The late 1990s represented a nascent yet rapidly expanding era for the internet in China, a period characterized by a surging user base that grew exponentially from approximately 620,000 in 1997 to nearly 8.9 million by 1999. This rapid adoption was, however, significantly hindered by limited domestic technological infrastructure, including inadequate server capacity, nascent data center development, and significant international bandwidth constraints that made accessing international content slow and unreliable for the average user. Simultaneously, there was a substantial and largely unmet demand for localized digital services, particularly search engines capable of effectively handling the intricacies of the Chinese language and delivering culturally relevant results. Against this backdrop, the concept for what would become Baidu began to crystallize in the mind of Robin Li. Li, an individual whose professional trajectory had been deeply intertwined with the foundational technologies of internet search, brought a critical set of experiences and insights to this nascent market. Educated at Peking University, where he specialized in information management, and later at the State University of New York at Buffalo, where he earned a Master's degree in computer science with a focus on information retrieval, Li possessed a formidable academic background. This specialized knowledge laid the intellectual groundwork for his future contributions to the field of web search and his profound understanding of the challenges in organizing large, unstructured datasets.
His early career in the United States provided a crucial training ground within the burgeoning internet sector. While working at Infoseek, one of the pioneering search engine companies that was among the dominant players in the mid-1990s alongside AltaVista and Lycos, Li was instrumental in developing and refining search algorithms. Infoseek primarily utilized keyword-based indexing but was also exploring methods for improving result relevance. It was during this intellectually fertile period, specifically in 1996, that Li devised and patented Rankdex, a sophisticated site-scoring algorithm designed to evaluate the quality and relevance of web pages based on an analysis of their hyperlink structure. The patent application for this innovation, titled "Hyperlink analysis and retrieval system and method," was filed in February 1997, well before similar widely recognized advancements. This represented a significant theoretical leap in the efficacy of information retrieval, moving beyond simple keyword matching to infer authority and relevance from the network of connections between web pages. The core principle of Rankdex—that the importance of a page could be inferred not just from the number but also the quality and contextual relevance of links pointing to it—formed a bedrock upon which future, more intelligent search technologies would be built.
The intellectual property embodied in Rankdex provided Li with a unique competitive advantage and a clear vision for an improved search experience. However, the existing market in the U.S. was already becoming crowded and highly competitive, with established players like AltaVista, Lycos, and Excite vying for dominance, and new entrants such as Google (founded in 1998) rapidly gaining traction with its PageRank algorithm. Li perceived a more fertile ground for his ambitions in his home country. China, despite its burgeoning internet penetration which saw user numbers approach 9 million by the turn of the millennium, largely accessed via internet cafes, lacked a truly effective and culturally resonant search engine. The complexities of the Chinese language posed significant technical hurdles, including accurate word segmentation (as Chinese script does not use spaces between words), the variety of character encodings (e.g., GB2312, GBK, Big5), and the need for sophisticated semantic understanding to process nuanced queries and cultural idioms. Moreover, the preference for localized content and the nascent state of Chinese web infrastructure meant that dominant global search engines of the era often struggled with slow loading times due to distant servers, poor relevance for Chinese queries, and display errors, thereby presenting an undeniable market vacuum for a locally optimized solution.
In the summer of 1999, Li relocated to China, driven by the conviction that his expertise in information retrieval and the proprietary Rankdex technology he had developed offered a unique opportunity to address this critical domestic need. He established an initial base in Beijing, recognizing that success would require more than just a superior algorithm; it would demand a deep understanding of the Chinese user's search behaviors, content preferences, and specific linguistic challenges, coupled with a robust technical infrastructure localized within China to ensure rapid service delivery and data management, and also to navigate the emerging regulatory landscape concerning internet content. This vision resonated profoundly with Eric Xu, a fellow Chinese national who had also gained significant entrepreneurial and operational experience in Silicon Valley. Xu, who had previously held management roles at the genetic information company Stratagene, brought valuable expertise in establishing and scaling technology-driven businesses, managing complex projects, and navigating the venture capital landscape. Their shared understanding of both the profound technological possibilities presented by internet search and the distinctive market conditions in China formed the intellectual and strategic foundation of their core partnership.
Preliminary discussions between Li and Xu focused intensely on the architecture of a new search platform, one that would not only leverage Li’s patented Rankdex technology but also be meticulously optimized for the unique characteristics of the Chinese internet. This involved tackling challenges such as developing efficient web crawlers capable of navigating the vast and diverse Chinese websites, establishing large-scale indexing systems for billions of Chinese documents, and building query processing engines that could accurately segment Chinese phrases and understand nuanced search intent. Critically, their initial business model was not solely consumer-facing. Instead, it strategically envisioned providing search capabilities as a service (SaaS) to other prominent Chinese internet portals, which at the time served as the primary gateways for the vast majority of internet users. Key players included Sina.com, Sohu.com, and Netease.com, which aggregated news, email, and community features, attracting millions of daily users. This B2B model offered several strategic advantages: it provided a stable, albeit indirect, revenue stream; it allowed the nascent company to focus intensely on refining its core search technology and accumulate invaluable user query data through its partners; and it mitigated the immediate, substantial costs and complexities associated with direct consumer marketing, brand building, and extensive infrastructure development. This approach allowed Baidu to build its technological prowess and reputation behind the scenes before a full public launch.
Securing initial capital was the next critical step to materialize their ambitious vision. Through extensive networking and by leveraging their established Silicon Valley connections, Li and Xu successfully attracted seed funding from prominent venture capital firms. Key investors included IDG Technology Ventures, known for its early and strategic investments in the Chinese technology sector, and Draper Fisher Jurvetson (DFJ), a respected firm with a track record of backing innovative startups. These early investments, totaling approximately $1.2 million, were secured during the height of the dot-com boom, reflecting the robust investor confidence in the internet's future, even as the market showed signs of exuberance. This crucial capital provided the essential financial resources to establish the company's initial operations, including securing office space in the Beijing University Science Park, purchasing the necessary server hardware and networking equipment for data storage and processing, and recruiting an initial core team of engineers and technical staff, likely numbering around 10-20 individuals. This capital injection was instrumental in translating their intellectual property and market vision into a tangible corporate entity. The foundational elements were now firmly in place: a compelling technological innovation in Rankdex, a clearly identified and underserved market opportunity in China, a seasoned founding team with complementary skills, and crucial early investment. With these components strategically aligned, Baidu.com, Inc. (officially incorporated as 北京百度网讯科技有限公司, Beijing Baidu Netcom Science and Technology Co., Ltd.) was formally established in Beijing on January 1, 2000. This date signaled its official entry into the competitive and rapidly evolving Chinese internet landscape, poised to address the formidable challenge of organizing and making accessible the vast and growing repository of online Chinese information.