Building credit worldwide with Juvo
Welcome back to What’s NEXT, a podcast from Samsung NEXT exploring the future of technology. In this episode, I speak with Juvo’s Head of Data Science, Aristotle Socrates, about how his company can help the billions of mobile users in the developing world gain access to credit — and why he believes “data science” is the wrong name for his field of work.
Ryan Lawler: Welcomes the the What’s NEXT podcast. Let’s start with your name. How did you get that name?
Aristotle Socrates: Okay, you said we only had 45 minutes? It’s a long story, but it’s a pretty good story actually. My parents are from India, South India. They both grew up in villages in the state of Tamil Nadu, and the story goes … So, it comes from my paternal grandfather. He was a farmer, and his father died when he was six years old. And so, it was just him and his mother, and they were getting by. And out of nowhere, when he was around 20 years old, he decided that he wanted to educate himself. So, he sold two out of four of the cows that they had against his mother’s wishes. And took that money to go to a town pretty far away.
And after that, he came back and then he ended up going to medical school. Ended up getting his bachelors and ended up becoming a school teacher, and resided in that village his whole life. It was impossible to get him out. I’ve only seen him in that village. So, in order to distinguish himself from his fellow villagers, he named his first son Socrates, and that’s my father. And so, then his second son, he named him Aristotle, and then his third son he named him Archimedes. And so, that’s sort of the tradition of my family. My grandfather is an extremely, extremely brilliant man. I’ve worked with a lot of smart people, and he is a really, really smart guy.
So therefore, my name became Aristotle Socrates. So, I don’t think my parents fully understood the repercussions of that name. So, here I am. That’s probably the 12,000th time I’ve explained that.
Ryan Lawler: Wow. Wow. That seems like a lot to live up to though.
Aristotle Socrates: It is. It is. It’s a little bit of a weight to carry on your shoulders. But on met, on average, it’s more of a plus than a minus.
Ryan Lawler: Okay. That’s incredible. So, tell me a little bit about your background. You studied physics and astronomy, I believe?
Aristotle Socrates: Yeah. I have a PhD in physics, yeah.
Ryan Lawler: Okay.
Aristotle Socrates: And I was a theoretical astrophysicist for quite a while.
Ryan Lawler: So, how did you find yourself doing data sciences, find yourself with Juvo?
Aristotle Socrates: Yeah. It’s actually a pretty common story nowadays. There’s a lot of talented people in physics, astronomy, biochemistry, applied mathematics, mathematics, who are getting PhDs and postdocs, and get really great jobs out of their graduate programs. But the ability to carry on your passions, or the ability to conduct really original research that’s creative and fulfilling become less and less as time goes on in your career. It’s just getting kind of harder and harder to have a really fulfilling life in academia.
As a result, I thought, “Look, they’re in California. There’s this new field called data science.” And so, I figured out how to use a computer, and we moved out to California. And then, I got my first job in data science. And one thing led to another, I met the CEO, and he told me the story of what they’re doing and the core idea behind Juvo. And I thought it was the best idea I’d ever heard, at least in the bay area. And so from that, I was like, “Okay, I’ll take a chance.”
Ryan Lawler: All right. Well, let’s talk about that. Let’s talk about Juvo, and the mission, and what you’re doing.
Aristotle Socrates: Yeah. So, how I pose it to people is, and how I think about it is, if I were an alien and I were to come to the earth and see the situation that half of the planet is underbanked, which means that they don’t have access to elementary financial services such as credit or savings accounts, how would I fix that quickly? And then, if I were an alien and the laws of intergalactic or interstellar economics are the same as they are here on Earth-
Ryan Lawler: God, I hope not.
Aristotle Socrates: Or at least the basic laws of interstellar economics are the same as here on Earth, I would ask myself, “What are the necessary elements to have a fully functional economic system?” And then, you would think, okay, well if you go back to the foundational works in economics by people like Karl Marx, and Adam Smith, and David Ricardo, and John Stuart Mill, they’re looking at the same problem from little bit different perspectives, but they all agree more or less on things that are very important.
That for a fully functioning, or highly functioning economy, you need two elements. And the first element is capital, which is pretty obvious to understand why you need capital.
But the second element, which is a little trickier to understand is, you need information. And you need organized information. And the reason why you need organized information is so that each individual stakeholder in an economy can make a rational decision. And so, in other words, financial institutions need for example, organized information about the population in aggregate to make, just to construct the financial products to begin with. They also need organized information about each individual, to know whether or not to make a decision about that individual. To give the person a loan, or not to give a person a loan.
So, we got to get a hold of people really fast, and from what I was just talking about, we got to get a hold of data on these people really fast. And so, as a result, what we do is, we partner with mobile network operators in the developing world.
