Baidu in the Past Year

Published: 2017-01-03

Preface

This article is the first in the business commentary section of my personal blog. Why choose this topic? It’s simple: for a first business commentary I needed a subject that was broad, rich in material, and highly topical. Naturally, BAT came to mind — a big IP that occupied tech media headlines for nearly half the year. I didn't write about Tencent first because I plan to read Wu Xiaobo's new book The Tencent Story before commenting. I didn't write about Alibaba because, although I've read many stories, I lack deep understanding and need to study it further. I chose Baidu because the past year with Baidu has been like something stuck in the throat — I had to write about it.

Baidu in the Past Year

The name “Baidu” comes from a line by the Southern Song dynasty poet Xin Qiji from eight hundred years ago: “Searching for him thousands and thousands of times in the crowd.” The line describes the poet’s persistent pursuit of an ideal.

In December 2015, Baidu announced a major organizational restructuring, forming a new Financial Services Business Group with Vice President Zhu Guang as general manager. Zhu joined Baidu in 2008 and had long overseen Baidu’s large-market, public relations, and government relations teams. Deeply experienced in China’s internet ecosystem, he has staged several PR cases regarded as textbook examples. After this change, the public relations team was handed to Baidu President and head of the Emerging Businesses Group, Zhang Yaqin. Zhang, with a background in multinational firms and technology, is mild-mannered and refined; he had just joined Baidu from his prior role as Microsoft’s Senior Vice President and Chairman of the Asia-Pacific R&D Group. At the same time, Baidu more firmly pivoted its strategy from connecting people to information toward connecting people to services, aiming to become a service-oriented company centered on AI and big data with finance, O2O, mobile social, and social commerce as core businesses. Baidu Tieba — Baidu’s only social product within its product framework — offered limitless possibilities for advancing the “connect people to services” strategy starting from the user end. This renewed hope for Baidu, which in recent years had gradually lost its halo among the BAT trio. In 2016, a “new Baidu” seemed poised to emerge.

In January 2016, just after the new year’s bells had barely faded, users on platforms such as Zhihu and WeChat began reporting that Baidu Tieba had sold the management rights of numerous disease-related forums, including “hemophilia,” to so-called “experts,” dubious hospitals, and third-party organizations. These included many cases of fraud and false medical claims that misled patients. Baidu faced harsh criticism from users and media and was swept into a public relations maelstrom; CCTV even ran a segment titled, “Ask Baidu: How many degrees of conscience?” The company then announced on January 12 that all disease-related Tieba forums would cease commercial cooperation and only be open to authoritative public welfare organizations; the hemophilia forum would introduce an NGO specializing in hemophilia prevention and treatment, “Home of Hemophilia,” as the new forum owner. However, the remediation did not meet external expectations: the diabetes forum’s moderator remained of “uncertain origin,” the Changge forum, previously focused on voice acting, was bought by an online game company and overnight shifted discussion topics from voice acting to gaming, effectively being ruined. Many high-quality forums related to finance, education, and regions also saw original moderator teams removed and new moderators parachuted in.

In May 2016, four months after the “hemophilia” Tieba incident, Baidu again became the target of fierce media criticism. A 21-year-old college student, Wei Zexi, died from synovial sarcoma after receiving treatment at the Beijing Armed Police Hospital No. 2; in reality the treatment was provided by a Putian-system hospital that had contracted a department of that hospital. The hospital used a treatment it knew to be ineffective, charged the patient nearly 200,000 yuan, and Wei tragically passed away. The information that led him to that hospital was obtained through Baidu’s paid search ranking results.

In July 2016, only two months after Robin Li was summoned by the Cyberspace Administration, The Beijing News reported that after 10 p.m. each night, searching Baidu for terms like “New Lisboa” would show gambling websites among the top promoted results. These gambling sites, which had stolen many companies’ business licenses, quietly launched at night and were offline again by 9 a.m. the next morning, restoring natural search results. Baidu responded that the involved sites opened accounts in April and that gambling information was content privately altered by companies late at night in violation of rules. But The Beijing News reported that the issue had been discovered and reported as early as December 2015 and remained unresolved after seven months.

