This wonderful interview is with Shivani Siroya. Shivani is the CEO and founder of Tala. Tala is changing how credit and loans are doled out in emerging markets like East Africa. To do this they’re using data from cell phones to better understand their customers. Tala can disburse loans within minutes.
Tala was recognized by FastCompany as the #1 most innovative company in money in 2015 and the #8 most innovative company in finance in 2016.
Tala seems like the future of credit both in developing countries and in developed countries like the United States. And Tala has built some incredible credit models to make this happen.
Here are some other things we talk about:
-How did you come up with the idea for Tala? Shivani was probably one of the most informed people on Earth about credit in developing countries before starting Tala.
-What did your first product look like compared to today?
-What inputs do you use for your credit models?
-What input surprised you that makes someone a better credit candidate?
-What’s a good case study where you helped someone?
-Why did you start in Kenya?
David Kruse: Hey everyone. Welcome to another episode of Flyover Labs and today we have a really interesting guest with us, Shivani Siroya. And Shivani is the CEO and Founder of Tala. Tala is company changing how credit loans are given out in emerging markets like East Africa. To do this using data from cell phones to better understand our customers and Tala can disburse loans within minutes, which is quite fascinating. And they are recently recognized by fast companies as the number one most innovative company and then Money in 2015 to have a number 8 most innovated company in finance in 2016. So I invited Shivani on the show to just hear more about her background and about Tala and it seems like what they are working on is the future of credits and lending. So something we should all hear about and learn about. So Shivani, thanks for coming on the show today.
Shivani Siroya: Thank you so much for having me.
David Kruse: Definitely. And so first off, you have an interesting background. Could you just give us a little overview on how you got to where you are now?
Shivani Siroya: Sure. In terms of my background and how I got there, there is all kinds of mix, investment banking and then also working for the UN Population Fund. So I started out in Banking and was doing pretty traditional equity research and realized that I wanted to do more and kind of work on something that was more thoughtful and meaningful and then I got microfinance, and had the opportunity to work with a small microfinance institution and realize that while I really felt the benefits of microcredit, I really felt like you know the micro borrowers that were in our system weren’t necessarily moving out of the informal financial services market into the formal financial system. So into a system where the rest of us are, where they have really choice and control over the financial products, that they have access to. So I got very interested in trying to understand why these problems existed, went to go study funds economies and econometrics and from there I ended up actually working at the UN population fund and you know studying the effect of micro credit on peoples quality of life and other development programs and whether these development programs or micro credit programs were actually improving that quality of life for these individuals. So if we could really understand where this capital went in the cinicom, we could then understand who to give more additional capital to and how to customize these products to better suite their needs and as I did that research, I ended up actually interviewing about 4,500 people and…
David Kruse: What?
Shivani Siroya: Yes, it’s real crazy, but what I realized was that was really the issue. It was one, this data really wasn’t digitized. So a lot of that information is obviously in cash and it’s in paper records and so I wanted to find a way to bring this data into you know real life and create really the foundation for financial systems and these credit scores. And then the second thing was I wanted to find a way to use that information to then be able to create customized products for these individuals and so that’s kind of that’s how Tala started.
David Kruse: Interesting. And so where did you do that research? You know what countries and how long did it take you? And did you do it personally, all those 4,500 interviews?
Shivani Siroya: Yes. I mean it was almost like being a walking clickbook. So yeah, I mean it was across nine different markets. It was across West Africa, Southern Africa and South Asia and it – I was there for about two and a half years.
David Kruse: Wow! At that point you were the most educated person on micro finance in the world or one of the most by the end of that, wow, okay. And so then how did you go from that to actually starting Tala. So you had some amazing research and you understood maybe who should be giving – who should you be giving these loans to. But then how did you take the next step and then start, actually start the company and build out some technology and yeah, how did that go? Shivani Siroya: Yeah. I mean I think it was a little bit – it was pretty tough. I think the first piece is I started collecting the information by hand, right and then from there it was realizing hey, all of these customers do have telephones in their hands, you know mobile phones in their hands. Can we use basic technology so text messaging and interactive voice technology in order to collect this information and so collecting both the revenue and expense data and demographic information from our customers. And then the next step, which is really what the product is now is realizing that Smartphones were starting to become more and more prevalent in these markets. You know I think right now there’s about 1.8 billion Smartphones in emerging markets and so it’s – I mean it’s growing very, very rapidly. And so what I started realizing is that all of this information, the revenue data, the expense data and the financial transactions, how much people are paying on bills, where they go throughout the day and kind of the behavioral information or the credit, kind of the likelihood of whether someone would be willing to pay back a loan, a lot of that data was sitting on the operating system of these phones and it was already digitized. So when we developed Tala, it was a software that’s embedded within an Android App that allows the customer to then give us access once they download our app, to give us access to that key – to those key metrics and information that’s sitting on the device and so then we can seamlessly bring that in. We bring in about 10,000 different data points per customer and then we’re able to score them in about 20 seconds.