Ryan Lawler: So, what does that actually look like? How do you turn these mobile users into individuals with a financial identity?
Aristotle Socrates: So, say you’re in Jakarta, Indonesia, and you’re walking around and you come to a zero balance, you go to the kiosk associated with some mobile network operator. And you give them five bucks, and then they give you a scratchcard. And then you scratch off the scratch card, and there’s a pin code. You send an SMS with that pin code in it, and then your account gets charged up for three days, a week, two weeks, a month, et cetera. Okay?
Now, when you come back to a zero balance, it doesn’t matter if you’re in the middle of a packed bus, it doesn’t matter if you’re at a meeting, it doesn’t matter if you’re at home watching your sick kids, it doesn’t matter if there’s a typhoon outside, you have to go back and get another scratch card. And that’s a really crappy user experience. And so, as a result, people are off air five or six days out of the month.
And so, what we do from a physical point of view, to help resolve this problem on the road for making a foundation for the next, what we would consider the next financial system for three billion people is, from a physical point of view what we do is, we integrate our data systems with the billing systems of mobile network operators. So, when a user comes to a low balance, we have the ability to predict that. And then when we predict that, what we do is we send them an SMS that says, “Hi. We noticed you’re at a low balance. Why don’t you download the Juvo app, which is co-branded with the mobile network operator. Upon doing so, you qualify for an airtime loan.”
And so, that’s the entry point into finance. After they pay these loans back, we then give them bigger ones, and then bigger ones, and then bigger ones, to the point where we’re giving out about a month’s to six weeks worth of airtime for a typical person at a pop.
At the end of that trajectory, we then offer them access to additional financial services, which can cross the carrier society boundary that are provided to them by additional financial partners. Thereby, it’s more or less creating a mechanism for establishing credit at scale.
Ryan Lawler: Okay. So, first of all, when you talk about the underbanked, let’s talk about just that one piece in and of itself. Why do those people have trouble getting access to capital or credit?
Aristotle Socrates: Right. So, you’ve built your credit, I’ve built my credit, everybody builds their credit in the United States, more or less without thinking about it. And what basically is in place that we don’t really know, or really see, is a pretty extensive data infrastructure.
And so, the underbanked people in the world, it’s the viewpoint of Juvo that they’re living lives for the most part, very similarly to the people in the United States. They’re consuming services like utilities, cellular services, transportation, et cetera, et cetera, all very similarly to people in the United States. And the one thing that’s lacking is the data infrastructure to collect that transactional history, aggregate it, and then serve up that information to stake holders in a timely way.
Ryan Lawler: The next question is basically, what’s lacking in that data perspective? Is it just that those people are paying mostly in cash? Why are my transactions being aggregated in a way that gives someone the ability to create a credit score for me, but folks in these other markets don’t have the same capability?
Aristotle Socrates: Well, I mean, it’s a complicated, complicated question. The data infrastructure that is in place in the United States for aggregating data to collect payment history and transaction history in individuals, has been there in way or another for nearly 100 years.
And so, people 50, 60, 70 years ago were establishing their credit, and it was mostly all cash based. Most people didn’t have credit cards, but there was a concerted effort to collect transactional data in analog and aggregate that in analog, because financial institutions found it extremely lucrative to get a hold of that information.
So, it’s just in the developing world, there just hasn’t … The economies haven’t gotten to the point where financial institutions have come to the realization that it’s that lucrative to do that. So, what we’re doing at Juvo is realizing that we can side step or circumvent that by establishing people’s financial identities, by using a data infrastructure system that’s already there in place, cellular networks. It has absolute penetration.
Ryan Lawler: So, we talk about from the consumer standpoint, the benefit is they don’t go offline, they don’t lose their access to the communication infrastructure. What do your partners, these telecom operators, get out of this?
Aristotle Socrates: Right. As I said before, we’re solving a very important problem for them, which is that the user has no affinity, typically, to any of the mobile network operators if you don’t have any credit. Because when you’re a prepaid user, you can go to one mobile network operator, to another mobile network operator, the services are all going to be the same.
An easy way to understand that is, if you drive a car in the United States for example. Where was the last place you got gasoline? Who knows? I don’t know where the … Shell, Chevron, who knows? And that’s because the product is the same everywhere. The price is the same everywhere. The reason why the gas companies aren’t suffering from a pricing war against each other is because the oil companies have an agreement with each other to maintain a certain level of price.
So, the mobile network operators, what they’re getting out of it therefore is, a way of avoiding the pricing war with one another of trying to steal relatively low value customers. In turn, making their relatively low customers that they’re stealing from everybody else into high value customers. And so, they don’t have to reacquire somebody multiple times. They acquire them once, they become a Juvo user, and they just stay in that network. So, it makes their business a lot happier.