In the subsequent months there were further incidents: pornography on Baidu Cloud, Baidu black-market takeout scandals, two investigations by the Cyberspace Administration, and agents coming to demand explanations. 2016 was an intense, highly visible year for Baidu; the Wei Zexi case in May was a peak — unprecedented in breadth and endurance. I suspect Baidu itself would have been prominently featured on any 2016 list of hot topics.

Losing Standing in the Field

For a time, alongside China’s rapid economic growth, the Chinese internet rode that wave and enjoyed fast, often brutal early growth. Eventually three major camps formed — Baidu, Alibaba, and Tencent — colloquially “BAT,” which established the industry’s order and rules: the Matthew effect and jungle law. But over time, with shifts inside the BAT companies and waves of startups, that earlier BAT dominance began to wobble.

An employee recently joked that BAT has quietly become TAB, which they tolerated, but now it’s slowly turning into ATM, with ‘B’ being dropped from the lineup — and they can’t sit still about that. Changes in how companies are referred to reflect external doubts about Baidu. In capital markets, Tencent — initially the underdog among BAT — is now the favorite, with a market cap of about $231.6 billion. Tencent’s early diversified bets are now generating scaled profits; it leads the industry in gaming and social, offering investors big upside. Alibaba, though less dazzling than before, remains steady and resilient with a market cap roughly comparable to Tencent at about $224.7 billion. Baidu, once China’s top internet company, reached a peak stock price of $251.99 on November 28, 2014, with a market cap of $86.8 billion. Two years later its market cap was about $58.4 billion, roughly one quarter of Tencent’s at the same time. What caused Baidu’s position to erode? I’ll examine Baidu’s troubled 2016 to find clues.

Debate over Profit Models

As a mature domestic internet heavyweight, Baidu’s revenue model is now relatively clear and stable. But a comparison of BAT revenue composition exposes a flaw in Baidu’s structure: compared with Tencent and Alibaba, Baidu relies excessively on online marketing. For a company of Baidu’s scale, a single revenue source increases operational risk. Baidu’s situation demonstrates this: controversies over Baidu’s paid search ranking have hit its online marketing revenue hard and even risk contraction, while its earlier investments in big data and AI remain loss-making. This leaves Baidu in a precarious, uncertain state.

Baidu’s online marketing revenue primarily comes from its paid search and the upgraded but fundamentally similar Fengchao (Phoenix Nest) system. Google, another search-origin giant, also derives a large share of revenue from online marketing, yet Google’s search has not been criticized the way Baidu’s has; Google is often praised in comparisons with Baidu. Let’s analyze differences in Baidu’s and Google’s profit models.

For both Baidu and Google, advertising is their main revenue source, so neither is essentially a hardware or software company — they are advertising companies. But while both sell ads, their selling methods differ significantly.

Take Baidu: when searching, you typically see two main parts on the results page — the left side with ranked search results and the right side with sponsored recommendations. The left side reflects the domestic-first approach Baidu pioneered: a performance-based network promotion charged by visits. In simple terms, advertisers bid on keywords; the higher the bid, the more likely their result will appear higher for that keyword, attracting more clicks and yielding better advertising performance. The right side is the “Baidu Hot Zone,” where advertisers also bid on keywords but pay a fixed fee for time slots (e.g., a year), with different positions priced differently.

Google’s main ad product, Google AdWords (now Google Ads), sells keywords to the highest bidders and displays ads in order of price on the right side of search results, charging only when visited. The key difference is that Google’s left-side organic rankings remain determined by metrics like traffic, popularity, and PageRank, so ads do not interfere with natural search results, preserving fairness and impartiality.

Further, Google leverages powerful data analysis to deliver precise ad placement and thereby increase revenue. Data analysis requires raw material like user behavior data. For example, many use Google Maps; aggregated user map data reveals who lives in an area, what they consume, and their purchasing power. Such data is critical to advertisers and greatly improves ad effectiveness. As advertising guru John Wanamaker said: “I know half my advertising budget is wasted; I just don’t know which half.” Precise targeting solves that problem. By offering many free services, Google gathers the user data needed to predict tastes and needs accurately, enabling targeted ads that increase clicks and ad spend. In this three-way loop — users receive timely, relevant product and service info; advertisers gain efficiency; Google earns significant ad revenue — everyone benefits. Baidu also invests in data analysis, but its paid ranking model cost it the moral high ground.