David Kruse: Yeah, that’s amazing. And can you give us some examples of if it’s not confidential, some of the examples of the data points that you bring in and analyze?
Shivani Siroya: Sure. I mean I think a lot of what we’re looking for is get past to the information, so financial transaction data is very important to us, because it allows us to understand what can this individual actually afford and what are they looking to do with their capital, right. And so we see how much they pay on their electricity bills, their rent, their water. We see how much goes to school fees, how much comes in and out of their bank accounts or their saving balance, so that’s one key metric for us. Another thing is we’re looking at consistency. So we want to see are the bills that they are paying on time, are they delayed. Are they paying about the same amount for their cell phone bill every single month and it gives us a really good proxy to understand what their actual income is. The last thing is then more the behavioral stuff and so we want to understand what is this person’s support system? You know do they get that on social network, they have a distributed social network. We want to understand what kind of employment do they have? Is that consistent, do they go to work every day, do they move around in the same places and make the same phone calls, things like that and so that’s very much saying, ‘okay, we know you have the capacity because you make a certain amount of money. But then we want to understand what kind of character do you have and are you someone that is not going to be fraudulent and can verify that identity.
David Kruse: And you can get all that from somebody’s phone?
Shivani Siroya: Yeah.
David Kruse: Because I can understand like the social and the vocation, but I don’t think my water bill you’d get from my phone. I know it’s different in other countries and that’s why you’re targeting certain countries, but – yeah, so you can get all that from the phone?
Shivani Siroya: Yeah, it’s pretty amazing.
David Kruse: Wow, so that way they are much far ahead than here in the United States, interesting. All right, so you had – so did you start raising money right out of the gates, because you had this great research and you had all of this research on paper that could be translated into kind of creating models of ideal people to lend to, you know translating over the phone. So how – what did your initial kind of prototype look like and how is that comparative today?
Shivani Siroya: Yeah, so as I was mentioning, a lot of this data we initially collected just using that basic technology as text messaging and the interactive voice, and so we actually had customers self report in this information. And then as we started to learn about what kind of android technology was out there, we then shifted from doing the self reported to actually collecting it directly off of the device.
David Kruse: Interesting. Okay, got you. And what countries are you operating in now?
Shivani Siroya: We’re currently working in East Africa and now in South East Asia and so our main market in East Africa is Kenya and then we’re also in the Philippines.
David Kruse: Got you. And why did you start in Kenya?
Shivani Siroya: It’s just it’s a wonderful market and that there you’re seeing, it’s kind of this perfect convergence of technology and kind of the financial services system. And so in a population of 44 million people, you have about 27 million people that have a mobile money account. So that allowed us to really you know kind of seamlessly be able to automate our disbursements and collections within the Kenyan system. We didn’t have to rely on any other third parties or bank accounts. Then the other thing is that you know in terms of prevalence of kind of just micro credit and people understanding financial services, the Kenyan market is very sophisticated and you already have a lot of products like mobile savings accounts, you have group savings accounts and a lot of this is already happening on the phone. And you do have a very high prevalent android technology with an English speaking population and so just all of these things together really made it a great market for us to pilot in.
David Kruse: That makes sense. And so what you do over in Kenya, well could you even do that in the United States because of our lack of money transactions on the phone. Do you think it would even be possible?
Shivani Siroya: Yeah, it does make it difficult, but now with things like [inaudible] and AppleK you know I think the U.S. is starting to catch up. I think the difference here is actually the kind of data that’s on our devices. A lot of the receipts that what I am talking about don’t exist on our phones here, but they do exist in emerging markets.
