How to Do Good AND Make Money

Carnegie Mellon math professor Po-Shen Loh set out to prove that social entrepreneurship can achieve economic self-sustainability without relying indefinitely on grants or venture capital. By pairing mathematically brilliant high schoolers with professional acting coaches and building custom moderation software for 30-student live streams, Loh created an online math experience that makes group problem-solving more engaging than 1-on-1 tutoring. His platform operates with over 50% profit margins at $23/hour per student, fully funding free STEM education for underprivileged kids in rural America and developing nations. In this interview, Loh shares how to design positive-sum network systems where 2+2=5, why unit economics are essential for doing good, and how AI enables solo creators to build impactful ventures.

Hi, I'm Po-Shen Loh. I'm a mathematician, but more than that I'm a social entrepreneur. What that means is that I'm interested in trying to solve real problems for real people, and that's my main driving force for what I do. But in order to do that, you need to have money, so we use the method of entrepreneurship to have enough money to try to solve social causes. It's not quite a nonprofit. We don't want to only do things that will cost us money and require donations for every single person we help, because if we lose value on every person we help, we are then completely dependent on philanthropy. Instead, the question is how to create an engine so that we can make money on each person we help so that we can make more help.

Don't start by thinking of something super grand like making a new foundation AI model. You can start much, much simpler. It's a totally different mindset. Of course there are things that you're passionate about, because you're watching this from a social perspective. That means that you see that there are people who need help. What are they paying for to help? Give them that help.

I've run three different ventures for about 10 years. I came to really appreciate social entrepreneurship as figuring out how to create value in the world, solving other people's problems in a way where everybody who's involved gets some gain in the entire system. It's important that we also keep some of the value. If you need money for every person you help, you are completely dependent on philanthropy, and you need to ask for more and more and more money.

My eventual observation was that for social entrepreneurship, if what you're doing really does on average make the world a better place, then you're actually creating value out of nothing. That's true. That's actually possible. Sometimes the whole is bigger than the sum of the parts. Two plus two equals five.

Sometimes that's a good thing, so I go and find the situations where two plus two equals five. I try to make sure that everyone who's involved in the whole thing, employees, partners, customers, you name it, gets a net gain of value. A customer can get a net gain of value even if they pay you, because they feel like they paid you for something and got more than what they paid out of it. Once I figure out this whole system, then on the net we actually gain money per person we help. Well, actually, so far that's just normal entrepreneurship.

What makes it social is that I happen to have control of the company. If the control of the company is in the hands of people who just want money, you have to just make sure you make a lot of money. But with what we did, we managed to reach profitability early enough that I actually still have control of the company. So the kinds of decisions that we make are not purely designed to maximize profit. The whole point of entrepreneurship, I guess, is to scale stuff.

That allows us to have a very rich experience. At that stage it cannot just be that I do everything myself, so I started to think about the ways to do this. There are people who try to do it through technology, maybe recording videos of themselves and then letting everyone see those videos. But I guess because I'm a network theorist, I like to think of ways that I can use my abilities to construct a network, or to cause a network to come together so that the network itself has value.

I'm always looking for where two plus two equals five. I want to find all of these, and one such place is when you bring people together. A very simple example of why is that you could imagine there's a person over here who has too much of one thing and not enough of a second thing. Then you could imagine there's another person who has not enough of the first thing and too much of the second thing. They're opposite.

So if you can find these two people and introduce them to each other, they will thank you, because they will just go switch. The one that has too much of one will give to the one who doesn't have enough, and the other way around. In fact, they might even pay for it, and they're thankful to you. You've created value. Two plus two is now five.

That's the value of a network in the simplest way, for one link. It can also get more complicated, with all of these different people whose needs and wants balance out so that everyone gets a net gain. You can create a lot of value by creating a network instead. That's how I invented the live product.

The live product is where you have a win-win situation. The experience we deliver is two high schoolers who are kind and really clever and really good at math. We incentivize them to teach by hiring professional actors who supervise them and give them real-time feedback while they live stream math lessons to 11, 12 and 13 year olds. This exchange is very beneficial for the older kid, because there are a lot of people in the world who are very good at math for whom learning communication skills would be good. It helps them in their entire life, so they want to join. We find professional actors who are looking for part-time paid remote work, and we give them a great deal. They then guarantee the class is good.