Ryan Lawler: All right. So, there are a lot of other companies out there that are working on this problem of helping the underbanked get access to credit. So, how is Juvo thinking about this differently, or what do you do differently?
Aristotle Socrates: Well, I think they’re doing a lot of great work. They have really talented people working there, and a lot of them are doing really, really well. And some of them are not just doing it in the developing world, they’re doing it in the United States. There’s many people in the United States that are underbanked or unbanked. Like, most … A significant fraction of Oakland, for example, I would consider a credit desert.
I think the difference in approach is that we are grafting a credit system on top of very typical everyday behavior. So, people’s consumption of mobile network operator services. Compared to the United States, that’s exactly how we build our credit in the United States. You go to college, you get a credit card, you get your first job, your parents co-sign the apartment, you get a car, you have a car loan for that car, and you’re not thinking, “Oh, you know what? When I’m 38 years old my credit rating’s going to be a 784, and I’m going to be able to get a house in Westchester.” Nobody thinks that, it just happens naturally.
From Juvo’s point of view, what were saying is people in the developing world are just like that. And the one thing that we can help provide in the developing world is not really a hand, a helping hand. Just basically, and infrastructure to collect their good behavior. Or, it’s really not even good behavior, it’s typical boring normal behavior. And typical, normal and boring, if you’re a financial institution, is good.
Ryan Lawler: Okay. That makes a lot of sense. One of the other things that you are able to do with this is sort of walk people up this ladder of progressive finance. Curious where that goes, and for those who have been using Juvo for a while, what does that enable in the long term?
Aristotle Socrates: Right. Great question. Just to give a little prologue to that answer, what we’re trying to do is establish credit for three billion people within a time span of 10 years. That is a tough problem. It’s a tougher problem then I’ve … I worked on some tough problems in astrophysics, it’s tougher than those problems. So, there’s probably a few ways to go about it, but how we are going about it is sort of step by step.
The initial plan is to create a reservoir within any mobile network operator’s walled garden of users, and it’s a significant number of users. A reservoir of users who have a transaction history, like say, people who have taken out 10, 15 airtime loans of significant value. So, when we are progressively stepping them up, we’re deliberately stepping them up to basically a month’s worth to six weeks worth of airtime. Which really distinguishes us from a lot of competitors in the mobile network operator space, because that means we’re taking on a lot of risk.
A lot of what my team is managing is risk, is how to deal with all that risk and make sure that we are in good standing on our balance sheet, et cetera.
Ryan Lawler: Okay. When you look at the way that these markets work, is the data or the model replicable across different markets and different carriers, or does one market have particularly unique characteristics that might fall outside of the model in another market?
Aristotle Socrates: Yeah. That’s a great question. That was the assumption to begin with when we were doing this, and it is part of some of the concern I had about whether or not Juvo was scalable. So, the first carrier we ever really launched the hardcore full platform on, where we’re doing everything and all the risk management, and the notifications, and this and that, was in the Caribbean with a carrier. And there is 14 countries that we did. And we had to roll this out in like six months in 14 countries.
Now, these are small countries. One of them is Montserrat, and I think there’s more people who live on my block than in Montserrat. But these countries have real differences from an economic point of view. So for example, Jamaica, the per capita GDP is four of five thousand dollars per year. Whereas, in the British Virgin Islands, the per capita GDP is more like $40,000 per year, if my memory serves me correctly.
And so, what we found is that people, as I mentioned before, people spend very similar amounts on airtime, when scaled to their income, or at least the per capita GDP, in all these countries. And it’s a linear relation with a constant in proportionality of about 1%.
And so, what they meant is that we could equate convenience … Well, what does convenience really mean? It’s really time, right? So, we looked at the data in all these different countries, and we’ve looked at it in many other countries as well. And people top up about once a week. That means that they’re doing that 50 times a year.
Ryan Lawler: All right. So when you say 50 top ups a year, what does that mean for Juvo, what does it mean for the product, and how does it help you reach users?
Aristotle Socrates: If any chore takes 15, 20, 30 minutes or whatever, and you’re doing that 50 times a year, well then you equate that into hours, that’s like 20 or 25 hours. So, you’re queuing in line the same amount of time that you spend working to earn the money to consume the service you’re queuing in line. And that’s universally true across all these countries.
So, what we did is, we therefore scaled our products and the timing of our notifications to per capita GDP. And what we found is when we did this, well the prediction was that people would descend the credit ladder at the same rate, if we scaled everything to per capita GDP. If, and it’s a very important if, if people in poorer countries are just as financially reliable with respect to their income, as people in wealthier countries.