If you want a more detailed, in-depth understanding of Google’s ad system, you can read the discussion on Zhihu“How does Google make money?”Regarding this question, Sim Chen’s answer quotes parts of Professor Wu Jun’s The Wave Crest and explains the topic thoroughly; I recommend reading it. Google’s sophisticated revenue model was not perfectly designed by Larry Page and Sergey Brin at the company’s founding; many adjustments were driven by external regulation, which I will discuss in the final section. Now let’s look at Tieba’s monetization model.

Baidu set a 10-year goal in 2010 to reach 100 billion yuan in revenue. With slow growth in online marketing, Baidu tried to shift away from its single revenue stream and find new growth sources. As Baidu’s only social product, Tieba needed to monetize after years of operation; Baidu’s search profits were generated via agency sales, so Tieba naturally followed a similar agency model. Commercializing Tieba inevitably created conflicts between original forum owners and the new agents, and Baidu’s oversight didn’t keep pace, which led to the Tieba incidents.

Over years of development, Tieba achieved accomplishments to be proud of: more than one billion registered users, over 300 million monthly active users, and 19 million topic forums. Tieba’s daily active users and visits far exceed Tianya, Zhihu, Douban, and Renren combined. Tieba aggregates large, precise niche user groups, so many highly popular forums inherently possess marketing value.

Since 2010, Baidu set a 10-year target to reach 100 billion yuan in revenue. With tepid growth in online marketing, Baidu internally tried to shift its single revenue structure and find other profit points. As Baidu’s only social product, Tieba has a solid user base and strong user stickiness. Given Robin’s deep fondness for Tieba and hopes for social commercialization, after years of operation Tieba needed to monetize.

Monetizing Tieba was imminent. In 2014, Tieba announced a commercialization strategy and many companies began taking over related forums, including NetEase Games, iQiyi, Tmall, and JD. But monetization proceeded slowly. As with Baidu search, which monetized through agency advertising sales, Tieba adopted a similar agency model. In 2015, Tieba piloted a “Tieba Partner” system intended to let commercial entities co-manage forums with users. Previously, moderators were appointed after user applications and Baidu’s assessment of activity and influence. A forum’s quality depends heavily on its moderator. After the “Partner” system, moderator appointments, ad displays, and management rights were outsourced to agents. The problem lay in this blunt outsourcing: agencies prioritized commercial interests, undoing the painstaking work of original moderators. It’s understandable that original moderators and users pushed back. Thus, Tieba’s problem was the conflict between original moderators and agents; the rushed commercialization lacked matching oversight, leading to the 2016 Tieba incidents. The resulting management chaos and trust erosion will make future commercial efforts far more difficult.

Executive Cabinet Turmoil

I mentioned Baidu’s December 2015 reorganization earlier; most notable was Zhu Guang leading the newly formed Financial Services group and Zhang Yaqin taking over PR and government relations — a change in leadership for a team that had been under the same person since 2008. What followed in 2016 were one crisis PR case after another. Shortly after the Tieba scandal broke, Robin Li said in an internal meeting: “We’ve done poorly in operations, product, and public relations. Especially this time, our PR response was not timely, producing a major crisis; public opinion ran away with things and our voice was drowned out.” Clearly, senior leaders chose to let the PR team take the fall. Zhang Yaqin, inexperienced in crisis PR and less than a month into his role, was understandably overwhelmed by the sudden, sharp public relations crisis. During the Tieba outbreak, Baidu’s PR office burned the midnight oil daily trying every possible way to limit negative impact.

Below is an organizational chart after Baidu’s major restructuring:

In 2016 Baidu fired or sent more than 30 employees to court. The number and seniority of executives involved shocked the industry. In April 2016, E-staff vice president Wang Zhan was dismissed for harming company interests. In early November, another E-staff member, and once-touted “Baidu prince” Li Mingyuan, resigned amid allegations that his participation in an acquisition, management of design business scope, and external personal investments might have harmed company interests. Coupled with multiple internal corruption cases, the external perception became that product lines handling money — Tieba, games, key-account sales, group-buying, takeaway services — had become breeding grounds for corruption.

