[Top 15: Baidu] Tieba Is Full of Contradictions. Can It Give Baidu the Future It Wants?

Baidu makes its money from search, but it also owns Tieba, one of the longest-lived social products on the Chinese internet. In the mobile era, can Baidu still rely on it?

"Tieba's best days are still ahead. I promise you will like the Tieba of the future even more." Robin Li said this on December 3, 2013, at the 10th-anniversary celebration of Baidu Tieba. He spoke not as Baidu's CEO but as the moderator of the "Robin Li Bar."

The "Tieba of the future" Li had in mind is a mobile Tieba, part of the mobile transformation of Baidu that he talks about at nearly every public appearance these days. At the end of 2014, Baidu's mobile traffic overtook its PC traffic for the first time, and Baidu said 70% of the posts made on Tieba each day came in through mobile. In the first quarter of 2015, mobile brought in more than half of Baidu's revenue for the first time. But that money came from old strengths such as search and maps, not from Baidu Tieba.

"Born for interests" has long been Tieba's brand promise. At the start of this year, though, Tieba changed its slogan to "Join Tieba, find your crowd," saying it wanted to meet users' many social needs as fully as possible and make Tieba more appealing. It hired the ad agency Changrong Communications to promote the rebrand. The agency's subway and online ads deliberately blocked out the word "crowd" to create buzz and suspense, and Tieba later released a dedicated HTML5 ad as well.

That got us thinking: Baidu is playing up its social side, so can Tieba make up for what Baidu lacks as a social company?

As a search engine, Baidu connects anyone with any piece of information, and Tieba is its only product with a social side. Yet on mobile, social looks far more promising than search. That is why Google tried so hard to make its own social product, Google+, work. Despite repeated efforts, and even a change of CEO that brought co-founder Larry Page back to take charge, Google got nowhere.

And what has Baidu, the company most like Google in everything from business model to core service, done with Tieba? Can Tieba strengthen Baidu's user identity and become the future Robin Li hopes for?

Compared with Google+, Baidu Tieba has survived better than we expected, though as a social product more than 10 years old it cannot shake a sense of decline.

Tieba now claims 1 billion registered users, 8.2 million topic forums (called "bars"), nearly 300 million monthly active users, more than 100 million topics a day and more than 2.7 billion page views a day. These are the figures Tieba keeps repeating in its promotion.

The platform's finest hour came with the "Super Girl" singing contest. Before the final ten years ago, posts about the top three contestants passed 10 million, an average of 4 users posting at the same moment every second. Alexa data at the time showed Baidu Tieba's traffic even exceeded Sina's, making it the largest Chinese-language site in the world.

This is actually how they looked back then...

"This was where I first saw the power of UGC [user-generated content]: a group of people gather around an interest, and as they bounce ideas off each other they produce so much excellent, no, brilliant content, some witty, some beautiful, some expert. ... Anything an official body organizes and produces looks pale next to UGC." That is how the Zhihu user "Wang Dongming" described to QDaily how he once felt about Baidu Tieba.

Tieba once stood for a different idea of social life online, built on shared interests, apart from forums such as BBS boards and Tianya. Set it beside the internet products that appeared around the same time 10 years ago, and it is one of the few still alive and kicking.

Xiaonei, the forerunner of Renren, founded in 2005, was once hailed as China's future Facebook, yet most of its users probably posted their last status update two or three years ago. Kaixin001, launched in 2008, once had office workers hooked on "stealing vegetables" and "grabbing parking spaces"; it was beaten by Renren and has been silent ever since. Douban, which has also lasted 10 years, is doing somewhat better, though more and more of its longtime users complain about it.

Without exception, these were products that thrived on PC, and many more like them have vanished altogether. Baidu Tieba still matters, but it is riddled with problems and contradictions.

Acquaintance networks versus interest groups

On a social network, a person is a person. On a traditional BBS, people are grouped by interest and topic.

