Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
The Lending Business That Never Charges Interest or Late Fees | Nalin Agrawal (SnapMint)
0 Playsin 3 hours
Nalin Agrawal, Co-founder of SnapMint, is a three-time entrepreneur and IIT Bombay alumnus who has quietly built one of India's most efficient consumer financing platforms, scaling from a tiny ₹5 crore revenue to ₹350 crores while serving 7 million monthly users across 23,000 pin codes.
In this candid, wide-ranging conversation with host Akshay Datt, Nalin reveals the contrarian principles behind SnapMint's success: why they have never charged a single rupee in late fees, how their data science moat achieves industry-beating credit loss rates, and why they believe India will leapfrog credit cards entirely and go straight to EMI on UPI.
What you will learn in this episode:
👉How SnapMint built a 2.5% credit loss rate versus the industry average of 6-8%, using machine learning models powered by 3,000 data factors and a sophisticated real-time fraud detection engine that catches organised fraud patterns in under 10 minutes
👉The four-quadrant framework, Market-Product fit, Product-Channel fit, Channel-Model fit, and Model-Market fit, that Nalin uses to evaluate every business idea and what investors are really looking for at Series A versus Series B
👉Why SnapMint calls itself a transaction-led business and not a lending business, and how this distinction creates fundamentally different and more predictable unit economics compared to traditional balance sheet lenders
👉The story of how a failed advertising campaign in 2016 revealed a 300 million consumer opportunity hiding in plain sight, and how that insight became the founding thesis for SnapMint
👉How India's digital public infrastructure stack, UPI, Aadhaar, Account Aggregator, and the Unified Lending Interface, is enabling fintech companies to serve tier 2 and tier 3 consumers at a cost that was previously impossible
If you find this episode valuable, subscribe to the Founder Thesis Podcast (Listed as one of the Top 45 Indian Entrepreneur Podcasts by FeedSpot) for weekly deep-dives with India's most compelling founders and operators. Follow Akshay Datt on LinkedIn and X for curated insights, episode drops, and startup ecosystem commentary delivered straight to your feed.
#NalinAgrawal #SnapMint #BNPLIndia #IndiaFintech #EMIonUPI #ConsumerFintech #FounderThesis #AkshayDatt #FintechIndia2025 #SeriesBFunding #DigitalLendingIndia
Disclaimer: The views expressed are those of the speaker, not necessarily the channel
Recommended
Transcript
SnapMint's Consumer-First Lending Approach
00:00:00
Speaker
So far, we have almost given out close to about 8 to 9 million loans. We actually haven't charged a single penny of late fees ever. How do you make money by giving loans which are interest-free and you're not even charging late fees? That sounds like magic. The difference between us and typical lending company, in one model, you rely on the consumer making mistakes versus for us, we are consumer first when we want to be honest and transparent with the consumers. We are working exactly in the opposite way. Nalin Agarwal is the founder of SnapMint, a leading fintech that has raised $180 million dollars and serves 7 million monthly users across India, providing them zero-interest loans with no credit card required. Digital frauds are growing with technology. We have fraud engines to figure out if this transaction is either an identity fraud or a ring fraud. What exactly is the account aggregator framework and how does it help?
00:01:00
Speaker
Nalin, welcome to the Founder Thesis podcast. ah You are the founder of Snapment, which is a massive lending business. ah Let me start with my first question to you. ah Give me an elevator pitch of Snapment. Why should I care to learn about Snapment?
00:01:21
Speaker
So Snapment, essentially, it's ah um it's an EMI payments platform catering to the next 100, 200 million digital native consumers of this country.
00:01:32
Speaker
um we We cater to almost 5 million consumers today. And in the last four years, have scaled it up from almost 5 crores in revenue to as high as 350 crores today.
00:01:45
Speaker
And growing pretty fast in in
Partnerships and Merchant Collaboration
00:01:48
Speaker
India with a simple, clear focus on providing interest-free credit for consumers to actually make purchases for what they like.