Quoting a report in Computer World, Li Mingyuan was seen by many inside Baidu as a casualty of internal infighting with Xiang Hailong; more bluntly, a failure of the Robin Li faction against Robin’s wife Ma Dongmin’s faction. According to an internal employee, even Wang Zhan didn’t have economic issues mentioned in internal emails, and internal emails went only to director level. Li Mingyuan’s case, ostensibly letting him off, in fact left an indelible mark of internal exile. Such infighting had become routine across Baidu’s departments. “When even internal employees anxiously speculate and internal infighting becomes mainstream, the company’s death is not far off,” an employee said sorrowfully.

In September 2016 Baidu announced Zhang Yaqin’s transfer away from public relations and government relations, replaced by Wang Lu, former CEO of Yihaodian. In the same month, Baidu strategy advisor He Haiwen (Helen) quietly resigned. Over the year, a string of previously silent then abrupt personnel changes invited speculation about hidden truths. When power rent-seeking expands, corruption and infighting naturally sprout. Can we blame only these professional managers? For a Baidu so deeply marked by Robin Li, we cannot avoid examining the founder’s responsibility.

Failing to Guard the Domain

Twenty years ago Robin Li wrote in Silicon Valley Business Battles: “Do not cling rigidly to the company you founded — that lesson comes from many painful experiences. Not everyone who builds an empire is suited to run it indefinitely.” Twenty years later, he must confront the advice he once penned.

Robin Li was born in Yangquan, Shanxi Province, and has long been a source of pride among Shanxi merchants. As the only son among five children, he was treated as a protected child and received extra family attention. Through hard work he entered prestigious schools like Peking University, went to the U.S., and in a few years rose to become a star entrepreneur in China. That pampered upbringing and repeated successes fostered pride, stubbornness, and a tendency to insist on his own decisions. This contributed to the distorted paid-search monetization model; despite criticism, Robin Li took pride in it because he had personally shaped it in China — like a child of his. His self-centeredness made product strategy unpredictable: some projects were stopped abruptly; employees wrote him letters but received no explanation.

For example, in mid-2015 Baidu announced a 20-billion-yuan investment to develop Baidu Nuomi, widely interpreted by media as a pivot to O2O to link LBS and payment services. Yet in April 2016 Robin Li declared the company’s focus would shift to autonomous driving and AI. Whether O2O would remain or be abandoned remained unclear. During the May Wei Zexi controversy, a former employee’s post titled “Former Baidu Employee: Why Baidu Can’t Let Go of Medical Information” trended in tech media. The post blamed founder Robin Li as the main responsible party. “Factory director Li Robin now finds it hard to hear critical opinions and outside criticism; his first reaction is often ‘I didn’t do anything wrong — they’re attacking me!’” This transformed Baidu’s culture from a “wolf culture” to one of sycophancy: “Whenever negative news hits Baidu, the exec group adopts a victim stance with little reflection. Some even chant slogans of love and loyalty to reassure the ‘factory director.’”

Baidu felt like Robin Li’s personal company. Over the years figures including Panorama CEO Pan Haidong, writer Han Han, and Robin’s Peking University classmates wrote him letters; Zhou Hongyi repeatedly criticized him by name — yet he remained unmoved, never responding or arguing, living in his own world like a spoiled child. Of Baidu’s seven founders, by 2010 only Robin Li remained; the early co-founders and later professional managers had left. Those still around had joined as early as 2004. This made it hard for an already solitary, reserved founder to find long-trusted colleagues to confide in.

In a company that grows from a startup, you need a public-facing soul and also several trusted lieutenants to handle internal affairs. Tencent’s Pony Ma has Zhang Zhidong and Zhang Xiaolong; Alibaba’s Jack Ma has Joe Tsai, Peng Lei, and a roster of senior lieutenants. Although Baidu still had Zhu Guang, Zhang Yaqin, and Xiang Hailong, it lacked a definitive number two; everything was tightly controlled by Robin Li. Today Baidu is essentially Robin Li and tens of thousands of employees. In such a large company, one person cannot manage everything; you need seasoned insiders who treat the company like home and handle detailed internal matters. But early founders left, eroding startup culture. Under Robin Li, people work to complete tasks, hit goals, and earn bonuses — they won’t take on unpleasant housekeeping. With cultural and value collapse, KPI metrics and data became the chief communication tools between Robin Li and employees. I believe the collapse of culture and values is the root cause of these problems.

