Baidu’s Transformation — From a Search Giant to Building a Mobility Empire

Shuai Chen
7 min readSep 21, 2021

--

Over the past 21 years, Baidu made its glory with the rise of the Internet in China, fell behind in the mobile boom, and is embracing the AI & IoT era with ambitious goals in commercializing autonomous driving technologies. It created miracles while had setbacks and faced lots of criticisms. This article walks through the 21 years of Baidu, a story about innovations, growth, dilemmas, survival, and perhaps nirvana.

In January 2000, Baidu was established in Beijing.

The name “Baidu” is two characters in Chinese coming out of a famous poem from the Song dynasty that described one’s persistence towards his dream.

2000 — 2005: a search giant’s birth to IPO

Robin Li, prior to founding Baidu, worked in the US for 7 years as software engineer on Internet search technologies. During the time he created RankDex, a site-scoring algorithm for search engine page ranking that was awarded a US patent.

With the strong technical background, Baidu quickly became the provider of search engine technology to major web portals in China like NetEase, Sina, Sohu, etc. in 6 months.

In August 2001, an independent search site “baidu.com” went live, which was the start of Baidu’s transition from B2B technology provider to B2C search service. Although investors were initially against the shift of business models, Li saw the growth limits serving web portals. Meanwhile, the initial business model he had in mind similar to “Inktomi” (acquired by Yahoo in 2001) turned out to be vulnerable on the US market.

(Baidu homepage in the early days)

Throughout 2002, Baidu executed a “Lightning Project” aimed at intensively optimizing its search quality to improve user experience and win traffic in 9 months from its competitors, one of which was Google. As a result, the site was getting tractions from more and more Chinese Internet users.

Following the success of “Lightning Project”, Baidu had more confidence with the B2C direction and launched some products linked to its homepage in the next couple of years. Just to name a few that made huge impact:

  • Baidu MP3, music search engine.
  • Baidu Tieba, an online community organized by topics of interest.
  • Baidu Zhidao, a Q&A platform.

As the products became widely adopted, Baidu formed a strong community of contents. The search engine was like a distributor and the content community was like an aggregator. Baidu seemed to have built an empire of its own, where Internet users of the time constantly came back or never really left. In August 2005, Baidu went IPO on NASDAQ as best performing overseas deal on the first day.

2006 — 2010: thriving and hitting a plateau

During this period, Baidu further expanded its empire by launching over 20 products covering nearly every aspect of life, which formed a “Baidu Family Bucket”.

Some generated a huge user base, becoming leaders in their verticals, such as:

  • Baidu Baike, an online encyclopedia.

But some lost to their competitors, for example:

  • Baidu Space that went live in July 2006 rivaled Qzone (a.k.a. “QQ Space” from Tencent)
  • Baidu Hi that went live in March 2008 rivaled QQ (from Tencent)
  • E-commerce platform and payment solutions launched in September 2009 (similar to Taobao and Alipay from Alibaba)

In 2008, Baidu faced several crises. It was first suspected for covering a brand scandal. Then Baidu got challenged for its interventions in search ranking for commercial interest. Towards the end of year, it got exposed for allowing false information spreading under its algorithms. All of this pushed Baidu to launch a new advertising system “Phoenix Nest” (a.k.a. Fengchao) that weighted a lot more factors for fairness and distinguished search results from paid promotions, etc.

In 2009, Baidu initiated a new concept “Box Computing” and launched the feature in its search engine. It was an innovation and a breakthrough in the Chinese Internet industry. However, while Baidu was building an ecosystem with hundreds of 3rd party developers under its “Box Computing” strategy, mobile had begun gaining momentum in China.

2011 — 2015: missed the mobile wave

Since the mobile app “WeChat” was launched in January 2011 and growing fast in the following years, there were no reports of major actions by Baidu. Some thought Baidu believed mobile was just a different screen and search would still be the entrance to everything. Some thought the giant had lost its way in the transition.

Until 2013, Baidu decided to double down on mobile by acquiring 91 Wireless, a leading app distribution platform of the time, for $1.9 billion. But the value of 91 got diminished after a year as smartphone OEMs like Huawei and Xiaomi started launching their own app stores. Meanwhile, apps with the most daily active users like WeChat served as leading access portals independent of app distribution.