00:01:59
Speaker
that That sounds like magic, interest free credit. Okay, so I'm going to get into that, like how you do interest free credit, just very quickly, what what's the funds you raised till date?
00:02:11
Speaker
So till date, we've raised almost $150 million dollars in funds. but yeah And what is the, I think one of the metrics for lending book, for a lending business is the, either the loan book or how much you have disbursed so far, right? Yes. the I mean, the clear way to understand our business, it's very different from a book-based business. It's more of a transaction-led business.
00:02:35
Speaker
It's all about, you know, what is the, kind and what are the consumers purchasing with you and how much are they purchasing with you? So, so far, for example, we have raised, I mean, we have close to 5 million consumers who transact with us. And on an average, our consumers spend close to about 20, 25,000 rupees with us over a period of a year, for example.
00:02:58
Speaker
So that's, But but but ah does the consumer know this or you are behind the scenes? From what I understand, you are ah so you are in the EMI enablement space, if I understand correctly. So for example, when somebody is buying a phone,
00:03:16
Speaker
ah you help the seller to tell that customer that, hey, ah do you want to pay this in three interest-free installments? And that is what you mean by interest-free credit. Am I right and understanding the business? yeah Yeah, exactly. So I think ah so any merchant that you work with, based on what the merchant is selling, ah
Fraud Prevention and Data Utilization
00:03:39
Speaker
either three months, six months, nine months, or 12 months EMIs,
00:03:44
Speaker
is what the merchant offers to the consumers, typically on interest-free. And we are the ones at the backend who underwrite the consumer and then give the credit out to the consumer. And then we collect the money back from the consumer.
00:04:00
Speaker
Okay. Do consumers know SnapMed as a brand or they know, for example, I bought an iPhone and that iPhone I was able to pay a over 12 months interest-free. ah And like, ah is there such a thing as stickiness in this business or the nature of the business is the merchant is the one you want ah to be sticky with you?
00:04:20
Speaker
ah So um consumers, obviously they know about the brand because it is us that they need to pay the money back to. Right. So they do know about us. And it's it's actually a consumer first business, because what we believe is that if your business, if you made a business which is consumer first, then the entire business model will also be attractive to the merchant.
00:04:45
Speaker
The merchant then will be I mean, so if if you and the merchant both are thinking consumer first, right, it's only then can the product or solution actually work for both the consumers and the merchants.
00:04:56
Speaker
So it it is a consumer-led business, and we have extremely high retention rates as high as 70% retention rates year on year are the kind of retention rates that we have.
00:05:10
Speaker
What does it mean to be consumer first in your line of work? Are there some trade-offs that you can share as examples that you would do this but not that so that you are consumer first or some design choices or things like that?
00:05:24
Speaker
Yeah, there are... um I think there are quite a bit of things. right so I think the the entire um idea behind the way we look at SnapMent, today consumers have two choices of payments on credit. One is your credit cards and the other is a platform like us where you give EMIs.
00:05:42
Speaker
And where we differentiate ourselves against a lot of the work that we do is is essentially being very transparent to the consumers, right? So simple things like late fees, we actually haven't charged a single penny of late fees ever.
00:06:00
Speaker
Wow. okay Which is absolutely opposite that of any credit card or a so-called lending business. um There are things like, you know, for example, a credit card model, a simple credit card model, what they would do is that, um you know, so if your monthly income is 50,000,
00:06:17
Speaker
they would actually give you a credit limit in the credit card which is worth one lakh, which is twice your income. Because ah what the credit card companies want to do really is that if you land up spending a 60 or 70,000 rupees that you cannot afford, ah they want you to say that, okay, towards the end of the month, I will pay just this minimum due of, 3,000 or rupees, right?
00:06:44
Speaker
And for the rest, you would be charged interest, which is as high as 48, 52%. And that's how the credit card companies work. And so they always give you credit limits, which are much more than your income, because if they don't do that, they will not be able to make money in the model.
