Koolearn (K-12) \China

Koolearn was the K-12 online education subsidiary of New Oriental Education & Technology Group, China's largest private education company. Founded in 2005 and spun off as an independent entity, Koolearn operated a dual-revenue model combining live online classes with recorded course content for primary and secondary school students. The platform leveraged New Oriental's brand equity and teaching resources to deliver test prep, subject tutoring, and enrichment courses across China's massive K-12 market. The 'Why Now' was compelling: China's education-obsessed middle class was rapidly adopting smartphones and broadband, creating unprecedented demand for scalable, affordable tutoring that could reach second and third-tier cities. Tencent's $164M investment validated the thesis that online education could democratize access to elite instruction while achieving superior unit economics versus brick-and-mortar tutoring centers. Koolearn went public on the Hong Kong Stock Exchange in 2019, riding a wave of EdTech euphoria that saw competitors like Yuanfudao and Zuoyebang raise billions. The value proposition was clear: parents would pay premium prices for outcomes (gaokao scores, school admissions) delivered through a trusted brand with proven pedagogy, now accessible from any device.

SECTOR Communication Services
PRODUCT TYPE EdTech
TOTAL CASH BURNED $164.0M
FOUNDING YEAR 2005
END YEAR 2021

Discover the reason behind the shutdown and the market before & today

Failure Analysis

Failure Analysis

Koolearn's collapse was a black swan regulatory event, but the underlying fragility was years in the making. On July 24, 2021, China's State Council...

Expand
Market Analysis

Market Analysis

The global EdTech landscape today is bifurcated: China's for-profit K-12 market is dead, but the US, India, and emerging markets are growing at 15-20%...

Expand
Startup Learnings

Startup Learnings

Regulatory risk is not a tail risk in education, healthcare, or fintech—it is a core business risk that must be modeled into unit economics...

Expand
Market Potential

Market Potential

The global K-12 online tutoring TAM today is $180-220B, with the US market at $30B, India at $15B, and Southeast Asia at $8B. China's...

Expand
Difficulty

Difficulty

The technical rebuild is straightforward with modern tools—Vercel for frontend, AWS/Cloudflare for CDN, Agora.io or Daily.co for live video, Stripe/Paddle for payments, and GPT-4/Claude...

Expand
Scalability

Scalability

Koolearn's model had inherent scalability tensions. Recorded content and AI-driven practice tools have near-zero marginal cost and can scale virally—once built, serving 1,000 or...

Expand

Rebuild & monetization strategy: Resurrect the company

Pivot Concept

+

An AI-native, outcomes-based tutoring platform targeting high-stakes exam prep (SAT, ACT, AP, IB, A-Levels) and gifted student acceleration in the US and international markets. The core innovation is a hybrid model: AI tutors handle 70% of instruction (practice problems, instant feedback, personalized learning paths), while human coaches provide 30% high-touch support (motivation, complex problem-solving, college counseling). Revenue is outcomes-based—families pay a base subscription ($99/month) plus performance bonuses (up to $500) if the student hits score improvement targets. The platform is built on modern infrastructure (Next.js, Supabase, Vercel, Stripe) with AI powered by fine-tuned GPT-4 and Claude models trained on 10 years of exam data. The wedge is SAT/ACT prep, a $2B US market with measurable ROI and low regulatory risk. Growth loops: students share progress on social media (gamified leaderboards, score improvement stories), parents refer friends (dual-sided incentives), and the AI improves with every interaction (data moat). The long-term vision is to expand into gifted education (students needing acceleration beyond school curriculum), homeschooling support, and international markets (UK A-Levels, Indian JEE, Singapore O-Levels). The business is designed for profitability from month one—CAC under $150 (organic, referral-driven), LTV over $1,200 (12-month retention, upsells to college counseling), and gross margins above 70% (AI reduces instructor costs by 60% vs. traditional tutoring). The moat is the AI's adaptive learning engine, which gets smarter with scale, and the outcomes-based pricing model, which aligns incentives and builds trust. This is not Koolearn 2.0—it is a fundamentally different business built for the realities of 2025: AI-native, niche-focused, outcomes-driven, and designed to survive regulatory scrutiny.

Suggested Technologies

+
Next.js 14 with App Router for frontend (React Server Components, edge rendering)Supabase for backend (Postgres, real-time subscriptions, auth, storage)Vercel for hosting and edge functions (global CDN, instant deploys)OpenAI GPT-4 and Anthropic Claude fine-tuned on SAT/ACT/AP curriculum and past examsLangChain for AI orchestration (prompt chaining, memory, tool use)Stripe for payments and subscription management (outcomes-based billing logic)Daily.co or Agora.io for live video coaching sessions (WebRTC, low latency)Mixpanel for analytics (retention cohorts, funnel optimization, A/B testing)Resend for transactional email (progress reports, coach scheduling)Cloudflare for CDN and DDoS protectionGitHub Actions for CI/CDSentry for error trackingFigma for design system and prototyping

Execution Plan

+

Phase 1

+

Step 1 - AI Tutor MVP for SAT Math (Wedge): Build a narrow, deep product—an AI tutor that handles SAT Math practice problems with instant feedback and personalized learning paths. Use GPT-4 fine-tuned on 5,000 past SAT questions and student performance data. Launch as a free beta to 100 students via Reddit, College Confidential, and TikTok. Goal: prove the AI can deliver measurable score improvements (50+ point gains) in 8 weeks. Collect testimonials, iterate on UX, and validate that students will use the product daily (target 4+ sessions per week). Timeline: 8 weeks, $20K spend (API costs, design, initial marketing).