The idea for Tieba grew out of search itself: people who search Baidu for the same keyword may share an interest or taste. If a product could bring them together and make the new content they produced searchable, it would draw in more people like them, creating a virtuous cycle and eventually a network built on shared interests.

Baidu Tieba's HTML promotional page

Tieba uses a discussion structure close to a BBS, with no strict limits on topic. Ordinary users have great power to create a bar for any keyword that does not have one yet, so any niche interest can quickly gain a community of its own. Baidu's technical muscle simplified or hid the once complicated work of building a site, filtering spam and handling SEO.

Qu Bo, now CEO of "Jiaoge Yazi," was Baidu Tieba's first official administrator. He told QDaily that two things let early Baidu Tieba beat specialist forums and grow fast: the search engine sent it traffic, and a back-end system let bars be set up quickly.

More important, Tieba is a social product. A search engine is plainly a tool: it has a huge and loyal user base, but people don't linger.

After Tieba took off, Baidu also tried some more "pure" social products, such as the instant messenger "Baidu Hi," launched in 2008, and a microblogging service, "Baidu Shuoba," launched in 2010. But cut off from the traffic and users that search supplied, the first is now used only by Baidu employees to talk to each other, and the second was shut down less than a year after launch.

Social networks versus search engines

On a quarterly earnings call in 2011, Robin Li spelled out Baidu's social strategy for the first time: Baidu would fold social into search rather than build a separate social network. He also revealed that social search already accounted for a quarter of Baidu's total traffic and was still growing, faster than web search.

But search remains Baidu's core competitive strength, and paid-ranking ads are how it makes money now and will for a long time to come. Highly relevant search results give ads an obvious edge in value, but as the ways people get online splinter, that value has begun to slip.

For Google, at least, the trend is easy to see. In 2014, advertising made up 89% of the company's revenue, but growth slowed markedly in the second half. In the fourth quarter, the price of ads on Google's own sites fell 8% from a year earlier, and on third-party partner sites it fell 11%. Advertisers were no longer willing to pay more for their keyword ads.

In future, social networks may gradually replace search engines. That is the view of Facebook CEO Mark Zuckerberg, and the company's soaring ad revenue has begun to bear it out. In 2014, Facebook earned US$9.3 billion from advertising, up 53% from the year before, and its mobile ad revenue grew a full 136%.

Because they draw on real personal information, real relationships and individual needs, ads on social sites such as Facebook, Twitter and Instagram get closer to customers and hit their targets more precisely.

Facebook has held up one advertiser as an example: a wedding videographer from Michigan who set his ads to be shown only to "women" aged "24-30" whose Facebook status read "engaged." Over 12 months he spent just US$600 on ads and took in US$40,000 in revenue.

Facebook's rapid growth pushed Google to launch its own social product, Google+, in 2011, and to bring back co-founder Larry Page, who had spent several years running projects at the Google X lab. Like Facebook, Google+ shows comments, photos and the latest posts shared by friends and contacts, and users can back a friend's post with a "+1" button, much like a "like."

To promote it, Google at one point forced new users of its products to sign up for G+. It also hoped users would share interesting search results, which would in turn help it improve its search algorithm.

But just as the company had already shown with a string of social products such as Wave, Buzz and Orkut, Google+ lost its shine in under three years, and related services, photos among them, are being split off one by one. Figures compiled by the blogger Kevin Anderson in January 2015 showed that of Google+'s 2.2 billion registered users, only 0.2%-0.3% had ever posted a status there the way people do on a social network.

On this score, Baidu Tieba is in far better shape than Google+. A platform of finely divided interest communities is a natural draw for brand advertisers who want precise targeting. Search, Baidu's main business, keeps sending traffic to Tieba, and Tieba does not seem to need to "beat" anyone, even though China still has nothing as dominant as Facebook, whether in user numbers, in ad revenue or in the pressure it puts on search quality.