We also put two teaching assistants in every classroom. The teaching assistants usually are not from the US, for the only reason that our classes are US evenings and I had a hard time hiring people who wanted to work their main job every day at night. But then I realized that in Malaysia and in the Philippines it is perfect daylight hours when ours is evening, so we can make another win-win situation.

As a net result, we make all of these classes which can teach the middle school kids, and the classes that we make end up being really fun. They're taught by people who are a little bit older. The people are coached by actors, so they're funny. The people are super smart anyway. Suddenly we're creating a product where everyone wins.

Expii is the big organizing company for all of them. It's actually a for-profit company, and underneath it we can have all kinds of different things. We do have an Expii.com website, we also have the Novid app, and we also have the live classes. The live happens to make enough money that it also pays for Expii at this point, and Novid we also pay for through the live. So these two things, which are the do-good stuff, we are actually just paying for ourselves.

People are curious how in the world this supports all of these different social ventures. The live classes are designed to be approximately $23 per hour for each student who's taking the class. We set that by looking at the entire market and seeing what people are willing to pay for extracurricular math classes.

Honestly, we could probably have charged way more. We could probably have bumped that price way higher, to 30 or 40, because the people who have been teaching are extraordinary. Out of these high school students who have taught these classes, 30 of them ended up going to MIT. Six of them ended up going to Harvard. I don't think that is necessarily what makes you strong, but I mention it because many people who go for the tutoring ads will say, get tutored by a Harvard person, and then the price goes nuts.

So what I'm explaining is that the kind of service we're delivering should be very expensive. Then why in the world do we only charge 23? It's because we can. It's because we have another breakthrough. We made a 30 person class more fun than a one-person class. Not just more fun, more effective.

This is where, when you're trying to make a breakthrough, you do need to have innovation in a startup. Startups cannot usually just go and say that they will do what everyone else is doing. No, no, no. There has to be some innovation.

Before us, everyone thought that if you do an online math class, math is boring, and you cannot have more than five kids in the room because no one is going to pay attention. With five kids, you stare at them. Like with a Zoom call, if there are five kids and one teacher, you can see everyone, and if they're not paying attention you say, "Jimmy, what do you think?" That is how everyone thought an online math class has to be, because they assumed it's boring.

They also assumed that the way a math class worked was that you tell the kids what to do and you walk them through the steps. We fundamentally changed both of these philosophies. The first is that our math class is not trying to teach you how to do a standard method. Our math class is trying to teach you how to generate your own idea if you've never seen something before. So we constantly ask the students, here's a problem that you've never seen anything like before, what do you think you should do?

We've actually observed that if there are only five kids, brainstorming is hard. They just very quietly sit and stare at each other, because no one has any ideas. You need enough people. I've just proven why five is worse than more.

But the second thing is that I've actually seen examples of online interactions where more is better. Live streams are one. If there are only five people watching the live stream, it feels dead, and that's because of the chat. What I discovered was that in a live stream format, which is what all the kids are used to today, they like to be in something where the chat is flowing with lots of nice, fun, happy and productive contributions. That happens when you have 30. If you have five, you don't have that.

If you do 30, though, you have to be careful, because there could be nonsense. With Zoom, the problem is that if anyone put a bad message, it was there forever. There was no way to delete a bad message.

So I was analyzing this and realizing, aha, because those bad messages stay forever, that's why the teachers shut off the kids chatting, or they told the kids to only say the answer, because they were afraid a kid would say something bad. So I said that I need to build a little piece of software to put moderation in. On our system we actually use a different chat system, by which the students can suggest things and then we go and filter out every bad message and only let the good messages through.

Good could include the wrong answer. That's fine. Being bold and trying to suggest the wrong answer is fine. But saying that was dumb, nope, that gets cut out. Once we built this system, it became more like the online live stream vibe. Suddenly 30 kids is better than five kids, and now 30 times $23 an hour is good.