Encouragingly, Robin Li’s statements over the year suggest he began to rethink Baidu’s state. After the January “hemophilia” incident, a purported screenshot of Robin Li’s remarks circulated online saying: “Baidu’s values are good and noble, and there is no perfect business model; Baidu’s business model is fine.” After being summoned by the Cyberspace Administration over the Wei Zexi case in May, Robin Li wrote in an internal letter: “If we lose user support and fail to uphold our values, Baidu will be bankrupt in 30 days!” He began to recognize that “the chase for short-term KPIs squeezed and distorted Baidu’s values and distanced us from users.”

Failures of the Regulatory and Business Environment

Stepping back from Baidu: in the Wei Zexi tragedy, blame also falls on fraudulent Putian hospitals and the heavily fortified Beijing Armed Police Hospital No. 2. Together they reveal a chaotic medical field that forces the question: where is regulation? In China this problem isn’t limited to healthcare — food safety, securities, real estate, and other traditional industries have produced regulatory scandals. In the rising internet sector, regulation is even more patchy; many legal provisions are blank and unenforceable. In a market lacking effective oversight, wrongdoing carries minimal cost and companies tilt the balance toward profit over business ethics. Didi watch-ride passengers were murdered, fake goods proliferate on Taobao, fake tickets on Trip.com, fraudulent listings on 58.com, black-market takeout on Ele.me — these examples show that after nearly twenty years of rapid growth, China’s internet companies often prioritize profit over ethics.

Can a company’s behavior be constrained by its integrity, social responsibility, self-regulation, or the conscience and charisma of executives alone? History suggests not. A mature market economy relies on sound legal systems and penalties for wrongdoing. Google has also erred: in August 2011, Google reached a legal settlement with the U.S. Department of Justice over illegal pharmaceutical ads, paying $500 million to avoid prosecution. Google reported blocking 10,000 sites selling counterfeit goods in 2015, closing 18,000 offending accounts, and blacklisting 30,000 sites selling weight-loss products. Baidu, over 17 years, faced issues like IP infringement, paid ranking, and pornography, but each crisis resulted only in minor fixes without addressing root causes. From that perspective, had Google grown in China, could it have become today’s global tech leader? It might have become a second Baidu.

Too-low costs of violation — and penalties that vary by a company’s nature or size — are the root of many commercial sins of Chinese firms. The core of a market economy is fair competition, and fairness depends on rule of law; thus a market economy is fundamentally a legal-rule economy. Only proper regulatory mechanisms can create a healthy business environment.

Finally, a personal view, perhaps narrow and for reference only. Wei Hanfeng, former editor-in-chief of Bloomberg Businessweek China, once said: “Undoubtedly, a company’s fundamental purpose is to maximize commercial returns.” If pursuing profit is natural for businesses, then a ruling party pursuing political interests is likewise to be expected. We are a socialist country with a single-party system; in theory the party represents the fundamental interests of the people, so its political goals need not conflict with those interests. But this is like an idealized perfectly competitive market — approachable but never fully attained. In Western systems the ruling party faces opposition parties, and market order is regulated purely by laws enacted by successive governments. In our country, the hand of the state is more visible in markets, making our regulatory mechanisms less pure and less autonomous even if well designed. Google’s 2010 exit from China, which left Baidu unchecked, is a good example.

Conclusion

Above I analyzed Baidu’s problems from four angles. The most direct result is how it affects product experience. Below are two random screenshots I captured from mobile, showing Baidu’s most proud and most profitable product — the mobile search product that underpins the company’s survival.

I imagine a foreign friend hearing about a Chinese high-tech company doing work as impressive as Google’s, typing the domain and hitting enter, then seeing a homepage that bears no resemblance to a high-tech firm — they would surely say, “Are you kidding me?”

Finally, a joke from Lin Mo, the ‘old teacher,’ seems especially apt; I’ve adapted it slightly:

2016 was the year the whole internet cooked Baidu. Stir-fried it, steamed it, braised it, tossed it in a chaotic stew — in short, vented rage. Yet tomorrow we still go to work and have to keep using it.