with exposure to China: “Both Yahoo and eBay have interesting plays in the Chinese market. So our simplistic point of view is that one way to play the Chinese market is by owning Yahoo or eBay.” The report quoted Whitman, raising the stakes, as predicting, “Whoever wins China, will win the world.”
Meeker’s vocal support for Whitman, including cheering on her China strategy, helped boost eBay’s share price by 80 percent in 2004. But the climbing valuation obscured growing challenges for the company. A series of hikes in commissions sparked protests from its virtual store merchants, tens of thousands clubbing together to denounce “FeeBay” or “GreedBay.” Complaints reached a crescendo in February 2005 when eBay increased commissions by almost 3 percent (to 8 percent) on its worldwide mall. Whitman remained sanguine. “The thing to know about the eBay community . . . is that it’s been vocal from Day One.” She did concede, however, that the dissatisfied merchants had “perhaps been a tad more vocal than in the past.”
China became a useful way of distracting eBay’s investors from problems at home. Worse, before the company had even secured its position there, a “we’re winning in China” attitude, at both eBay and its newly acquired business, PayPal, ensured a form of collective denial even when confronted with signs that things were not going to plan. China was considered so important that managers, keen to present a positive story to Whitman and other senior executives, made sure that everything looked great on PowerPoint and sounded smooth on conference calls. But thanks to its own missteps as well as Alibaba’s competitive moves, this was increasingly at odds with the facts on the ground.
eBay’s biggest mistake was in getting the culture wrong. A “leave it to the experts” attitude demoralized the original EachNet team in Shanghai, as eBay executives were parachuted in from headquarters in San Jose or other parts of the eBay empire. No matter how skilled the new arrivals, most spoke no Chinese. They faced a steep learning curve to understand the local market. Key EachNet team members started to leave, their exit interviews revealing concerns that San Jose no longer involved them in key decisions. eBay had sent over a number of China-born executives, but most had studied or worked in the United States for many years, sparking misunderstandings or friction with the local team. EachNet found itself at a serious disadvantage to the 100 percent local Taobao.
This gap was reflected in the design of the two rivals’ websites. eBay moved quickly to align the EachNet site with its global site, revamping how products were categorized and altering the design and functionality of the website. This not only confused customers, but also alienated a number of important merchants who saw their previously valuable China account names had been deleted. This invalidated their trading history and forced them to scramble to reapply for new names on an unfamiliar global platform. Worse still, the Chinese website lacked a customer service telephone number. eBay’s China site, modeled closely on eBay in the States, looked foreign to local users, who found it “empty” when compared to local sites.
In website design, culture matters. In the West, websites like Google had become popular for their clean lines and uncluttered “negative space.” But to the mass market of Chinese Web users, accustomed to pop-ups and floating banner ads, they seemed static and dull. As you can see for yourself by opening taobao.com, successful Chinese websites are typically packed with information and multimedia graphics, requiring many scroll-downs to see the whole page. From its outset Taobao has been a website built by Chinese for Chinese. And it worked.
It’s not just the graphics that helped Taobao connect with consumers. Taobao structured its website like a local bazaar, even featuring innovative ideas such as allowing male or female shoppers to click on a button to display products most suited to their interests. The design of the site makes it the virtual descendant of the Yiwu wholesale market, where Jack and many other Zhejiang entrepreneurs draw their inspiration. The founder of another, niche e-commerce venture explained, “If you go to Yiwu, you can order as little as three pairs of shoes. One factory specializes in soles, another in the uppers, another factory—or perhaps a small village—specializes in the laces. Taobao tapped the motivation of those small merchants to make money.”
For companies like eBay and Amazon, their experience in the United States and other Western markets proved to be of little use. “E-commerce in China is very strange,” the rival e-commerce founder continued. “It started with C2C (consumer-to-consumer) and with nonstandardized products. This was unlike Amazon, unlike the conventional wisdom where you need
to start with standardized products, like books. The more standardized the supply chain, the higher the barriers for e-tailers. All the smaller, mom-and-pop stores selling nonstandardized products are more accommodating, more flexible in supplying goods. That’s unique to China. The lack of national supply chains removed the barriers to entry that exist in the West, making it possible for individuals to make money.22 By starting with C2C, it made the price factor very appealing. Individuals23 can be happy to make even five mao (less than 1 U.S. cent) on a sale.”
