[2014] Baidu: A Year of Tech Slogans, So What Happened to the “Wolf Spirit”?

Top 15 Year-End Reviews: Baidu

Regular article

Of the BAT trio, Baidu had the dimmest 2014

Baidu's failed box-office forecast for The Golden Era squandered one of its few chances in 2014 to show the world what its big data could do. Meanwhile Alibaba and Tencent made acquisition after acquisition, and Alibaba grabbed the headlines with the largest IPO in US stock-market history. People began to ask where Baidu really stood within “BAT.”

In 2012, Robin Li likened the mobile internet to drunk driving: “crazy and dangerous.” By 2014 he had to pour more of his energy into mobile, where demand for search is inherently weaker and other products have already staked out the vertical niches. If Baidu wants to repeat its PC-era dominance on a small screen, it needs more disruptive ideas.

Its maps had already won Baidu an early ticket to O2O (online-to-offline services). But that lead has been worn down as Tencent and Alibaba fight over offline gateways: ride-hailing, food delivery, local services and more. Worse, a year on, Baidu still seems to be dithering at the fork between payment tools and offline promotion.

What has left Baidu some grounds for confidence is its traditional strengths: search and its stockpile of data, its expansion into voice and image search, and its bets on deep learning. It may also have been short on luck. It took a box-office forecast that was meant to be icing on the cake and turned it into thoroughly bad publicity.

As a China-concept stock that has been listed on Nasdaq for 9 years, Baidu has one striking habit: whenever people doubt its growth, it turns in a decent earnings report. But now that industry leaders like Alibaba and upstarts like Momo have jumped into the same US market, hot money seems to be drifting away from a Baidu whose next source of growth is hard to see.

Still, it is trying hard to look like a technology company

To remind itself that it is a technology company, Baidu did put its technical strengths to some use. It worked with the United Nations Development Programme on big data for environmental protection and health. It followed Google's lead in smart hardware closely, releasing a string of concept smart-hardware products in the second half of the year. And when Wanda announced at the end of August 2014 that it would set up the Wanda e-commerce project with Baidu and Tencent, Baidu contributed some funding and big-data forecasting.

Robin Li's technical background has inevitably shaped Baidu's culture and management philosophy: an engineer culture that respects algorithms, keeps human hands out of how products actually run, and relies on technical innovation to drive growth.

But at Baidu, where sales and business-partnership staff make up as much as 70% of the workforce, engineer culture means something different. Unlike Google, Baidu pays closer attention to the experience of ordinary consumers, and its product managers, not its engineers, decide what gets built. The engineers' job is to find the best technical solution for whatever product the product manager wants. To some extent, this has limited Baidu's ability to innovate.

Baidu still spends heavily on R&D. In the first three quarters of 2014, its R&D spending was 1.277 billion, 1.737 billion and 1.832 billion yuan, slightly higher than Google's and up more than 50% year on year in every quarter.

But Baidu's best-known news this year was that it poached two top technologists, one after the other, from old rivals Google and Microsoft: Andrew Ng, the AI expert hailed as the “father of Google Brain,” and Zhang Yaqin, the guiding spirit of Microsoft Research Asia. Ng now runs Baidu Research, which is expected to build the world's largest artificial neural network in 2015. Zhang has become Baidu's president, in charge of overall strategy.

Taken as a whole, though, Baidu's moves this year added almost nothing to its results, nor did they show a competitiveness to match the company's standing and size.

Take Baidu Brain. According to Robin Li, artificial intelligence, machine learning and hardware support on the scale of tens of millions have already given “Baidu Brain” the intelligence and learning ability of a 2-3-year-old child. But no one can give a firm timetable for how fast it will grow, or for when and how it will be merged with Baidu's search.

Or take smart hardware. Behind these devices lie Baidu's need for more varied user data and its ambitions in smart-hardware operating systems. In this it resembles Xiaomi, which is using MIUI to compete for several hardware gateways. But Baidu has no advantage in hardware or in hardware distribution, so the promise of more user data may well turn out to be a contradiction.

Then there are Andrew Ng and Zhang Yaqin. Hiring people from big companies with far stronger R&D than its own may indeed help Baidu make technical breakthroughs. The world's largest deep neural network that Ng is building, for example, could one day help in government, NGOs, manufacturing, finance, retail, education and other fields. But Zhang spent 15 years at a Microsoft Research that was already in decline. With him as president, can Baidu still give the market what it expects?