Social networks versus making money

Baidu has always been relatively relaxed about profits from businesses other than search. Not until 2009, at a meeting with third-party partner sites, did Robin Li say Baidu should "dig for gold" in the far larger traffic of its non-search businesses: "Even if they can earn only 20% as much as search, there is still enormous room."

Among the first attempts to make money outside search were targeted ads and brand ads on Baidu Tieba. But Baidu's quarterly results for 2009 showed these contributed almost nothing to total revenue. As late as 2012, Xiang Hailong, then a Baidu vice president from a marketing background who oversaw search, said the brand marketing products meant to make money from Baidu Tieba had not yet begun formal trials with major clients.

After inviting companies that know the fan economy well, such as Xiaomi, to run trials, Baidu formally launched its business platform in July 2014. From then on, Tieba had two kinds of bars: grassroots "interest bars" run by users, with no badge, and "official bars" run by companies, marked with a blue "V." The first companies to sign up included not only big brands such as Tesla, NetEase Games and Tmall, but also China Telecom, Hubei Radio and Television, China CITIC Bank, and even the magazine Popular Cinema (《大众电影》).

For a verification fee of at least 60,000 yuan a year, a company gets full administrative control of the bar bearing its name, plus the chance to place all kinds of ads all over the page. The more features it adds, of course, the more it pays. Few companies run their bars themselves; the great majority hand the day-to-day work to local agencies, of which there are now more than 30 across the country.

For industry bars whose topics are broader and whose commercial interests are harder to pin down, such as "makeup," "cycling" and "phones," Baidu introduced a "Tieba partner" scheme in 2015. A company whose application succeeds can control all the promotion channels in an entire category, just as it would its own brand's bar.

"The most obvious change once companies move in is that the bars aren't as 'clean' as they used to be." Several Tieba users QDaily interviewed said Tieba had long kept a fairly simple, clear look, with plenty of white space on PC and an emphasis on text. But once companies changed the page backgrounds, put ads across the top and down both sides, and could even dress up their promotions as posts barely distinguishable from ordinary threads, the platform lost much of its appeal.

And a bar with commercial potential that its company has not yet claimed can be flooded overnight with ads from a rival that bought the ad space, until users can't find their way around. As one user put it: "It's not really that the game company wanted to advertise in our bar. What we saw at first were usually Baidu Tieba's own banners. But if the company didn't pay Baidu in time, the bar would fill up with competitors' ads. So in the end it paid up, and got moderator privileges thrown in."

Moderators versus a network centered on people

This problem, too, emerged after Tieba began selling itself. Moderators seemed to add value to Tieba's later commercial push, but they also limited the chance of building a social network around individual users. And a bar under a company's full control may, after a while, become just another corporate forum, even if the original moderators manage it on the company's behalf.

Then there is a second layer of conflict, between the company administrators who parachute in and the original moderators.

Every Baidu Tieba bar has a management team drawn mostly from its users: three head moderators and a number of deputy moderators they appoint. In some large bars, a moderator can spend nearly a full office workday each day on routine tasks such as banning users, deleting posts, marking featured posts, pinning posts and handling appeals.

When a company or an agency takes over a bar, it has to spend time talking with the existing management to ensure a smooth handover. Wei Liqian, who ran the company bars for both Xiaomi and LeEco, told QDaily that he finished one handover in a single day by patiently talking things through with the old management team over QQ and WeChat. But he had also heard of brands that fell into bitter fights with the previous management when they took over their bars.

"Tieba users who become moderators of a bar generally deal with all sorts of people every day, and they are very reasonable. As long as the company shows a little basic courtesy and respect, the original moderators won't push back that hard."

