In 2001, a college student named Arnold Fu started a place on the internet where people could learn together. Hujiang spent its first five years operating without a commercial business. When it became a company in 2006, the founding team numbered eight and its startup capital was RMB80,000. The classroom came before the cash register.
- Free language tools and a learning community built the audience.
- Paid courses turned some of that attention into revenue.
- CCtalk extended the idea to independent online teachers.
- Rapid growth carried costs that a big user count could not settle.
By the time Hujiang sought a Hong Kong listing in 2018, it reported roughly 170 million registered users and around 2,000 courses. There was another pair of numbers worth reading together: RMB555 million in 2017 revenue and RMB537 million in net losses. A company could be crowded with learners and still have an expensive education ahead of it.
01 / The dictionary is the front door
Hujiang’s insight was to meet a learner before that learner wanted to buy a course. Someone needs a word translated. Someone else wants to practise Japanese vocabulary on a phone. Free information, Little D Dictionary and Happy Word Field create reasons to visit that are smaller, more frequent and less intimidating than enrolling in school.
The paid destination is Hujiang Online School. Its catalogue spans languages, overseas-study tests, postgraduate examinations and workplace skills. The customers include university students, examination candidates and adults fitting study around work. The problem is practical: how do you get organised instruction without arranging your life around a physical classroom?
Its mobile school makes that proposition tangible through notes, questions, progress tracking and peer interaction. The app even gives learning a virtual tree. A tree is a charmingly patient mascot for an impatient activity: you return, complete something, and watch growth accumulate. The interface gives an abstract ambition a visible shape.
and dictionaries
and community
and instruction
The distinction from a tutor marketplace is meaningful. Hujiang can supply the surrounding study environment as well as the lesson. Compared with New Oriental or Youdao, its historical identity rests particularly on the community and tools feeding its school. That is a distribution advantage, although it does not by itself establish better learning outcomes.
02 / One company, two classrooms
Hujiang also developed CCtalk, an interactive teaching platform launched in 2012. Here, independent teachers and educational businesses could bring their own courses. Hujiang was selling instruction in one part of the business and supplying infrastructure for other instructors in another. Those activities share software, but they demand different commercial arithmetic.
Proprietary courses require content, teaching and student support. A marketplace needs teachers, buyers and a reason for both to stay. Historically, Hujiang described CCtalk’s monetisation through transaction-related charges. Its own branded courses supplied the overwhelming majority of revenue during the period covered by its listing materials. Platform ambition was ahead of platform income.
The company explored other routes to learning, too. It invested US$3 million in language-teacher marketplace italki in 2016. In 2017 it launched Hitalk for adult spoken English, describing adaptive learning and nearly 1,000 real-world scenarios. The pitch addressed working professionals who needed usable conversation, rather than another pile of vocabulary they could recognise but never say.
03 / Growth has a delivery charge
The financing was substantial. Hujiang’s reported 2015 Series D was about US$157 million. Its 2018 listing application put the company’s expansion before public-market investors. The appeal was obvious: a large audience, a broad course catalogue and a business sitting between consumer software and education.
But the school still had to pay for expertise, technology, administration and selling courses. Contemporary reporting tied widening losses to those expenses. Revenue increased strongly in 2017; losses increased as well. The useful question was whether the next paying student would help carry the system, or require still more spending to attract and serve.
Two separate accounting measures, shown at the same scale. Historical figures from the listing period.
In March 2019, Hujiang acknowledged staffing adjustments while disputing dramatic layoff rumours. It pointed to efficiency and turbulent capital markets. Yicai also reported management salary cuts of 20 to 50 percent and changes to office arrangements. The public shift was concrete: expansion now had to accommodate cost reduction. There is no need to invent a private moment of revelation.
04 / The classroom beyond the customer
Hujiang’s rural education work supplies a different test of usefulness. Its Hu+ project, started in October 2015, connected schools to online lessons and teacher development. In an April 2018 essay, Fu reported reaching more than 3,000 rural schools and over one million students. Those were the founder’s reported programme figures, rather than a measure of paying demand.
He also identified a limit that software businesses sometimes overlook: access to an application was insufficient without accessible teaching resources and trained teachers. The project worked with educational organisations on professional development. A connection could bring an art lesson into a small school; people still had to make that lesson work.
“They teach as they learn.”Arnold Fu, describing rural teachers, 2018
05 / Copy the habit. Price the lesson.
Today, Hujiang’s school site still displays language and professional courses. Its corporate training offering combines content with digital learning platforms for employers and institutions. CCtalk continues to display courses, but its current site names Shanghai Xianli Network Technology as operator. The platform Hujiang developed should be distinguished from assumptions about present ownership.
For a learner, the practical approach is to start with a defined goal, inspect the course level and support, and try the material before committing. For a founder, the transferable idea is equally specific: build something people repeatedly need, then offer a paid next step closely related to that need.
That approach depends on useful free tools, relevant instruction and enough revenue to cover delivery. It becomes weaker when casual visitors need no course, when teaching costs outrun fees, or when a sprawling catalogue makes support expensive. Hujiang makes the question wonderfully concrete. A dictionary can open the door. The school still has to pay the teachers.