You see, the point is that there's a unit economics. The way I think about money is that I think about the unit economics, which means that for every unit of product we deliver, how much did it cost to create and how much do we earn from doing it? If you were selling at $23 an hour and you have five people inside paying you, you will only make $115 an hour, and you have to share that among all the different people who are doing this thing.

We can take 23 times 30, and 23 times 30 is much bigger. That's $690 an hour. That's close to $700 an hour. So our situation is that our unit economics have about $700 an hour going in, and then we have to spend some money going out. The net is that we probably only spend about half of that.

The net result is that we end up getting profit margins over 50%. You have to take into account the fact that we have to train the high school students, and that's actually very expensive, because the high school students leave after two years. Oh my gosh, this is a terrible business method. I don't want to say model, because business model is a word. But this is a business method that most other people would never want to do, because your teaching staff has to renew every year. This is ridiculous.

There's a huge amount of effort that we spend teaching our high schoolers how to teach. The reason we want to do that is because I'm a social entrepreneur, and we see that as a very valuable output, that we're creating a network of kind and clever and very well spoken high schoolers who grow up and will change the world. We don't get money back from that, but that's a good we're providing for society. We will do it just because we think it's a good thing to do.

But this is why I also know, as an entrepreneur, that I don't expect any venture-backed startup to try to fight with us on this. It will be too painful for them. It's a lot of pain to keep working with all these high schoolers unless you intrinsically think there's a value in spewing these wonderful people out into the world, because they're not going to make any more money for you. You train them up to do great things and then you don't get any of the money, but we're willing to do it.

So the net result is that we end up getting profit margins over 50%, and that's how we can go and support all of the other things that we do. All the other things that we do are funded by the fact that we have an excess on this one thing. Why do we have an excess on this one thing? It's because we invented a better model.

When I started, I thought that every company that gets to lots and lots of users makes money. That's what I thought at the beginning, and that's actually wrong. Even the XP.com website at its height, we would get half a million people coming through every month, and we never made any money off of it. Ultimately, it's useful to always be thinking about whether there are ways that you can go all the way through to success.

It's also good to understand whether those people are coming through but only spending 15 seconds. We didn't see a good way to convert into revenue. We actually ran out of money.

So you see, the original Expii.com was funded initially by some investors, and at that point we thought maybe we could make an absolute ton of money in the usual Silicon Valley way. Then we realized we're not going to be able to grow that fast, and we also found out that what we were doing was of interest to philanthropy, because we were doing good work for people. So I thought that philanthropy is even better, because that's not with equity attached and you don't lose control of the company. Let's try doing that for a while. In fact, we actually did get a big amount of philanthropy, and I got very excited, because I thought we could do this with philanthropy now. Then at some point the philanthropy got not interested in funding anymore.

That's how I learned that nonprofit is dangerous, because if you ever stop getting funding, you're out. So unfortunately, I couldn't get new funding on the philanthropy side. But the reason we were able to keep going is because we accidentally found a way to make some money with a totally different product.

What's happening here is that there's an Expii.com free website to learn math, and we started that in 2014. In 2018, in December, a Chinese startup company found me and said, can you make some video recordings of your math class, people will buy it. I said, are you serious, you can't sell recorded math classes. I just didn't think it would work, but they thought it would work, so I said okay, fine, I'll try it.

I was on the way to Bangladesh anyway, trying to go and do some other social impact stuff there. So I stopped in Shanghai for three days, filmed something, and then they actually started making money. It was a small amount, but enough so that we didn't need to raise any more investment. This is actually how I know there's another way to make a company, which is through bootstrapping. Even though the original Expii.com didn't go anywhere, for these other things that we built we didn't use the original investor money. We actually used bootstrapped money earned from these math classes sold in China.

So I want to emphasize that it is okay to make money. Actually, I learned this from a philanthropist. It was a life-changing conversation, because I was asking him for money. By that time we also got donations for the kinds of work that we did, which was more on the nonprofit route. When I was talking to that philanthropist, he said that since I have a way of making money, I should just make money and then use the money to do good.