David Kruse: That makes sense, okay. And then so how do you – how does it work updating kind of your scoring models. So I assume you kind of have some kind of a score that you might give each potential person you would lend to and then over time either those people pay back or they don’t and then do you feed that back into the model to continually improve it?
Shivani Siroya: Yes, exactly. And so you know you initially start out by giving planned lending in order to really get the outcome data and you do want to actually see people repay as well as default, because that’s how you start to build the correlation. But then yes, as we’ve gone on and we’ve now built more than four different scoring models and so it allows us just over time to continue refining that.
David Kruse: So I was kind of curious, is there – why isn’t there kind of a good story of how if somebody has X, there is no way we’ll lend to them, because they always default. Have you ever made a concept like that that’s kind of like a pilot you know that you didn’t expect?
Shivani Siroya: No, I mean it’s not – there is no hard and fast rule to it, right. I mean in that sense it’s like every feature has to be – I mean there is no one feature that is going to determine it. I mean I think the best thing I can say is you know when we can verify someone’s identity, that’s probably the only instance when it’s going to be difficult to lend to someone, right, because then you have to think about exception clause. But I mean I can give you definitely some fun ones. The ones that I think is always really interesting is you know the way someone organizes their contact book or their – you know within their phone book. And we found that customers that have more of their phone numbers listed as you know first and last name; so in this case my name would be Shivani Siroya. If more of my contacts, more than 40% of them are listed with two words as opposed to one, I am actually more likely to be a good borrower and so…
David Kruse: That’s fascinating.
Shivani Siroya: And I think that what’s really interesting about that one is that alone is not the feature that we use right. So we really like to think about the data as the story and the context of the data in terms of someone’s daily life, and so we’ll take a feature like that and we’ll interact it with other variables that we know about them. So it could relate to the kind of phone that they have and so if they actually have a low end device, even if it’s a Smartphone, it may be very cumbersome for someone to actually input first and last name. You know they may not be a native English speaker and so actually typing all of that out is going to be difficult for them. It may depend on their kind of business that they run. You know, if they are running their own business, we’re more likely to hopefully see that they are very organized and that they have many contracts. But I think it’s always – I think that’s the most interesting part about our work; it’s that it’s not just data points for us, but it is really the daily life of these customers and that’s how we actually customize the product to them.
David Kruse: Yeah, that’s interesting. How big is your data science team? It sounds like a very complex data science. I mean who would think to look at the contacts? I mean sure you guys have gone through the entire phone looking at what can you look at to start incorporating that and then waiting that. It seems like you have a lot of different data points to wait. Yeah, how big is your data science team?
Shivani Siroya: In total our engineering and data science team together is about 20 people.
David Kruse: Okay. Interesting, interesting. They must stay busy. And so yeah, I guess on that what – can you give us some other stats on Tala, whether it’s the customer is – well, I don’t know if your disclosing revenue. Its fine if you’re not and then money that you’ve raised so far.
Shivani Siroya: Sure. So inclusive of our team, in total we’re about 45 full time team members and so the majority of the team is based here in Los Angeles. We’re based in Santa Monica and that’s where our headquarters is. And then we also have local teams based in Nairobi and then also in the Philippines, in Manila and then what were the other questions? Sorry.
David Kruse: No problem, no problem. Well, revenue and money raised if you’d disclose revenue?
Shivani Siroya: Sure. So in terms of revenue we’ve now got I would say over $2 million in revenue and in terms of – and that’s in the last year I would say. Actually sorry, lets scratch that one. I’ll just say we’ve generated over $2 million in revenue and what was the other questions? Oh! And money raised.
David Kruse: Yeah.
Shivani Siroya: So in terms of money raised, we’ve raised more than $10 million.
David Kruse: And are you – I think I read that you are now lending directly. Before it was just more of your scoring models here and developing that. Are you lending now, your money to these people as well?
Shivani Siroya: We are, yeah. So we not only acquired the customer to underwrite the loan, but we also service and do our own direct lending.
David Kruse: Wow! That’s amazing. You only have 45 people and you’re doing that. And do you have to – in Kenya are there a lot of banking regulations you have to go through in order to start doing that?
Shivani Siroya: We don’t actually and that’s yeah, so we don’t actually take deposits and so because of that lender the banking regulations.