00:07:01
Speaker
um So outside the late fees, this is the second area where as as an EMI product, we are giving extremely, I would say, definitive terms to the consumers. The consumers exactly knows that he has to pay back in three or six months.
00:07:15
Speaker
ah If we if if you believe that the consumer at any point of time is getting over leveraged, we actually don't allow the consumer to make that transaction, which he called as responsible lending.
00:07:27
Speaker
So it's all about not letting the consumer get over leveraged for every transaction that you i mean that they work with you on. And that is really where the difference between both the models is. in in In one model, you rely on the consumer making mistakes and that is what you earn from.
00:07:48
Speaker
Versus for us, when we are consumer first, when we want to be honest and transparent with the consumers, we are working exactly in the opposite way. We are trying to make it work for the consumers that we don't make money from their mistakes.
00:08:01
Speaker
Okay, interesting. I just want to define over-leverage, the term you use, which may not be universally understood. but So over-leverage is basically when you have a lot of debt, like like you've taken a lot of loans or your credit card ah has a lot of overdue amounts, so more than what your income can justify for you to have. So so such consumers, you...
00:08:22
Speaker
ah don't approve ah them for the EMI plan. Yes, yes. Okay. okay So ah I want to understand the magic behind this. As I said, this sounds like magic. ah Interest-free EMIs.
00:08:36
Speaker
There is no such thing as a free lunch. Who's paying for this? How do you make money by giving loans which are interest-free and you're not even charging late fees? There there is essentially...
00:08:46
Speaker
nothing that you are earning from the person who's taking the loan the consumer so how are you earning I think it's a standard uh business model which is prevalent globally right it is essentially if you look at the business model the merchants want to increase the sales by giving this liquidity to the consumers of not paying for the entire amount upfront but over three six nine month installments merchant is able to increase his sales right so it's a merchant who gets the benefit and it's the merchant then who subsidizes the interests for the consumers uh in the offline world you would see a bhajaj finance providing zero percent emi to the ka consumers for smartphones televisions and washing machines where someone like a bhajaj finance would actually get the interest subsidies uh
00:09:36
Speaker
The brands like Samsung and the other brands are selling out there. In a very similar way, when we work online, we actually get our commissions from the brands because we help them increase the sales by 10 to 20%.
00:09:50
Speaker
Okay. Okay. 10 to 20%. So this is a, like a industry standard that by offering pay through EMI option, merchants get 10 to 20% more sales. Like this is like across the world, et cetera, like an established thing. Yeah.
00:10:06
Speaker
it's It's an established across the way that is the way it needs to work across categories. It might differ for certain categories, which are say like iPhones or, uh,
00:10:18
Speaker
you furnishing businesses, it can go as high as 20 to 40%. Okay. Like high ticket purchases, the ah yeah EMI option can drastically increase. Yeah. but I remember I was buying a DSLR camera where the only reason i actually bought one was because of this emi free EMI facility, which was available at that time. Otherwise I wouldn't have bought it at that time.
00:10:41
Speaker
ah Yeah. I totally get it. Um, so, uh, what is the subsidy which a merchant is giving you? Like, uh, normally in the lending business, uh, when it's like an unsecured lending where there is no asset to back the ah the loan, uh, banks or nbfcs digital lending companies they will charge anywhere from say a 15 to 25 rate of interest um do you earn a similar rate of interest on the money that you lend out from the merchant like when the merchant is giving you some sort of a payment to make the emi available sure it's not like i mean as i said it's not like a lending business where you charge an a roi it's about it's a transaction-led business where um
00:11:28
Speaker
that where we work with the merchant essentially on what the commission that the merchant would pay on the transactions. Now ah basis who the merchant, what the category is, this number can go from as low as 3, 3.5% to as high as 10%.
00:11:45
Speaker
um Because it it it really depends upon the margin that the merchant has. on the particular product and if it's a high ticket size product then the kind then the merchant would get a lot more benefit if he were to say give a nine month or a 12 month EMI at 0% where a merchant would be willing to give a lot more commissions.