After launching official company bars last July, Tieba introduced Tieba partners this May, opening specialist bars such as the beauty bar to advertisers. As the Zhihu user "Wang Dongming" sees it, "Ads are just an eyesore. It isn't a big problem even if you don't block them, because they don't damage the way Tieba is organized; they're just ugly. It's like someone scribbling in the margins of a book. It looks bad, but if you want to read it, you can still finish the book. Bringing in company bars and now partners, though, changes the way Tieba is organized. Pages go missing, the content gets cut and altered, and it's no longer the book it used to be. The authors don't want to write a sequel anymore, and it becomes completely unreadable. Tieba partners, like company bars, damage the way Tieba is organized."

Mobile versus the old PC

The Baidu Tieba young people see on their phones looks less and less like the original BBS and more and more like a social app. Besides a layout and gestures redesigned for mobile, it has added features such as groups, chat, voice replies and location display.

But in the process, the weakness of Baidu's user ID system, a long-standing soft spot, has if anything grown more visible.

In Tieba's early days, to win traffic and attention fast, it was at one point open enough that people could post anonymously under an IP address without registering. Only as Baidu launched more and more personal services did it adopt a fairly formal ID registration system, and even then enforcement stayed loose: a Tieba user could often keep several, even dozens, of spare alternate IDs.

On mobile, however, precise marketing is hard unless an account system links data across products. Baidu Tieba's phone apps now require users to register an ID with an email address, and offer rewards such as "check in every day to earn triple experience points."

This has angered some longtime users who have been on Baidu Tieba since the PC era: "My level on Tieba came from the time I spent reading and replying to posts, and I don't care much about it. Clocking in at work is tiring enough, and now people even pay to check in every day. It all feels like cheap tricks to tie users down."

Recreating its PC-era glory wholesale on mobile has taken Baidu, which began championing the goal in 2012, rather a long time, and it has not been easy. In the third quarter of 2014, Baidu's mobile traffic overtook its PC traffic for the first time. In the first quarter of 2015, mobile's share of revenue also passed 50% for the first time, though this pushed up the company's traffic acquisition costs (TAC) slightly.

Of the 14 apps with more than 100 million users that Baidu cites publicly, the ones carrying its mobile traffic and revenue are still the company's old strengths: search, maps, app distribution and online video. Baidu Tieba, which started its move to mobile relatively late, is not among them.

Whether they are the post-1980s and post-1990s generations who first found Tieba on a computer at school, or the post-2000s generation who came to know it by reading posts on their phones, Tieba's users share some traits: they are self-confident, think for themselves and keep a close eye on new things and trends.

The fans in Baidu's "Lu Han Bar" took only 2 hours and 29 minutes to stack up 1 million replies in one thread, setting Tieba's record for the fastest replies. At heart they are much like the young people who cheered wildly on Tieba during the 2005 "Super Girl" final.

"Join Tieba, find your crowd." It is indeed a catchier gimmick, but it also exposes Tieba's basic limitation. We think a social network should have these qualities:

• It suits mobile, always online and fully part of everyday life.

• One ID is one person, and that person is the node and the center.

• It is organized around individuals. True decentralization means every person becomes a center.

• It is close to real-name, not in the sense of government registration but in the sense of a real personality.

But Baidu does not seem to think this way. What it talks about most is still "monetization": a Tieba that started selling itself very early, and a social strategy that bundles social with search. How much can Tieba contribute to what Robin Li now wants, for "Baidu to shift from connecting people with information to connecting people with services"? We can't yet tell. What we can say is that connecting people with services still requires connecting people with people.

(QDaily reporter Wang Jiefu also contributed to this article.)

About Top 15

In every era, a few companies gather the most capital and the best talent and shape the development of society as a whole. QDaily believes the companies that shape how we live are the core of today's business world.

So we have chosen the 15 most important lifestyle companies in the world and in China: Apple, Google, Baidu, Alibaba, Tencent, Amazon, Disney, McDonald's, Starbucks, Nike, Xiaomi, IKEA, Uniqlo, Wanda and Vanke. The list will change as the companies' businesses, markets and trends change. We will follow them, along with their competitors and partners.

Over the next month, we will publish a series of reports on how these companies are developing.