From September 16 to 21, six days, seven open letters. FENBI (02469.HK) held the public opinion spotlight in the civil service exam training industry with an almost uncontrolled frequency. From "Salute to Excellent Peers" to "Who Exactly Is Feeling Wronged?", from admitting "pixel-level imitation" of Chaoge Education's courses to publicly dismantling a competitor's cost structure, and then to self-revealing that "the courses didn't sell more, but refunds increased quite a bit." A listed company used the "mini-essay" approach to push itself onto trending topics and into an awkward position.
Six Days, Seven Letters: A "Self-Exposing" Public Opinion War
The starting point of this storm was FENBI's "pixel-level imitation" of Chaoge Education. On September 16, FENBI released "Salute to Excellent Peers," announcing a comprehensive benchmark against peers and a price cut. On September 17, FENBI admitted in "Acknowledgments" that its National Day question-brush camp had "pixel-level imitated Chaoge's seven-day National Day course"—Chaoge priced it at 2,180 yuan, while FENBI priced it at 1,380 yuan, with nearly identical course structures and question-brush arrangements. On the same day, FENBI issued "Who Is the Conscience of the Civil Service Exam Training Industry?" which publicly broke down the competitor's cost structure, calculating Chaoge's profit per class at around 200,000 yuan and questioning the claim that it "doesn't make money." On September 18, FENBI issued "Apology," admitting that "after riding the trend, not only did the courses fail to sell, but there were also quite a few refunds." But before the ink on the apology letter dried, FENBI on September 19 proposed a "100,000 yuan reward": if Chaoge teachers could score 85 points multiple times or 90 points once in FENBI's offline mock exams, they would receive a 100,000 yuan prize. On September 21, FENBI followed up with a long article titled "Who Exactly Is Feeling Wronged?", directly responding to Chaoge's claim that its boss was "wronged." The article's core argument: amid the dispute between the two sides, traffic and revenue across multiple institutions rose sharply, so Chaoge's "hard work" and "being cyber-bullied" narratives did not hold up.
Seven letters in six days—this frequency is extremely rare among listed companies. FENBI chose the "mini-essay" tactic over conventional PR strategy. The short-term effect of this public opinion battle was not entirely absent. The imitative 1,380 yuan course attracted over 1,200 registrations on its first day, and the 12.8 yuan series course sold more than 200,000 copies. But FENBI gave a more honest assessment in its September 18 apology: "After riding the trend, not only did the courses fail to sell, but there were also quite a few refunds." There is a clear gap between traffic heat and actual course conversion. The cost of the public opinion war is being quantified in the form of refunds.
Founder Already "Crashed" Once, the Foundation of Brand Trust Was Shaken Long Ago
To understand why FENBI chose such an extreme marketing tactic, one must rewind three months. On June 3, Zhang Xiaolong, founder and then-CEO of FENBI, was invited to give a career planning lecture at the School of Philosophy, Renmin University of China. He changed the topic to "AI stock trading," claiming he had invested 80 million yuan in tech stocks and netted 53 million yuan in a month. When the student response was flat, Zhang lost his temper, publicly scolding students for "deserving to be unemployed" and calling civil service exams "a way to coast through life." The recording spread rapidly across the internet. Numerous FENBI students called for boycotts of its courses and demanded refunds on social platforms. On July 8, FENBI announced Zhang's resignation from all positions, including executive director, CEO, and board chairman. A company whose core revenue depends on civil service exam training—its founder publicly denied the value of its own target audience in public, which was almost a self-negation of its own business model. Zhang's departure did not quell the public opinion backlash, and the crack in FENBI's brand trust had already formed.
Loss-Making Performance: Operational Anxiety Behind the Public Opinion Battle
Compounding the founder's reputational crisis is a substantial deterioration at the operational level. In the first half of 2026, FENBI posted revenue of 1.248 billion yuan, down 16.3% year-on-year, with a net loss of 184 million yuan compared to a net profit of 227 million yuan in the same period last year. This marked FENBI's first half-year loss since its listing in 2023. Adjusted net loss was 158 million yuan, versus an adjusted net profit of 271 million yuan in the prior-year period. Training service revenue fell 17.5% to 1.069 billion yuan, accounting for 85.6% of total revenue. Book sales and other income dropped 8.8% to 180 million yuan.
The deterioration in gross profit is even more severe. First-half gross profit fell 35.2% to 522 million yuan, with gross margin dropping 12.2 percentage points from 54.0% in the same period last year to 41.8%. FENBI attributed the margin decline mainly to severance costs from optimizing its employee structure—by the end of June, total employees numbered 5,963, down more than 1,000 from the end of 2025, with full-time lecturers reduced by 716. Meanwhile, sales and marketing expenses rose 18.1% year-on-year to 362 million yuan, and R&D expenses increased 22.8% to 132 million yuan. Revenue is shrinking, expenses are expanding, and operating profit swung from 219 million yuan in the prior-year period to an operating loss of 177 million yuan.
AI Is a Way Out, But Distant Water Cannot Quench a Near Fire
FENBI has not been idle in seeking new growth engines. In the first half of 2026, revenue from its AI question-brush system courses reached 42.2 million yuan, already exceeding the total of 38.1 million yuan generated in eight months after its launch in 2025. But AI question-brush courses account for only about 4% of training service revenue, and against the backdrop of a 17.5% decline in traditional business revenue, this scale is far from sufficient to serve as a hedge. More critically, FENBI is already increasing AI-related spending before its revenue scales. The increase in sales expenses includes resources for AI course commercialization, while the R&D growth is mainly directed at AI course development. Investment comes first, returns follow later, and the cash burn during this window period is real.
Conclusion
As of September 21, FENBI's share price stood at HK$0.315, with a market capitalization of about HK$705 million, down more than 80% this year. From the industry leader at its 2023 listing to a market cap below HK$800 million today, the collapse in FENBI's valuation reflects the market's dual doubts about the competitive landscape of the civil service exam training industry and the company's execution capability. The founder hurt users' feelings with his remarks, and the company eroded users' trust with its "mini-essays." When a brand whose core value is "helping students pass the exam" treats students as pawns in traffic games, the short-term heat from traffic can never compensate for the long-term damage from lost trust. AI might be the next story FENBI can tell, but before AI truly supports revenue, FENBI must first repair the brand foundation it has shaken with its own hands.
Comments