00:12:06
Speaker
So it that it is to do a lot with the demand and supply balance that the merchant sees in his head to drive his business and that's how the commissions come up. Okay. ah So you've you've told me multiple times it's not a lending business, it's a transaction business. ah ah Can you just drive that point home? like Like as a CFO, how does the lens change? Like like i I'm continuing to ask you those questions, like what is the rate of interest you earn? But but how does the lens change when it's a transaction led business?
00:12:36
Speaker
See lens changes in a, I mean, say for example, if you are into a lending business or a personal loan business, right? um You typically would say give a personal loan of close to about two lakh rupees for 36 months.
00:12:49
Speaker
ah Now, the entire idea is once you have given that loan out, ah after that is all about managing what we call as your balance sheet, right? So how much money are you collecting every month?
00:13:02
Speaker
If the consumer fails, then, you know, how do you get that collection back? How do you manage? um I mean, because now you have lent out money and you're getting money every month from the consumer over 36 months, then how do you take loans from your lenders to actually manage that entire piece, right?
00:13:20
Speaker
So ah that is really what a lending business is. Uh, in a business like us, um, it's like, because you are giving three months, six month credit and you earn only on the transaction as a percentage of commission, the business is kind of predictable. You know, exactly how much money are you going to make in the entire business?
00:13:41
Speaker
Right? So, uh, you exactly know that, okay, if I'm making a 5%, then that hits my, uh, profit and loss account or, you know, my, uh, revenues today.
00:13:51
Speaker
Plus, because it is a six month loan, I exactly know within six months how much of how many of the consumers probably are not going to pay me back. That number becomes quite predictable.
00:14:02
Speaker
So, um which is unlike a loan where you rely on a 36 month kind of EMIs, where it is not very easy to predict for 36 months. You can still predict for 36 months. You can't predict the business for 36 months.
00:14:16
Speaker
so That way it' it's it's more about what you are doing today, how much you're making money today, what you're going to lose today. So you exactly know what your profit looks like as of today versus in an ah in in a ah pure lending business, you can't predict that profit.
00:14:36
Speaker
oh Okay, interesting. interesting So the the risk is lower ah is the one big thing because it's more predictable.
Market Dynamics and Business Model
00:14:44
Speaker
And the second big thing is your profit is getting into your account today because the commission is paid today when the customer buys. whereas In typical lending business, every time somebody's paying an EMI, some portion of the EMI is your revenue. ah so So these are like two two major differences. um Understood. ah Could you like you know take one category, let's say,
00:15:11
Speaker
phones and tell me what is the profitability? So like, let's say ah somebody, a merchant is doing 10 lakhs worth of sales of phones in a month, which you are financing for their customers out of that 10 lakhs, what commission would you earn? What would be the and the bad debt in that, like a non-payment happening in that? What would be the net ah revenue that you would generate as a business, just as a hypothetical, if you could talk through that calculation. Sure.
00:15:41
Speaker
I mean, it's ah ah i mean it's it's a very thin margin business. yeah That's why always say it's transaction-led because you need to rely on a lot of volumes. The margins that you get are extremely thin in this business model.
00:15:52
Speaker
so it has i mean So for example, if for six-month EMI, the a my ah merchant would land up paying you, say, a 7.5%. Okay. would go bad.
00:16:04
Speaker
roughly about two to two about two point five percent to three percent of that would go bad in terms of credit losses. um And that is the most important part because most of the businesses, they actually see a 6% to 8% kind of a credit loss, ah which hits their profit and loss accounts badly. And that's one the reasons why a lot of companies have found the business model extremely difficult to essentially manage these credit losses. That's a key item.
00:16:34
Speaker
ah The other piece is your cost of capital, right? So when if if we are giving out money to the merchant for by financing that particular purchase, then we are also borrowing from someone else, right?