Again aided by its roots in Zhejiang, Taobao outsmarted eBay by having a better understanding of the country’s merchants, for whom membership has been free of charge from the outset. Just as free listings was a core principle for the B2B Alibaba.com, it became a key competitive weapon for Taobao.com, too. Buyers pay nothing to register or transact; sellers pay nothing to register, list their products, or sell online.
EachNet had started out with free listings, but faced with spiraling costs in August 2001, it started to charge listing fees to sellers, adding commissions on all transactions the following year. These resulted in a sharp reduction in the number of auctions on the site, but given the state of the VC markets, EachNet management felt they had no choice. The decision to start charging fees, core to eBay’s model, ironically fueled its interest in buying EachNet. But once eBay was in charge in China, it pushed the fee culture much more aggressively than Bo and his team. eBay’s vice president for global marketing, Bill Cobb, summed it up:24 “We’re mainly interested in making sure that we structure this to have long-term sustainability. We have the essential eBay format—the insertion fees, final-value fees, and features fees—though at a lower level.”
Meanwhile, Taobao’s decision to forgo charging fees was not without risk, since it forced it to look to other ways of generating revenues, especially if the site became popular and drove up operating costs. But making the site free for both shoppers and merchants turned out to be the key factor in ensuring Taobao’s triumph over eBay. A research paper25 that analyzed more than a decade’s worth of transaction data on Taobao concludes26 that in the early phase of the company’s history, attracting merchants, who in China are especially allergic to paying fees, was more important than attracting shoppers. Taobao’s popularity was fueled by a “virtuous circle”: More merchants and product listings meant more shoppers were attracted to the site, which meant more merchants and products, etc.
In addition to being popular with consumers, offering free services ensured that Taobao was not distracted by a persistent problem that plagued EachNet from the beginning: worrying about how to prevent vendors and consumers from figuring out ways to use the website simply as a place to connect with one another, then conducting their transactions offline or through other means. As Taobao charged no fees, they had no incentive to police this behavior. On the contrary, Taobao actively encouraged communications between the transacting parties by setting up bulletin boards and, beginning in June 2004, launching an embedded, proprietary chat window with the unfortunately in English named AliWangwang.27 Buyers on the site use the service to haggle with sellers, which resonates well with the vibrant marketplace culture in China. Communication is a key underpinning of commerce, but eBay users struggled to communicate with vendors.
Designed with input from Taobao users, AliWangwang is an early example of the type of “consumer-driven innovation” that drives successful technology firms in China today, such as the role that cell phone vendor Xiaomi’s fan club plays in suggesting new product features.
To this day, AliWangwang remains a popular feature on Taobao, allowing consumers to maintain their own list of personal purveyors—one, say, for cosmetics, another for baby formula—who are at their beck and call around the clock. Customer service on Taobao is so good that it can be overwhelming. A purchase on Taobao is often accompanied by a flurry of messages on AliWangwang, a series of virtual bows and scrapes from merchant to customer, who may have a hard time exiting the conversation.
But whatever the “pull” of Taobao, a decision by eBay in September 2004 would serve to “push” many of EachNet’s customers away. eBay executives in San Jose decided to “migrate” the China website to the United States. Instead of hosting the website close to customers in China, it was shifted to the States. In a borderless Internet, where a website is hosted shouldn’t matter. But China’s is not a borderless Internet. Today the Chinese government is actively promoting its vision of “Internet sovereignty” around the world: a rejection of the idea that a nation-state’s virtual borders should be less meaningful than its actual frontiers. In China, the effects of the government’s long-standing efforts to build and extend the “Great Firewall of China” often means websites hosted overseas are much slower to load than those hosted in China itself. All Web traffic accessing sites hosted outside the mainland has to go through a series of chokepoints where the request is screened. This is to ensure that a foreign website does not display material the Chinese government deems “sensitive,” including the “three T’s” (Tibet, Taiwan, and Tiananmen Square). These and other sensitive topics, such as unrest in Xinjiang, are widely thought to have been the reason that China has blocked some of the world’s leading websites, from Twitter to YouTube to Facebook and, increasingly, Google.
