Not Just Censorship: The Stories Behind The Tech Giants’ Defeat In China

In 2016, when Uber China was sold to its Chinese ride-hailing competitor Didi, tech observers were much less surprised by the news than when Google shut down its engines in China back in 2010. Since then, the trend of Western tech giants habitually failing in Chinese markets has firmly set in. First, there were eBay and Facebook, then Twitter, Youtube, and more recently Uber, What’s App and Instagram. I can hear you saying: “Wait, Facebook or Instagram didn’t fail, they were banned by the Communist censorship”. This may be true, but at the same time being banned may have actually allowed them a more graceful exit, since none of those big names had managed to become mainstream in China before they disappeared from public view. Facebook, Twitter and Youtube, in fact, had been in the Chinese market for years before the exit, yet could never compete with their local competitors. Why?

The reasons are often phrased obscurely − for instance: “It’s all due to the failure to recognise that the Chinese market and the business environment are very different from that of the West.” This statement could lead us to speculate about anything, except (crucially) product design and product strategy themselves. Are there lessons to be learnt here?

This is a question I’ve been thinking about for years, especially after moving to the UK in 2013. Although many accounts have been offered from various perspectives, while exchanging ideas with people here in the UK, I unsurprisingly found that many stories well known domestically among practitioners were in fact often neglected overseas.


The first story I want to tell is about eBay and its Chinese competitor Taobao, the rising star of Alibaba at the time. It happened quite some time back, in 2006. But it is still relevant today, in terms of what it can tell us about relationship-building between an online service and its customers, as well as between machines and humans. For instance, what Alexa and the newly released Google Duplex are trying to achieve is of ever-growing interest.

Taobao had nothing to compete with eBay at the beginning. In the early summer of 2003, it was a start-up on a shoestring, launching their shopping site using a free open-source stack. In the same year, the already multinational eBay, which had just acquired Paypal, decided to enter China’s market by purchasing what was China’s top e-commerce site EachNet, which held a whopping 90% market share in the country.

Fast-forward three years to 2006, and the story is completely reversed. While eBay still had a larger number of registered users in China, Taobao gobbled up the majority of eBay’s market share and became the No 1 e-commerce website in transaction volume, which led to eBay’s withdrawal from China.

Many observers credited the victory of Taobao to free listing. While Taobao’s free listing policy certainly helped win over the cost-sensitive East Asian sellers from eBay (which is why eBay also failed in Japan), what won the hearts of buyers even more was, in fact, the instant message service it provides. From the very beginning, Taobao offered its users an instant message tool called Wangwang (‘good fortune’ in Chinese), the icon of which was a little chat bubble. The chat bubble appears on the top of every product listing page, right below the purchase button. The idea is that before a buyer commits to buy something, they can click on the chat button and speak to the seller directly.

Why is this chat bubble so important? First, bear in mind this dates back 15 years − in the early days of e-commerce, the biggest issue was the trust barrier. For people who were shopping online for the first time, how could a website establish a trust relationship in the easiest possible way? Wangwang was Taobao’s answer. By allowing buyers to instantly communicate with sellers, essentially Taobao brought online the familiar experience of dealing and bargaining with real humans. Human interactions shortened the virtual distance between buyers and sellers. After all, humans are social animals, and there is almost no problem that cannot be solved by communication. Unlike machines, human users can make small talk, give assurance, build long-lasting relationships and think outside the box. In fact, Chinese netizens enjoyed the experience so much that some of the e-commerce “slang” even influenced their everyday language − for example, to address a customer as Qīn (亲).

What did eBay do back then? As a matter of fact, eBay did nothing differently than what it does today. The only way a buyer can communicate with a seller is through emails. Aside from the fact that emails are less inviting than instant chat − you never know whether or how long it will take for the seller to respond to your queries − emails also face a culture difference in China. Unlike in the UK or the US, people in China are less used to using emails: thanks to Tencent (the producer of two dominating instant chat apps, QQ and WeChat), Chinese netizens are instant-chat natives. Email culture has never really existed in China except within co-operations. I remember the days when I used eBay and had to contact a seller by email about a return (which was never responded to). I wondered why the hell I didn’t just use Taobao.

For customers who like to bargain, offering buyers an opportunity to chat with sellers is certainly a better cultural fit. However, bargaining is just one of the reasons for buyers to click the chat button. According to a study of Taobao, buyers typically spend 45 minutes chatting with different sellers before committing to buy a product. Instant chat allows buyers to ask much more personalised questions. For example, is there an express delivery option for the town they live in? When will the sold-out dress get restocked? These questions can certainly be addressed in emails; however, in the fast-paced modern retailing industry, buyers would probably have turned away before even asking the questions if there wasn’t a quick and easy method of communication readily available.

Nevertheless, eBay’s failure in China was not entirely due to its strategic mistake not to invest in buyer-seller interaction. Although constantly being compared with each other, eBay and Taobao actually rely on two different business models − eBay’s strength lies in the C2C market, while Taobao’s is primarily B2C − meaning that a large proportion of eBay sellers are casual or part-time, whereas, on Taobao, most sellers are professional traders. It makes sense for professional sellers to prioritise customer service, but likewise it is not feasible for casual sellers to be ready to engage with customers online at any time. Ebay’s model may work well in markets like the US. But in a country like China, where millions of small retailers and manufacturers cannot wait to embrace e-commerce, B2C is the natural survivor.

Instant chat does have its drawbacks as well. With 15 years now passed, and the trust barrier of online shopping long gone, busy shoppers don’t see the necessity of interacting with sellers before purchase. And yet the bar of customer service on Taobao has been raised so high that it has become standard for a seller to provide 8 hours plus customer service. (A new profession has also emerged in the past 10 years: third party Wangwang customer service providers acting on behalf of Taobao shop owners.) This higher standard discourages small sellers, such as individuals, from competing with big sellers. C2C transactions, such as second-hand goods exchanges, are way too rare on Taobao. Some sellers of course also use chatbots to replace real humans, which may or may not generate the same user experience.

On today’s Internet, the number of chatbots are perpetually increasing. But do they really provide the same experience as human-to-human conversations? Even an optimistic answer would be “not yet”. AI technology nowadays may be able to provide information and even fool receptionists. However, in order to take on challenges such as establishing trust, it would still be ineffective. It is certainly clear that scientists and engineers are making rapid headway in “Affective Computing”. Take Google Duplex for instance: what wowed the audience most was not how well the machine can process information (both simple and complex), but its very humanish “mmhmm”!