00:16:49
Speaker
So because it's typically, say, if it's a six-month kind of a loan and we borrow at probably about 13% or 14% from the market,
00:16:59
Speaker
Then what that would translate for a six month loan is roughly about 3% kind of number goes there. And then you have your collection costs, your operational costs. So what you're left with across every transaction is quite small. It's between 1.5%. So you really have to manage the business model absolutely accurately because one small mistake you do, you will end up losing a lot of money.
00:17:25
Speaker
Okay. Interesting. um You're saying 6% to 7% is the normal credit loss in the industry. You're talking of this ah yeah yeah yeah EMI enablement industry. Is there a better term for it? ah Like this term I'm using yeah EMI enablement. Policy financing, purchase financing.
00:17:45
Speaker
Point of sale financing. and Consumer financing. These are the typical terms. Yeah. Okay. So for this industry, typical credit loss, essentially credit loss is people not paying up their EMIs, is six to seven percent, but you have kept it at under two and a half percent. Am I right in that understanding?
00:18:02
Speaker
How did you do that? i think this is where, I mean, the three of us, right, the three co-founders, Anil, Abhinit, and I, we actually go back together from IIT Bombay.
00:18:17
Speaker
And we have been into the world of data and data science and technology since 2006-2007 together. We understand this world extremely well.
00:18:28
Speaker
ah We have always looked at things from a consumer lens and then tried to always make the, I would say, the machine learning models and the data models ah basis trying to understand what the consumer is buying, who the consumer is, who the merchant is, and looking at a lot of these combinations to build your AI and machine learning models.
00:18:52
Speaker
And that is something that we have done for over 15, 20 years. We are comfortable with that. And we have been able to kind of make this business model work. So you're saying the key levers available to you are... ah who is buying, which is like the consumer's profile, what are they buying, the product, and ah who are they buying it from, which is the merchant. yeah ah These are the three things which you have to...
00:19:16
Speaker
ah like assess at the time when you're offering a loan and you might say no ah to some consumers because they are not passing these filters or you might say no to some merchants or you might say no to some products in order to keep that ah credit loss under two and a half percent. Yes, yes, yes. ah Help me understand whom all you would say no to.
00:19:38
Speaker
What kind of merchant would you say no to? What products would you say no to? And what kind of consumers would you say no to? um I mean, so from the consumer side, who we would say no to typically consumer with low credit scores, we actually don't underwrite them. So, you know, if if consumers from maybe the subprime category or consumers with less than 700 credit scores, we actually don't underwrite them.
00:20:04
Speaker
We do say no to them. ah The other pieces, but but new to credit consumers, we do invite them on the platform because If you see because they are new to credit, they do not have any loans with them. So they're not at all leveraged.
00:20:21
Speaker
So in fact, they are the best consumers who would land up paying back to you because they're not leveraged. ah So to credit consumers good credit scores what we accept. Subprime consumers is typically that we don't do. um in In terms of... ah Products, we actually don't deal with services much because even we are end use financing, right?
Consumer Disputes and Satisfaction
00:20:47
Speaker
So we don't work in the service sector because in the service sector, the um I would say the entire drama
00:20:55
Speaker
ah customer satisfaction is based upon who is delivering the product or the service to the consumer. And which is something which is not standardized, we don't have a control on.
00:21:08
Speaker
We do certain services which are productized, right? So we would say, for example, do flights, you would do trains, um ah we would do DTH boxes. So some of those areas we would do, which are more standardized transactional services, but which are more human led, probably like something like your...
00:21:30
Speaker
Education procedures. ah Yeah, education. So so not K-12, but professional education is something that we typically won't get into.
00:21:41
Speaker
Or wherever there is- You will do K-12, like school fees as EMI, that is something you are into. Yeah, yeah. K-12, we are not into that right now, but we would not say no to that.
00:21:53
Speaker
Okay. So K-12 is absolutely fine with us because again, over there, it's ah it's a standardized product that is getting delivered by the school to the consumers. right So it's it's standardized.