While e-commerce and online shopping typically don’t touch on these sensitive areas, the Great Firewall can often ensnare or seemingly block even anodyne activities or requests. For example, once eBay had moved its servers outside China, a user who happened to have a “64” or an “89” as part of his or her username might see their account blocked or be unable to access the Internet—the reason being that both numbers automatically trigger the censors in China as part of the effort to block any mentions of the events in Tiananmen Square on June 4, 1989 (6/4/89).
eBay had its reasons for the migration. As the business grew in China, the engineers in San Jose worried whether the platform built by a Shanghai-based start-up could cope. It turned out that EachNet had built robust technology, capable of scaling up by even a hundred times. But after a series of site outages that had damaged its reputation at home, eBay had become obsessed with the stability of its platform. eBay pushed ahead with the China migration anyway. The attraction of a unified worldwide site with a consistent set of features was just too hard to resist.
Some senior executives within eBay already knew that migration would be a mistake—the company had already seen the damaging impact in Taiwan—but bizarrely eBay managers in Taipei were blocked by migration-obsessed managers in San Jose from sharing their experience with the team in Shanghai.
As predicted, as soon as the China site was migrated and integrated into the global site, the impact on EachNet’s traffic was disastrous: It dropped off precipitously. Customers in China started to experience long delays and time-outs on the site. Why would they bother to wait for eBay in China—a site that charged fees—when Taobao was available instantly and for free?
Migration was also costly for eBay because the company typically carried out maintenance of its servers every Thursday at midnight on the West Coast, ahead of the peak traffic of Friday. But this meant the disruption happened at the peak of China traffic, fifteen hours ahead of San Jose. EachNet tried to adjust the maintenance schedule but with no success.
Meg Whitman had made China a key priority for eBay. But when migration caused traffic in China to plummet, no one told her. She found out only a month later on a visit to Shanghai, and she was furious at not having been kept informed.
Things quickly spiraled out of control for the company in China. Once the website had been migrated to the United States, all modification requests from engineers in China were stacked up in what the company called a “train seat” system. Departments would submit their requests for changes, and like an assembly-line process these were then lined up and consolidated into a “train of needs.” Changing one word on the site would take nine weeks. Changing one feature would take one year.
&n
bsp; How could eBay be so inefficient? There are two explanations. First, eBay had an effective monopoly in the States, and this bred complacency. Second, despite its Silicon Valley aura, eBay was never very strong at technology. One eBay executive famously once said, in public, “Even a monkey could run this business.” After the embarrassing site outages, stability and process trumped technology.
Once Taobao appeared on the scene, eBay’s “train seat” system quickly became a train wreck. EachNet executives desperately tried to signal the danger to senior executives in San Jose, but to no avail.
Although Taobao had its merits, Alibaba could hardly believe its luck as the ineptness of this supposedly world-renowned company became apparent. Jack compared eBay’s lumbering approach to a jumbo jet: “A global technology platform sounds great, like a Boeing 747 flying is great. But if the airport is a school yard, you cannot land. Even if you want to change a button, you have to report to, like, fourteen guys.”
Looking back on the fiasco eight years later in her new role as CEO of Hewlett-Packard, Meg Whitman was contrite about eBay’s missteps in China. “You’ve got to have a set of products uniquely designed for this market by Chinese. It is not a market where you can take a product or a system that works in Europe or the United States and export to China.”
She also concedes that migration was the fatal blow to eBay’s China ambitions. “We made one big mistake. We should have left EachNet on their own platform in China. Instead what we did was put EachNet onto the global eBay platform because it had worked everywhere. It had worked in Western Europe, it had worked all over. . . . We had bought all these baby eBays and basically migrated them to one common platform, which had a lot of advantages. One is cost. Second is speed to market, because when you roll ‘buy it now’ you could roll it to thirty countries as opposed to do it incrementally. But we made a mistake in China.”