What he commented was that if you make the money, then you can spend it. What a concept. Once I learned that that was happening, that's when I said, okay, how do we make better and better and better versions, always using what we've already earned to make the next one? So we were very, very tight on money and very disciplined. If we earn money, we can use the money we've earned to try to make more money, and then more out of that, and more out of that.

Because we had all of that going on, what happened in January 2020 is that we had to let go of a lot of our team, because the original thing that we were doing was not making money growth and it was also not what the philanthropist was interested in anymore. That then just wouldn't work. Unfortunately, that one had to fold down. That was painful, because these are people that I enjoyed working with, and unfortunately that had to end.

If my 2014 self was watching this, about to start Expii, what I would tell that person is to make sure that whatever you come up with has a way of actually making money that you can see. Don't start with something that just gets big and then figure out how to make money later. Figure out how to make money first.

But I think the bigger lesson is this. Can you reach profitability without losing control of your company? That's probably the bigger point, because the main thing is that if you have lost control of your company and then you start doing stuff like giving away free stuff, you're going to have to deal with some investors asking you why exactly you are doing that. Even if it makes long-term financial sense, it could be that by doing that you'll make even more sales later, but maybe the investor is short-sighted. You never know, so the control becomes important.

So the important thing is that we have a thing that can make money. The live classes, interestingly enough, come in two kinds. One kind is the classes where the students pay for the classes, and that one we make money from. We also have classes going for free to rural kids, and we are about to open some classes going for free to Africa too. We use some of our money to give away the free classes, but there's also a lot of donations which come in from people who like what we're doing.

The way that works is that there's a separate 501c3, which is the American language for a nonprofit registered charity. So there's a separate nonprofit entity, and they are our fiscal sponsor. What that means is that when someone wants to donate to the causes that we're doing, they actually donate to that charity. That charity then contracts us to do the work. For example, we're giving away free classes, so donations go into the nonprofit, the nonprofit then pays us, and then we deliver the free classes.

The easiest way to understand this flow is that there are often situations where people will donate laptops to schools. The way that always works is that the money is donated to a nonprofit entity, and the nonprofit entity then buys laptops from Apple or from Google. Apple and Google are for-profit entities, but they just happen to make really good laptops that are well suited for this particular market. And it happens to be that the nonprofit made the decision that the best vendor to source these laptops from is Apple or Google or whatever.

Our situation is that we just make the best advanced math lessons, from a cost perspective and a quality perspective, and then the nonprofit entity chooses us to deliver those services. It is actually possible to get philanthropic dollars for work that you're doing if what you're doing is delivering it to groups that would otherwise not have bought it for some reason. That's because we have the discipline to make a product that people will actually buy while also keeping the price point low.

Interestingly enough, that makes it easier to make the case to the nonprofit that this is the best way to spend those dollars. If a nonprofit was buying laptops for a school, they wouldn't want to buy overpriced bad laptops. Now, what was the most compelling thing to the people who were donating for charity? It was that the thing works. Now here's an important detail. When we started, a lot of people thought the things we were doing were way too ambitious.

You're going to get a bunch of high school students who are good at math, and you're going to teach them how to be funny? And you're telling me that there will be people who will take these classes, when there are so many things that are hard to do in that space? At the beginning, people were saying it's not going to work, how are you going to build this thing, it's a dream.

Okay, that's the value of being a social entrepreneur. We make money. So we used the fact that we made money on a different product to bootstrap the creation of the first two classes, taught by two students each. We made money, and once we made money, we made more money. Then we made more money, and then we made enough money that we had extra money.

With the extra money, I started going to low-income neighborhoods to see with my own eyes, and wow, there are all these kids in these regions who actually want to learn math. They really do want to learn math, and now we know how to find them. After we spent a lot of our money doing it, we're able to go and build out the whole rural side. At this point, we're teaching over a hundred of these underprivileged kids, and they're having a great time. Once we got the first class of Montana kids, those kids were awesome, just really interested in learning. They wouldn't have had the opportunity otherwise.