00:22:05
Speaker
ah So that's something from the um i mean that's something that you would typically not do um in financing on the product side. um merchants ah merchants, as long as you're selling some of these products, we are fine.
00:22:21
Speaker
But the merchants that you will not work with, this for example, may if you're not um a lot of small merchants ah who who would have opened the shop recently, right?
00:22:34
Speaker
ah Many of them, they have seen that they actually take quite a bit advantage of the consumers but not by not delivering the services correctly. So if you get too many disputes from the consumers with a particular merchant,
00:22:47
Speaker
we usually stop the relationship with those merchants. ah Because being consumer first is extremely critical. And we we typically like merchants who also think consumer first.
00:23:01
Speaker
So merchants who would not fall into that category or if they start working with us, we realize that they're not working with us to think about the consumer first. You typically would then go ahead and say no to them. Interesting. ah This ah ah unhappy consumer thread, I want to dig into that a bit. I remember when the whole Baiju's meltdown was happening, there were a lot of articles on ah how people are stuck paying off the EMIs for Baiju's product, ah which was missold to them. Their sales team used to do unethical practices to sell the products, whatever, and ah they were stuck paying EMIs. So what happens when your consumer is unhappy with what he has purchased?
00:23:45
Speaker
So the consumer is unhappy without his purchase, chase he obviously first raises a dispute with the merchant. right ah If the merchant is unable to resolve it, then then he has a facility to raise a dispute with us.
00:23:56
Speaker
Once he raises a dispute with us, we try and work this around with the merchant as much as possible and try to give the consumer the best possible solution. So we don't ah just leave it out and say that you know we will not help help you out.
00:24:12
Speaker
but That's how we typically work and respond back to the consumer. Obviously, the final resolution relies i mean lies with the merchant. ah So after a certain point of time, if we see too many such cases. Obviously, we don't stop working with the merchant.
00:24:28
Speaker
But ah in 99% of the cases, that does not happen because the merchant and as in the online world, we have to help the consumer resolve his disappointment. So one is like a legitimate merchant, and there are complaints even with good merchants, and they will work towards solving it. But suppose you ah
00:24:51
Speaker
ended up accidentally working with a bad merchant, ah and the consumer grievances are legitimate. In that case, who foots the bill for that?
00:25:02
Speaker
um I mean, so um there are there are two things out there, right? The way it would work. So one is that genuine refunds, even if, because we are a consumer first company would actually go, go ahead and give the refund back to the consumers.
00:25:15
Speaker
And you'll follow that loss like that. We do. We do. From the merchant, whatever you can, and if not, then you'll. yeah Okay. Yeah, absolutely. So, I mean, uh, we, uh, we have different ways of managing that.
00:25:30
Speaker
We're in, uh, but, but most of the time it's, it's going to be the consumer first approach again. ah We would land up taking some losses and hence I said, right, underwriting the merchant is as critical as underwriting the Kabjee one.
00:25:42
Speaker
ah Because we do lose money when we associate ourselves with merchants who are not thinking Kabjee one first. Okay. Okay. Interesting. ah How does consumer underwriting happen? Like I'm always amazed that it typically takes maybe a minute, right? To get a decision. What happens in that one minute when a consumer says, yes, I want a, like a,
00:26:07
Speaker
12 month interest-free AMI for this product and for you to like kind of offer it, like at what stage does it happen? Is it even before checkout that you've already done the evaluation or is it at checkout when there are different options to pay and someone is selecting interest-free AMI? Like just take me through that flow for a merchant.
00:26:25
Speaker
Yeah, typically the the in underwriting happens after a consumer selects Snapman. He enters the, the I mean the Snapman's, I would say payment window.
00:26:37
Speaker
Uh, over there, whatever information that we need from the country, will like, you know, probably a number of sectors, what we would land up taking from them. Uh, after that is where the end entire underwriting begins, right? Because I think over there, it's essentially at the backend. We, uh, so far we have almost given out close to what eight to 9 million loans.