David Kruse: Oh good, okay. Well, that’s nice. Okay and I’m curious and I know we’re getting to kind of – well, we got a little time left, that’s good. I was curious if you have some good case studies of how you’ve helped folks over in the past year too with lending money?
Shivani Siroya: Yeah, I mean I think we have some amazing stories. I think the biggest thing that I would say is that getting home from me and this is just as I really said in my previous background is I think it’s the fact that customers never believed that the financial system was really for them, right. And so the fact that they can download an app, answer about any question and in under five minutes they can actually get money into their accounts. You know our customers call the product ‘Magic’ you know and then I think that’s the thing I love the most, that the fact that they really are – they are amazed that people trust them. I feel like that’s what we are trying to do is kind of change our perception of risk and really unlock this capital in this market and really make it so that these customers have that choice and that ability to really control their own destiny and go seek out these opportunities. And so in that respect, one of my favorite customers, her name is Jennifer and she is 65 years old and she went from – when she first started working with us they were just running a small food stand and then over the course of the last year and a half with us, she has not only added two additional food stands, but she’s now gone on to actually get a small business loan from a traditional financial institution, because she now has that credit history and she is actually working on opening up her own restaurant.
David Kruse: Wow!
Shivani Siroya: And so you know it’s kind of – it’s the fact that this woman in her whole life was not able to achieve her dream and now at the age of 65 years old, she is still doing it right. She is still passionate about what she does and I think she’s ones of the most courageous people I have ever met.
David Kruse: That must be nice going to work with stories like that in the back of your mind and helping people like that.
Shivani Siroya: Yeah, we have something called ‘Our Wall of Love’ and its basically just testimonials from our customers. And so as they are going through the process they actually leave us comments and so it’s completely unfiltered and we – you know it’s kind of nice to wake up to that every day.
David Kruse: Definitely. And so someone like Jennifer, what would she have done without you guys. Like she just wanted to chose just not be able to get any type of loan at all?
Shivani Siroya: No, I mean she could have gone to a microfinance institution but the capital would have been kind of – you know it wouldn’t have been very customized to her needs. So if you think about it, you know with Jennifer she was running that food stall and she needed capital in the morning to be able to go and buy her vegetables and her inventory. And so we are open 24 hours a day and so she was able to take a loan out, she was able to pay it back you know in three to four days and then take it out again. But it was all really dependant on what her needs were, as opposed to us paying, here is a loan for the next six months and this is the structure that we are giving you. So what we really did is provide her with that kind of flexibility to say you know, the term of your loan can be anywhere between two weeks up to 90 days and it’s really depended on how she is going to use that capital. So we are dynamically pricing that.
David Kruse: Wow! That’s interesting, okay. That’s like a super efficient line of credit or a super fast line of credit. Interesting!
Shivani Siroya: Yeah exactly.
David Kruse: And so what type of you know it’s probably a range, but what type of interest rate or fees do you charge if that’s a…
Shivani Siroya: So as I mentioned we dynamically price and so it goes anywhere between – I mean our average interest rate is about 11%.
David Kruse: Wow! Interesting, that’s really good. That’s lower than the credit card rates over here by a lot. Interesting, I would not have expect that, okay. And there is no – you never talked to these people. I know you don’t have to. I mean it’s all they just download the app answer any questions and boom, if they pass they get the money and they start going, is that right.
Shivani Siroya: Yes exactly, yeah.
David Kruse: Interesting. Yeah, its defiantly the future and that would be nice over here to think. What’s a typical loan amount?
Shivani Siroya: So our average loan amount is about $50.
David Kruse: Okay, interesting. Yeah so you have to – Wow! You get that you have to loan a lot of money or have a number of – a lot of transactions to get to $2 million, that’s cool, okay. And do you disclose what your default rates have been on.
Shivani Siroya: Sure. Our non-performing rate is under 7%.
David Kruse: Oh! Wow, okay. And is that pretty typically; did you see that when you were working for the UN, did you see other default rates for other micro financing institutions?