You could say Baidu is laying groundwork for where the industry will go over the next few years. But it is also possible that Baidu is losing its knack for disruptive innovation.

Fortunately, Baidu is at heart a marketing company

Baidu is of course more than a company that toys with internet and hardware concepts. In 2014 its main job was still to defend the lead it had already built in the part of its business that makes money: advertising.

In each of the first three quarters of 2014, Baidu's total revenue grew more than 50% year on year. Net-profit growth fluctuated but stayed above 15%. Online marketing made up more than 95% of total revenue. By comparison, online marketing's share of Google's revenue has fallen below 70%. Baidu's dependence on a single source of revenue has begun to draw wide criticism.

Baidu has always been strong in marketing, and all its success rests on its share of search

Google has long held more than 70% of the global search-engine market. In China, even as Qihoo 360 has rapidly gained share, Baidu still holds more than 50% of Chinese-language search. Baidu currently handles about 5 billion search requests a day.

Baidu's search advertising is built on this enormous traffic. It is a classic asset-light, high-revenue business: Baidu charges only 42% of what Google does per ad impression, and its cost per click is only about 30% of Google's, yet it is highly profitable and growing fast.

Much like Google's AdSense, Baidu has gone beyond its own services and formed the “Baidu Union” with more than 600,000 websites, which jointly promote its services and carry its search to hundreds of millions of computers. Ma Guolin, general manager of the Baidu Union development department, expects the union's revenue share to reach 7 billion yuan in 2014, up 91% year on year.

While Tencent made games and Alibaba did e-commerce, Baidu stayed highly competitive in advertising revenue.

Baidu's search box also does more complex work as a gateway to the internet. In July this year, Baidu redesigned its search home page and built more functions and services into the box's suggested results.

But as we noted when Baidu invested in Uber, in the mobile era Baidu faces gateways that have been carved up and user data that is already scattered. After several years of cultivation (or cash-burning) by vertical apps such as Dianping, ride-hailing apps, and music and video apps, some Chinese users have grown used to getting services through separate gateways. With so many gateways, even small companies can control their own traffic without leaning on a giant, which gives them a chance to win users and make money on their own. Compared with its near-monopoly on the PC, Baidu's problems on mobile are more numerous and more complicated.

So investors don't really care about its technology; they care whether it can stay competitive in advertising as it moves into new businesses

In 2013, Wall Street split into bears and bulls on Baidu, roughly evenly matched. That sent Baidu's share price to a low at the end of 2013, followed by a strong rebound in 2014. In the filings Baidu disclosed for the first quarter of 2014, ten Wall Street heavyweights, George Soros among them, appeared prominently. Together they added 4.37 million shares, worth US$540 million, a sign of their confidence that Baidu's stock would keep rising.

Unlike US tech stocks such as Google and Amazon, Baidu is not judged on how strong its technology really is. What investors care about is whether it can stay competitive on revenue and profit margins, especially in mobile marketing, which it is still feeling its way into.

According to Analysys International, China's mobile internet market will reach 548.7 billion yuan in 2014, and mobile advertising revenue 27.3 billion yuan. As advertisers warm to mobile ads and shift more of their budgets there, mobile marketing is entering a period of rapid growth.

Baidu is moving aggressively to mobile, and at first glance its mobile strategy is going fairly well this year. Over the first three quarters of 2014, mobile's share of total revenue rose steadily from 25% to 36%. That figure, of course, combines 14 mobile apps with more than 100 million users each, including Mobile Baidu and Baidu Maps.

The other strand of Baidu's mobile strategy is to use its position as the market's largest app-distribution channel to move into new markets. A CNNIC report from the end of August 2014 put the combined user share of Baidu's distribution platforms (Baidu Mobile Assistant, Android Market and 91 Assistant) at 61.3%. On the Q3 2014 analyst call, Robin Li said Baidu might formally enter the games business through 91 Wireless's distribution channel.

But with Tencent and Shanda long entrenched in games, and Alibaba eager to join in, how Baidu plans to break in remains anyone's guess.

Everyone knows its rivals are “A” and “T,” yet out of habit it still seems to be watching Google

Everyone, Robin Li and Baidu included, knows that Baidu's real competitors today are Alibaba and Tencent, not Google. Yet in its strategy and in its habits, Baidu still watches Google's every move.

What kind of company is Google? One that gets to have its cake and eat it too. There is the “traditional Google” under Larry Page, which follows the market, strengthens today's business model and expands market share; and there is “Google X,” run by Sergey Brin, which sketches out the many things Google could become. In the Chinese market, the only part Baidu can realize and make the most of is the “traditional Google.”

