What Lei Jun’s Early Startup Failures Teach About R&D Management

In 1987 Lei Jun entered Wuhan University’s computer department, finished his credits in two years, and his Pascal programming assignment was selected as a textbook example.

After graduation he rode a broken bicycle, carrying dozens of floppy disks and thick programming books, and began developing encryption, antivirus, accounting, CAD, and Chinese system tools.

Together with classmate Wang Guoqian, he created the BITLOK encryption tool under the name "YellowRose" and later co‑developed the "Immunity 90" antivirus, selling dozens of copies for a few thousand yuan.

Lei also co‑authored the memory‑cleaning utility RI (RAMinit), which became a staple for programmers dealing with frequent crashes on early PCs.

Inspired by WPS, he reverse‑engineered a popular version and met its creator, which sparked his ambition to build software for every computer worldwide.

In July 1990 Lei, Wang, and two others founded the company Sunsir, hoping the primary colors would create a colorful new world; they split shares equally and sold their first order for a few thousand yuan, but their clone of the Han‑card technology was quickly stolen, leading to financial loss.

After university, Lei moved to Beijing, received authorization from Kingsoft founder Qiu Bojun, and established Kingsoft’s development department to support WPS Han‑card and plan a three‑to‑five‑year product roadmap, operating from a modest rented house.

He assembled a team of over twenty top programmers, many of whom stayed for three years, creating a strong sense of belonging.

Lei emphasized that loving programming makes the work feel like heaven, while disliking it turns it into hell, highlighting the intense dedication required.

He aimed to develop a flagship product named "Pangu," an office suite for Windows that would include word processing, spreadsheets, dictionaries, and business card management.

Recognizing the shift to Windows, Lei noted that DOS software was becoming obsolete, yet the transition required careful planning.

Despite the ambitious scope, the Pangu project suffered from over‑confidence, a fragmented product focus, and a lack of market experience, leading to its commercial failure after a costly launch in 1995.

Analyses pointed to four main reasons: dispersed effort without leveraging WPS’s brand, an inadequate name, incomplete features, and Kingsoft’s inexperience in sales.

Lei reflected on the failure as a painful but formative experience, acknowledging that the setback taught him humility and the importance of solid product quality and conscience.