Shivani Siroya: Well I think the biggest thing is that our loans are completely unsecured. And so not only are they unsecured, so we are not requiring out customs to put up any collateral. In addition to that we don’t need our customers in person or to pick up the phone and so this is completely automated. And so we have definitely seen higher repayment rates in microfinance, but the big difference is that they are meeting every customs, in person one by one and they are usually using a group, a group kind of a loan program. So they are actually attaching the risk to the group as opposed to the individual. And so if one person doesn’t pay back then their group is responsible for the loan. In our case we are really scorning just the individual and we do it in a completely automated way.
David Kruse: Yeah, it’s a huge difference. And all right so we are almost done with the interview, but yeah a couple more quick questions. One of them is, I’m always curious, you know when you are starting this up, this seems like a big task, a big project to start because you are in LA and you are serving clients all over the world. What was one of the major challenges you ran into like you didn’t really expect that you would run into or it could be any type of challenge. It doesn’t have to be – maybe you expected it, but…
Shivani Siroya: I mean, I think one thing that I always think about is, it’s not necessarily a challenge, but it’s something that dramatically I think changed my outlook and you know kind of the work that we do. So we always believe that it’s important to really like know our customs well and to build a relationship with them. And you know when I think back to our first product, we were using text messaging and interactive voice technology to start out with and in some ways we miss judged our customers. You know we thought that they didn’t have access to this kind of android technology and Smartphones and that they could be so sophisticated and so we were building for, we were building in a sense for you know customers that were just on the lower end and I think in some ways like what it made me realize is that, you can’t really, that you are always going to go – sorry, I’m like messing this whole thing up. But I think what I’m trying to say is that – I’ll start over. So one of the things that really surprised me and really changed our work was something that happened about two years ago at this point. So we were initially starting out and as I mentioned brining in self reported data from our customers on their revenue and expense information using text messaging and interactive voice technology. And what we did was we miss-judged them in assuming that they didn’t have access to Smartphone technology or android phone. And I had this experience when I was talking to one of our customs in Kenya and I was training him on the whole product and you know at the end of it, he is like yes, I’m excited, I’m going to use it and he is like, ‘but you know what would be really helpful?’ and he brings out his other phone and it’s an android phone. He is like, ‘it would be really helpful if you had an app for this’ and I was like, ‘What, I can’t believe that you didn’t tell me that you had another phone?’ And he just looked at me and he’s like, ‘but you didn’t ask’, right. The whole point is I think that the biggest thing I’ve learned is that you need to ask and you need to stay curious and continue to adjust to the market. And I think that’s what we take really seriously, it’s because of the fact that we now have access to all of this daily life data. It’s our responsibility as a financial services company and technology company to continue to refine our algorithms and to continue to refine the financial products that we give to our customs, because they are giving us all this access, right. They are giving us this information and so if we don’t do something good with it then it sort of, it’s on us, it’s our fault.
David Kruse: Well, that makes sense. I like that you have to ask. All right so, I think we just about done. Before we end, I was curious where in five years would you love Tala to be as far as you know punch more countries or would you improve your scoring model or what would a – kind of what’s your vision for the next five years?
Shivani Siroya: Yeah, I mean I think there’s two things we think a lot about, our geographic expansion. So we know that there are almost 3 billion people around the world that are currently under served and that are not getting access to the financial products that we need. So we want to bring this to more countries and to help open up those financial systems and then I think the other thing is thinking as mentioned about you know providing deeper value to our customs within the platform that we already have.
David Kruse: Yeah, that makes sense. And when you go to a new country, is it a big kind of reset. Do you have to learn a lot about those people or do you see similar patterns between countries?
Shivani Siroya: I mean it also depends. So you definitely have to look alike your product and your algorithms for each market, but we do see similarities and we see that you know I would say about 70% to 80% of the data that we’ve seen in Kenya does exist in other markets and so we are not necessarily having to rebuild everything from scratch. But you do want to take in to account all of those cultural differences and the differences in how people are using their phones in each market.
David Kruse: That makes sense, okay. Well Shivani, this is great. I think that just about does it unfortunately, but I really appreciate your time and what you’re doing for the people across the world and I wish you the best of the luck and hope that it keeps going really well.
Shivani Siroya: Thank you so much and thank you for the opportunity to share our story.
David Kruse: Definitely, and yeah its fascinating. Like I said at the beginning, I guess I was right. It sounds like the future of financial to me and thanks everyone for listening to another episode of Flyover Labs. As always I definitely appreciate it. Bye.