Po-Shen Loh, a Carnegie Mellon math professor and former US Math Olympiad coach, explains his philosophy of social entrepreneurship: create self-sustaining, profitable ventures that solve real problems, then use the surplus to fund social impact — maintaining control of the company so decisions aren’t driven solely by investor returns.
The Core Insight: Make Money to Do Good
If you lose money on every person you help, you become dependent on philanthropy; instead, build an engine that makes money on each person helped so you can scale the help.
Social entrepreneurship differs from traditional entrepreneurship only in who controls the company: when the founder retains control, they can choose to reinvest profits into social causes rather than maximize shareholder value.
Loh learned this from a philanthropist who told him: “Since you have a way of making money, just make money and then use the money to do good.”
The “2+2=5” Framework: Designing Win-Win Networks
Look for situations where the whole creates more value than the sum of parts — where every participant (employees, partners, customers) gets a net gain.
Simplest example: connect someone with too much of Thing A and not enough of Thing B with someone who has the opposite; both gain, and they may pay you for the introduction.
Networks amplify this: when many people’s needs and wants balance, everyone gets a net gain and the system creates value from structure alone.
The Live Product: A Concrete Win-Win System
Live online math classes for 11–13 year olds taught by pairs of exceptional high school students (kind, clever, strong communicators), supervised by professional actors who give real-time coaching, with two teaching assistants per class (often in Malaysia/Philippines where US evening = their daytime).
High schoolers gain communication skills and life-changing experience; actors get well-paid remote part-time work; TAs get daytime jobs; students get engaging, effective classes; the company makes money.
30 of these high school instructors went to MIT, 6 to Harvard — but Loh emphasizes the social value of spewing “kind, clever, well-spoken” young adults into the world, not the prestige metrics.
Unit Economics: Why 30 Students Beats 5
Market price for extracurricular math: ~$23/hour per student. At 5 students (industry norm), revenue = $115/hour — too little to cover costs after splitting among staff.
Loh’s breakthrough: a 30-student live class is more fun and effective than a 5-student class because kids today expect live-stream energy; a flowing chat with 30 engaged kids creates momentum that 5 cannot.
Built custom chat moderation software: students suggest answers (wrong answers welcome), but toxicity gets filtered instantly — solving the “bad messages stay forever” problem of Zoom.
Revenue at 30 students: $690/hour. Costs ~50% of revenue. Result: >50% profit margins funding all other social ventures.
Bootstrapping, Control, and the Nonprofit/For-Profit Structure
XP.com (free math website, 2014) got 500K monthly users but never made money; venture funding ran out; philanthropy funded it temporarily, then stopped — teaching Loh that nonprofit dependence is dangerous.
Breakthrough came accidentally: a Chinese startup paid for recorded math videos in 2018. Small revenue, but enough to bootstrap without investors.
Current structure: for-profit company (XP) runs Live classes profitably; separate 501(c)(3) nonprofit acts as fiscal sponsor — donations flow to nonprofit, which contracts XP to deliver free classes to rural/underserved kids (Montana, soon Africa).
Nonprofit chooses XP as vendor because it delivers the best quality per dollar — same logic as a nonprofit buying laptops from Apple.
Loh still has early investors only because he’s buying them out; if starting today, he would take zero outside investment.
Lessons from Failure: Profitability First, Control Always
Don’t start with “get big, monetize later.” Figure out unit economics first: can you see a clear path to profit on each unit sold?
Reach profitability before losing control. If investors control the board, they may block social-impact decisions even if those decisions create long-term value.
Cash-flow positive core = freedom. If your core product is profitable, you can slow down but never run out of money; you only quit if you choose to.
Try Everything, Quit Nothing: The 10% Success Rate Mindset
Aim for ~10% success rate when reaching out to partners/speakers/investors. If 80%, you’re aiming too low; if 1%, too high. 10% means you’re stretching.
Loh expects most things to fail (1% or 0.1% success on big bets). He doesn’t mind — he tries many things, learns, and the one that works becomes the new platform for the next jump.
Example: dozens of video interviews got 1,000 views; one got 3M+. The failures were the cost of the breakthrough.
Current “impossible” bet: explosion of science and technology across rural America. One year in, no visible result yet — but he’s confident because the pattern repeats.
The AI Era: Best Time Ever for Social Entrepreneurs
AI lowers startup costs dramatically: coding (Claude), marketing, slogans, soon communications agents. You no longer need to raise millions to hire engineering teams.
In the AI age, successful companies may employ very few people — wealth concentrates at the top. Social entrepreneurship becomes more important to distribute gains.
The real question: can you build something people pay for, reach profitability, and keep control? Answer is increasingly yes.
Start simple: find something people already pay for, do it better/cheaper, solve their pain point. Don’t start with “new foundation AI model.”
Network as Net Worth: Find Your People
Loh calls himself “selfish” — making others happy is what he enjoys. He wants more people to discover that joy.
Practical step: look for people whose eyes show they genuinely care about something bigger than themselves. Befriend them. Collaborate. Introduce each other to opportunities.
The network of mission-aligned people is the most important part of your net worth today.