They're very engaged, and they have very high potential as well. Now that we're already doing it, it's very easy to go and explain to other philanthropists that it works and we've got this. For the philanthropists, I found that the biggest thing that helped was to be able to show that it already works on some small scale.

Actually, these days, if I was doing it again, I wouldn't take a single investor. In fact, the whole thing I'm doing right now doesn't use the investors. The only reason we have investors is because I'm giving them back for what they supported me with a long time ago. If I didn't do that, we would have zero investors and we would still be able to do everything we did. And my first thought is that if you're trying to make something where you're getting investor backing, think very carefully about whether that is really the right thing to do.

Maybe the answer is no, unless you have investors who for some reason are themselves very interested in social things. So if you can see what I'm saying now, it becomes a question of whether you can find the right investor. It probably won't be a venture capital fund. A venture capital fund is a bunch of people who are responsible for taking care of other people's money and making it become bigger. But an individual investor who's already made it in life and wants to see some good stuff happen in the world is a different matter.

They might give to you out of this idea that some of it is also for good. I think those high schoolers who graduate out of our network, I expect some of them to get very, very wealthy because of the skill sets they have, and they'll be very wealthy people who care about people. They might actually be happy to do this kind of an investment. That's actually why I'm building this network.

If I had thought it would be hard, I don't know if I'd have dared to try. So I'm actually very happy that I thought it would be easy and it turned out to not be easy. But I also don't like giving up. I've never seriously considered quitting.

I think the reason I wouldn't ever quit is because I'm lucky enough that I'm cash flow positive. I want to emphasize that the earlier you can get revenue and get cash flow positive, the better, because then the only thing that might happen is that you go slower. If you're cash flow negative, you can just run out of money, and you can become cash flow negative by hiring more people.

What I mean by cash flow positive is that you have a unit in your business where that unit entirely can generate enough money to power itself, so to speak. Your company needs to have a piece where, if you lost everything except that piece, you would be profitable. With that, you can grow bigger and bigger and bigger, and then I don't ever need to give up. It just means I slow down, if I need to, down to that core, and then rebuild again.

If you're a social entrepreneur and you chose that you care about certain things, and if you have the power to not give up because you have a profitable core, then you would say, why in the world do I give up? I should just go and find a way to make this work. A lot of people give up because they don't like failing. I don't like succeeding too fast. This is actually how I operate.

I'll give an example of something that's totally different, but it's very easy to explain. I was at one point organizing a conference, so I was in charge of the conference committee and we were going to invite speakers. I had some people working with me, and we were going to have nine speakers. So I told the people who were working with me, "Make a list of 90 people. I want that when you reach out to speakers, your statistical success rate ended up being 10%."

If your success rate was too high, like 80%, that means you asked for people who are not famous enough. You see, you aimed too low. So if your success rate is 1%, you aimed too high. 10% is good. It just takes you 10 times as long, and at first the people were saying, what, 10 times, 10% success?

Honestly, the things I work on are like 1% success or 0.1% success. I don't want to attack something unless I think it's really unlikely. What that means is that if I get it, it's going to be huge. So I actually expect that I try something and it doesn't work.

I did this video interview with you guys the first time, the EO one that got to 3 million views, more than 3 million now. I had done a lot more video interviews with a lot more people, but I didn't mind that some of the ones I did got maybe only 1,000 views. That's because I expect it to be hard. And then finally we got one, and this one is successful.

But I don't look to always be getting this. Then I look to get even harder. So I want to go and try things that are even harder and even harder and even harder. Like the thing I'm trying to do now, which is to get an explosion of science and technology across rural America. That's hard, and no one's ever done that before, but I actually think we'll be able to pull it off.

When I first started thinking of this, that was one whole year ago, and I didn't have anything at that time. I was just thinking, let me try, I think there might be something here. So I think the important thing is to feel like you want to see that what you tried was too hard, because that's how you know you're trying hard enough things. If the things you try work too fast, you're not going to make a breakthrough. You're aiming too low.

I feel like I cheated myself if it was successful too early. It's like, yeah, of course it won't work, but I'm going to try. And you just try and try and try. One works, and bam, now it's big. That's a new story, and then you can make another big jump from there.