Phase 2

+

Step 2 - Outcomes-Based Pricing and Human Coaching Layer (Validation): Add a subscription model ($99/month base + $500 bonus if student improves 100+ points) and introduce human coaches for 2 weekly 30-minute sessions (motivation, strategy, complex problems). Recruit 10 part-time coaches (former teachers, college students) at $30/hour. Launch paid beta to 500 students via targeted Facebook/Instagram ads to parents of high school juniors. Goal: validate willingness to pay, achieve 60%+ 3-month retention, and prove unit economics (CAC under $150, LTV over $1,200). Expand to SAT Reading/Writing and ACT. Timeline: 12 weeks, $100K spend (ads, coach salaries, infrastructure).

Phase 3

+

Step 3 - Full-Stack Platform with Viral Loops (Growth): Build the complete product—AI tutors for SAT, ACT, and 5 AP subjects (Calc, Physics, Chem, Bio, US History), live coaching marketplace (students book sessions with top-rated coaches), gamified leaderboards (students compete on score improvements), and referral program (refer a friend, both get $50 credit). Launch social sharing features (students post score gains to Instagram/TikTok with branded templates). Scale to 10,000 students via organic growth (50% referral-driven), influencer partnerships (pay top education YouTubers $5K-10K for sponsored videos), and SEO (rank for 'SAT prep', 'ACT tutor', 'AP study guide'). Goal: achieve $100K MRR, 70%+ gross margins, and CAC payback under 6 months. Timeline: 6 months, $500K spend (engineering, marketing, coach network expansion).

Phase 4

+

Step 4 - Data Moat and International Expansion (Moat): Use 2 years of student performance data to build proprietary predictive models—the AI can forecast a student's final score after 10 practice sessions and recommend optimal study plans. This data moat makes the product defensible (competitors cannot replicate without years of data). Expand internationally: UK (A-Levels), India (JEE), Singapore (O-Levels), adapting the AI to local curricula. Launch B2B partnerships with schools and tutoring centers (white-label the platform, revenue share). Introduce premium tiers: college counseling ($2,000 package), gifted student acceleration (custom curriculum), and homeschool support ($199/month family plan). Goal: reach $10M ARR, expand to 50,000 students across 10 countries, and achieve profitability (20%+ net margins). Timeline: 18 months, $2M spend (international ops, B2B sales, advanced AI development).

Monetization Strategy

+
The revenue model is designed for alignment and scalability. Base subscription: $99/month for unlimited AI tutoring, 2 live coaching sessions per month, and access to all practice materials. Outcomes bonus: families pay an additional $500 if the student improves 100+ points on SAT (or equivalent on ACT/AP), capped at $1,000 per year. This aligns incentives—parents pay for results, not hours—and builds trust. Upsells: additional coaching sessions ($50 each), college counseling packages ($2,000 for essay review, application strategy, interview prep), and family plans ($199/month for unlimited students). B2B revenue: white-label partnerships with schools and tutoring centers (30% revenue share, $10K annual licensing fee). International expansion: localized pricing ($50/month in India, $150/month in UK) with outcomes bonuses adjusted for local exams. The model achieves 70%+ gross margins (AI reduces instructor costs by 60%), CAC under $150 (organic and referral-driven growth), and LTV over $1,200 (12-month average retention, 30% upsell rate). At scale (50,000 students), the business generates $60M annual revenue with $12M net profit (20% margins). The key insight: outcomes-based pricing is the moat—it differentiates from competitors (Chegg, Khan Academy, traditional tutors) and makes the product defensible against commoditization. Parents will pay premium prices for measurable results, and the AI's predictive accuracy improves with scale, creating a compounding data advantage.

Disclaimer: This entry is an AI-assisted summary and analysis derived from publicly available sources only (news, founder statements, funding data, etc.). It represents patterns, opinions, and interpretations for educational purposes—not verified facts, accusations, or professional advice. AI can contain errors or ‘hallucinations’; all content is human-reviewed but provided ‘as is’ with no warranties of accuracy, completeness, or reliability. We disclaim all liability for reliance on or use of this information. If you are a representative of this company and believe any information is inaccurate or wish to request a correction, please click the Disclaimer button to submit a request.