When we actually did this, the uniformity between all these 14 countries was extremely striking. So, I think that is something extremely interesting that Juvo has discovered. And I think it’s something that really could have very significant philosophical impact almost.
Ryan Lawler: Yeah, that’s pretty crazy. I never would have imagined that that would be true, and I think it really changes a lot of assumptions that people in financial probably have about the developing world.
Aristotle Socrates: Yeah, yeah. So look, like I said, I’m the son of an Indian villager, so I know what it’s like there. I’ve been there a bunch, a lot of times. People there are really a lot like people here. Especially after you spend a little bit of time there, you kind of forget that you’re there. You know, the same kind of human interactions that are happening here are kind of happening there.
And it turns out, at least from Juvo’s point of … Or at least what the experiments that we have conducted, that from a financial point of view, if you give people opportunity independent of how wealthy they are, they take it and they act good. They’re good actors for the most part.
So, that’s the central bet of Juvo. If Juvo really wants to be the most transformative, or a very transformative company, then that bet has to be true. If it’s not true, then Juvo’s not going to be as successful as it could be.
Right. Well, what’s one controversial opinion that you have, that you feel really strongly about?
Aristotle Socrates: Well, I think within the realm of data science, and I’m not sure if it’s controversial but it’s not really spoken that often; but a lot of people who I respect, prominent people in the field of data science that I respect also kind of secretly hold this opinion which is, what makes data science interesting, and what makes it profitable, and what it does to accelerate the success of companies is not necessarily the advent of smarter algorithms. But it’s the realization that you can take data from different but relatively obvious sources, and convert them into stuff that you can insert into products that just add value where you wouldn’t have been able to do that before. Where the type of technologies that we’re using didn’t exist.
It’s pretty simple. And actually, my viewpoint … You know, I have a background in theoretical astrophysics, and I actually think data science from a sciencey point of view, is most like, at least in the technology sector, it’s most like observational astronomy. Where you’re just … Observers in astronomy who … I never actually tooken observation before. My dad got me a telescope in fourth grade and I kind of used it a little bit. I’m not the type of person who goes out in the cold at night and looks at space.
But the observers that I know, they’ll go fly to Hawaii, and go to the top of the mountain where there are these huge telescopes, and they’ll just look out into outer space, and just look at stuff, and just take data night after night, night after night, night after night. And what they do is, they take the data and then they clean it up, and they reduce it, and they publish it in papers. And they do that over, and over, and over.
And then when you read their papers over and over, one paper in itself is not going to tell you the … But once you read a lot of them together, you start to get a picture of what’s going on. And there’s no underlying theory per se, if you’re kind of just pushing the edge, looking at data for the first time and just exploring it without any preconceived notions about it.
And I think that’s where you can have the most impact in data science. That’s the approach I think at Juvo, where I discovered we’ve made the most impact within the organization from a technology and product point of view.
Ryan Lawler: Okay. I’m curious, when you think about data science, and data science as a profession, there’s this entire field that didn’t exist maybe a decade ago, or maybe it was called something else. And I’m curious why that is. Is it just that we have so much data now to analyze, that we need scientist for it, or what is it?
Aristotle Socrates: I actually think it’s a terrible name. I think data science is a terrible name. And that’s maybe my most controversial thought, is that data science is a terrible name, ’cause you can’t have science without data. It’s like saying logic math. It doesn’t … You can’t have math without logic. Or it’s like fact history. It doesn’t really make a lot of sense.
And so, I think maybe a better name for data science would be quantitative technology, because what it is, is on the East Coast at all the Goldman Sachs and the JPMorgans of the world, you got other Aristotle Socrateses, and they’re coding away, writing C++ code, doing quantitative finance. And so, they’re in the finance industry, and they’re being quantitative, and their using math, and their using computers. In a little bit of a different way, but it’s more or less the same thing.
And I don’t think you need to be a scientist. I definitely don’t think you need a PhD to do data science. I think what you need is an open mind. And I think you need to have a little bit of math skills, but you really need to have the ability to grow your math skills. It turns out, you can also grow into a really great and powerful scientist, because all really great and powerful scientists didn’t have PhDs at one point. PhD is just a piece of paper.
So, I think having that open mind, and just coming above a certain kind of technical threshold, having a little bit of experience of knowing how to visualize and plot data, and phrase questions quantitatively is something that you learn. Kind of, it’s on the job training. And if you have the right mentors … I had a lot of great mentors. So, if you have the right mentors, you’ll figure out how to do it just by kind of imitating them.
Ryan Lawler: Okay, great.