I think that in the age of AI, social entrepreneurship is even more important, because there will be some very, very successful companies. In the old days a very successful company would translate into wealth for normal people. When a company became very successful, it had to hire lots of people, and so all the people under it would gain some money out of it. In the age of AI, a very successful, economically successful company might be able to do it with very few people, because it can use AI for the rest.

So then what you'd end up seeing is that the capitalistic company did well, the investors did well and the owners did well, but it didn't trickle down too much to the rest of the people. You see, that's because capital doesn't have a soul. Capital is just money, and the money makes more money. Investors often are controlling other people's capital, and they're just making money into more money into more money.

So these days, what I'm trying to do is teach the entrepreneurial skills to people who really do care about the world, and hopefully help more of these people be able to create their own ventures where they still maintain control. I think that this particular time in history is one of the best times for young people to get involved in business, and I hope in social entrepreneurship, because these days if you have some idea you can actually use AI.

You used to need to get a lot of money invested in order to get moving. You used to have to hire a software engineer, or maybe not just one, a whole team of software engineers. Just this morning, I was using Anthropic's Claude in order to do a whole bunch of software engineering. It's really good.

You can use AI to help you write slogans. You can use AI to help you with marketing. Pretty soon, there will be AI agents which can help you do communications. So it turns out that today you might not need to get much investment in order to get something off the ground. So the real question then becomes whether you can come up with something that people would actually pay for.

Could you become profitable without losing control of your venture? I think the answer is yes. So the idea is not to think first about how you make a big complicated thing where you need to spend a lot of money first. Many people hear about the glorious startups in Silicon Valley. Those are glorious, but there are actually very mundane ways to do things too, where you don't start by having something huge. Then you can maintain control of your company.

Some people wonder if I'm altruistic, because I seem to want to make this world a better place. Actually, I'm just really selfish. It turns out that what makes me happy is making other people happy, and that's what I do for fun. Now, this is a funny thing. I actually do really enjoy doing something and seeing a whole room of smiling faces. It's just fun, and that's why I like teaching, why I like performing and why I like giving talks. So actually, I'm just a really selfish guy.

But you see, the reason I said this is because I think that the more people in the world who discover just how much fun it is to make other people's lives really bright, the better the world is going to be. I do happen to think that you'll feel really good when you go and make someone else happy. So I would like to introduce that to you. I'd like to show you that this is a thing, and then you might decide that you like to do this too. I'm one of these guys who's not going to go out there and say that you must change.

No, I'll tell you, I think you might like it. If you don't change, well, it's your life, and I'm not going to get in your way. But I want to make sure that all of these people that we can show this to, we give them a chance, so to speak. Maybe after seeing these things, you're thinking that you want to try this too.

So a realistic thing to do right away is to think about whether there are things that people are paying for that you can do better than what they're paying for. Start thinking about whether you can make a much better version of that, or rather whether you can make something that really answers their pain point, and how much it would cost you to make that. Don't start by thinking of something super grand like making a new foundation AI model. Yes, sure, if you do that, you'll also be a billionaire. Very good.

But you can start much, much simpler. There are lots of products in the world that people are paying for. It's a totally different mindset. And of course, there are things that you're passionate about, because you're watching this from a social perspective. That means that you see that there are people who need help. What are they paying for to help? Give them that help.

Interestingly, you might find out that it doesn't cost you a huge amount of money to create. And if it does, keep shopping around. You don't have to do the first idea that comes to your mind, but once you find something that can generate some cash, that will give you more freedom. Then you get this little core that makes some money, and then you start to make the core bigger and bigger.

If there's one thing to take away and to remember, I think there's a huge value in people who really care about doing something bigger than themselves. Find those people. Our network is not meant to be exclusive, and we are not the only ones who know how to find us. As you go through the world, you can tell by looking at people's eyes whether some people really, really do care about these things. Become friends with them.

The fact that you also are this kind of person will mean that someday, someday, you guys might work together, or they might introduce you to someone else, and this will give you more and more and more ideas. The network is perhaps the most important part of your net worth these days.