When business is going smoothly and nothing stands in the way, it is very common for companies to stick to the way they have always operated. As Peter F. Drucker put it, every successful big company in history has, when faced with the “unexpected,” refused to accept it.

In the industry and inside the company, Robin Li has another label: “the pragmatist.” His forceful, final-word style of management at Baidu makes it hard for the company to broaden its path or its range of businesses. When “pragmatism” hardens into a short-sighted belief that one formula fits everything, it can become a “new barrier” to Baidu's own progress, or to its entry into other open markets.

Baidu should take a fresh look at who its competitors really are and, more important, draw up a strategy aimed at them.

Fortunately, it still has Baidu Maps

Robin Li once named traditional search, the mobile cloud, LBS (location-based services) and international expansion as Baidu's four future directions. LBS may be a slightly dated idea by now, but that has not stopped Baidu Maps from becoming another “super gateway” app after WeChat and Alipay Wallet. Baidu's location services now have more than 100 million active users and handle more than 2 billion location requests a day.

Analysys International data for Q2 2014 give Baidu Maps a 26.6% user share of China's mobile map and navigation apps, second only to Amap's 33.4%. But because Baidu Maps bundles in more local services, while Amap sticks to navigation, Baidu Maps edges ahead on daily use and visibility.

The core of location services is finding local information nearby and getting directions, and more than 200,000 merchants have already pinned their locations on Baidu Maps. Built around the maps, services such as Nuomi, Baidu Food Delivery, private-car hire and Baidu Wallet have begun to roll out. Each of these still has a tiny market share and needs a better user experience, but Baidu Maps remains the product most likely to give Baidu a closed-loop O2O ecosystem. Competition there is also fiercer. As the chart below shows, all three BAT companies have built complete ecosystem chains around maps and payments, but Alibaba and Tencent are far stronger than Baidu in the key gateways to everyday services and in payment tools.

All talk of buying, no buying: Baidu missed some chances this year

After Baidu released its 2013 results at the end of February 2014, CFO Jennifer Li told an analyst call that Baidu would keep expanding its platform in 2014, with strategic investment aimed at mobile search, cloud services, LBS, consumer products and international business.

Well, a year has passed, and apart from its strategic partnership with Uber, we have seen no big move from Baidu in any of these areas. Over the past year Baidu has been remarkably quiet.

Baidu told QDaily that one reason was that it was still digesting last year's “inventory” of acquisitions.

In 2013, Baidu bought the video site PPS and the mobile-app distribution platform 91 Wireless, for a combined total of nearly US$3 billion. At the start of 2014 it also announced that it would buy the group-buying platform Nuomi outright. Given the work of merging PPS with Baidu's iQiyi platform, and 91 Wireless's standing as China's largest third-party app distribution platform, Baidu does need a long time to absorb these assets.

Seen another way, Baidu's acquisitions still dig deeper into areas where it is already strong (video, mobile apps, O2O and so on). Unlike Alibaba and Tencent, it has not boldly moved into fields it is weak in, such as health, sports and payments. Even within its strong areas, Baidu has not used its signature marketing skills to tie the pieces together or to make them work as a platform.

This is one reason Baidu has failed to build an ecosystem that can rival Tencent's or Alibaba's.

Looking ahead, Baidu needs a bigger jolt

All in all, the Baidu of 2014 was a company whose results reassured investors and advertisers but whose products and strategy left users unsatisfied. Still, we believe that with its huge traffic and its market position, China's largest search engine and advertising platform has some good days ahead.

Revenue and profit growth have begun to slow, but the trend is still upward. Baidu's high price-to-earnings ratio holds strong appeal for investors, and its conservative valuation can still attract big-spending advertisers. And as traffic shifts to mobile, Baidu's mobile share of revenue is expected to pass 50% in 2015.

But that is not necessarily good news for Baidu.

In 2010, when Google pulled out of China, Baidu lost its biggest competitor, and every advantage showed up in its excellent results at the time. Then in 2012 Robin Li suddenly revived the “wolf culture” Huawei once championed. He hoped to shake the company out of its complacency and make Baidu wake up early to the threat from Qihoo 360 and Alibaba, and to the rise of the mobile internet as a whole.

As it turned out, that jolt worked reasonably well. But in 2014 we sense that the big-company disease may be coming back at Baidu. Robin Li needs a bigger jolt, yet this year he has stopped even mentioning the wolf spirit.