Then Baidu became aggressive with online-to-offline (O2O) business. Among a wide range of service products it launched, two were identified as key to dominate the O2O market: Nuomi, a group-buying service platform it acquired in January 2014, and Baidu Waimai, a food delivery platform launched in May 2014. By the end of 2015, the O2O market had gone through a consolidation while Baidu didn’t come ahead in the competition.

Besides the failed attempts, Baidu did make accomplishments that didn’t create much buzz at that time but laid the foundation for the future. For example, it launched the mobile version of Baidu and Baidu Map (in 2011) as well as invested in a series of app developers (mainly in 2015) to form its mobile ecosystem.

Most importantly, Baidu started applying AI technologies, mainly voice and image recognitions, in search that improved accuracy dramatically. In 2013, it opened an AI research lab in Silicon Valley.

2016 — 2021: back on track

Baidu decided to drop O2O completely by selling off Baidu Waimai and transforming Baidu Nuomi into an advertising platform.

It worked on strengthening its mobile Internet capability by focusing on 3 products:

  • Baijiahao, launched in 2016, was a content creation platform available on web and mobile allowing creators to publish content, distribute and monetize within Baidu’s ecosystem.
  • Baidu Smart Program, launched in 2018, allowed small applets from 3rd party developers to run directly inside the Baidu search app.
  • Baidu implemented in-feed ads in its apps since 2016 and devoted as much efforts as in “Lightning Project” optimizing the product in 2017 and 2018.

While mobile was contributing more and more revenue, AI had become the long-term mission of the company.

In September 2016, the Baidu World Congress was mainly about the company’s achievement in AI, which was like an official announcement of the “All in AI” strategy. Obviously, autonomous driving is the ultimate application of AI technologies. Around the same time, Baidu received a permit from the California DMV to test autonomous driving technologies on public roads, the first among Chinese companies.

In January 2017, Dr. Qi Lu, a renowned executive in computer science, joined Baidu as president & COO. Lu kicked off by reorganizing the AI related teams (NLP, KG, IDL, Speech, Big Data, etc.) into AI Group (AIG) and created Intelligent Driving Group (IDG) dedicated to autonomous driving.

In April 2017, the brand new IDG group announced an “Apollo Project” that was about building an open, comprehensive and reliable software platform for industry partners to develop autonomous driving capabilities on top of it. Nowadays the Apollo 6.0 has aggregated a community of over 45,000 developers and 210 industry partners worldwide. (For details on the IDG group, check out my previous post Chinese “BATH” Weight in Auto.)

(Apollo 6.0 roadmap)
(Apollo 6.0 architecture)

In September 2019, China released a long-term oriented guiding policy for building the country’s strength in the transportation sector, where intelligent transportation and new energy vehicles were included. Baidu reacted quickly by adding a new team to focus on intelligent transportation in 3 months.

In April 2020, the efforts were unveiled as “ACE Transportation Engine”, a full-stack solution helping cities to build intelligent transportation systems. “ACE” represents autonomous driving, connected road, and efficient mobility and it can be illustrated as “1 (digital foundation)+2 (intelligent engine) +N (application ecosystem)”.

(Baidu ACE transportation Engine)

In July 2021, Baidu announced the upgrade to ACE 2.0 that added mobility-as-a-service as an additional intelligent engine. In a bit over a year since ACE launched, the company had won a series of projects to implement its solutions in over 20 cities in China.

Over the years, Baidu has made a list of investments in external companies in the automotive and mobility service ecosystem.

On a higher level, Baidu has almost everything for the future of mobility — from autonomous vehicle to smart road, from robotaxi to robobus and robotruck, from software capability to vehicle manufacturing.

Baidu lost its premier status among tech companies in the mobile era. Now it seems to have navigated a path to return. Is “mobility empire” the ultimate pursuit of Baidu?

--

--

Shuai Chen
Shuai Chen

Written by Shuai Chen

Bridging the West and China Innovations in ADAS & Autonomous Driving | B2B Business Development | Go-To-Market Strategies & Execution (schen583@gmail.com)

No responses yet