Five Years After Double Reduction: Chinese Online Education Firms Recover Revenue, Pivot to AI, but Market Cap Still Lags
English summary
Five years after China's 'double reduction' policy, major online education companies have rebounded in revenue and turned profitable, but market capitalizations remain far below pre-crisis peaks. New Oriental and NetEase Youdao surpassed pre-policy revenue, while Gaotu narrowed losses and TAL Education lags in revenue but achieved over $500 million net profit. The industry has diversified into quality education and vocational training, though quality education faces limited pricing power and vocational training is a highly competitive red ocean. AI-integrated education has become the strategic focus, with companies developing proprietary large models, AI tutoring tools, and smart learning devices; however, monetization through hardware sales and subscriptions remains challenging. The next phase of competition centers on building AI infrastructure and data moats.
Chinese summary
双减政策出台五年后，头部在线教育公司营收企稳并转为盈利，但市值远未回到历史高位：新东方和有道营收超过双减前，高途亏损大幅收窄，好未来营收仍有差距但净利润超5亿美元。行业转向素质教育和职业教育两大方向，素质教育承接了K12流量但因非刚需存在天花板，职业教育则是竞争激烈的红海。AI+教育成为主线，各机构纷纷自研教育大模型、落地智能硬件和AI辅导，但硬件收入下滑、用户订阅付费意愿低，商业模式待验证。未来竞争将聚焦于AI基础设施积累和数据壁垒构建。
Key points
Revenue recovery but market cap lags: New Oriental and NetEase Youdao revenue surpassed pre-policy levels; Gaotu narrowed losses; TAL Education’s revenue still below pre-policy but net profit exceeded US$500 million; all four companies’ market caps are far from peak.
营收修复、市值难回：新东方和有道营收超双减前，高途亏损收窄，好未来营收差距大但净利超5亿美元，四家公司市值均远低于高点。
Transformation divergence: Quality education absorbed K12 traffic but has a ceiling; vocational training is a red ocean; New Oriental and Gaotu focus on vocational, TAL on quality education, Yuanfudao/Zuoyebang shifted to AI hardware.
转型路径分化：素质教育承接K12流量但存在天花板，职业教育是红海；新东方和高途重点布局职教，好未来深耕素质教育，猿辅导/作业帮转向智能硬件。
AI+education becomes main theme: Companies are developing proprietary education large models and smart hardware, but hardware revenue declining and subscription willingness low; monetization models remain unproven.
AI+教育成主线：各机构自研教育大模型、落地智能硬件，但硬件收入下跌、用户订阅付费意愿有限，商业模式仍待验证。
Future competition will center on AI infrastructure and data moats; firms offering full-scenario AI solutions will build true competitive advantage.
未来竞争聚焦AI基建和数据壁垒，能提供全场景解决方案的企业将建立真正的护城河。