00:26:58
Speaker
And, uh, out of those 9 million loans, we have actually closed 8 million loans. So what that means is that there is we are sitting on a good amount of data to build our machine learning and AI models to actually do that underwriting. And we actually and it is these models which help us take that underwriting decision within a second.
00:27:22
Speaker
It relies on a lot of data points, obviously, basis the past purchase behavior of the consumer. ah his credit histories, his repayment behavior with the merchants. It's a combination of a lot of factors, are close to about 3000 factors that are there at the backend, which which go into the model.
00:27:41
Speaker
ah For any point of time, there would be about 180 to 200 of these data points so that would be available to us for taking that decision. That's where those ah machine learning models, they take very quick calls.
00:27:54
Speaker
Apart from that, we obviously have to manage. Where are these data points coming from? Like if it's not a returning customer, it's first time customer. where So just the PAN ah number alone is giving you so many data points? Or like what's the source for all the data? So it's typically credit bureaus. It's typically credit bureaus from which we get most of this data.
00:28:15
Speaker
But outside of that, it's also. The PAN number is linked with the credit bureau. So that that data you're able to pull in. Yeah, yeah, yeah. Obviously that is one, then there is is this KYC data that would come in. When has he done his KYC's?
00:28:28
Speaker
ah So um what kind of credit lines has the consumer taken? right um you know How many credit cards does he have? So a lot of those things actually go on to take that credit decision extremely quickly.
Technological Advancements in Underwriting
00:28:44
Speaker
um Apart from that, the second huge piece for us, which is I think one of the most difficult pieces in the market is the frauds, managing frauds. ah We have the way the digital frauds are are growing with technology.
00:29:00
Speaker
Managing frauds is one the biggest pieces. I think we probably have one of the largest fraud engines that actually work to figure out that if this transaction is either an identity fraud or is it like, you know, a ring fraud where probably there are 10 people working together ah who have fake IDs and trying to build that fraud around.
00:29:29
Speaker
Or there's something called an account takeover fraud, which wherein, um you know, some fraudsters, they figured out that these are the set of consumers who already use SnapMint.
00:29:42
Speaker
And then they would go ahead and approach them and start transacting on their behalf. And consumers may not know that. And there are different ways in which these account takeover frauds happen. So there are multiple such frauds that keep happening.
00:29:56
Speaker
And it's it's a fraud engine which works to manage this. And that is where you see maximum number of carried losses in the online world.
00:30:08
Speaker
ah Can you zoom in more on how the fraud engine works? What, like, how does it know this is an identity theft or this is someone impersonating and ah ah consumer that's not the real consumer themselves and things like that?
00:30:22
Speaker
So, I mean, um I mean, typical telltale signs are that, you know, like, say, there is, you So there is a, kind I mean, you suddenly start seeing that, say there are three same smartphones being shipped to the same zip location or the pin code.
00:30:41
Speaker
and Right. And that is where the pattern starts to erupt. Right. Because there is something wrong that's happening because um it is fraud is all about pattern matching. What you've seen over a period of time is that, you know, you you never see three smartphones being shipped to the same location again and again and again in a matter of 15 minutes. You have never seen that pattern with that merchant.
00:31:05
Speaker
And suddenly that pattern emerges. Now, that pattern can be dealt in two ways. That either say there is a merchant who has launched a new advertising campaign for that particular product or he has not.
00:31:18
Speaker
So if he has, then you probably would need, then the system would need to increase those thresholds, say from three smartphones to six smartphones in a period of 15 seconds.
00:31:29
Speaker
ah But if he hasn't, then this is a huge problem. So once that is identified right now, ah now you will, now you'll start seeing a lot more patterns.
00:31:43
Speaker
Like maybe start seeing that, okay, the same IP address, you're seeing the same IP address being used on other merchants also. So that means there's a group of people who are doing these frauds with other merchants. And then you start blacklisting some of those IP addresses and some of those other data points that we have, all other merchants do.
00:32:07
Speaker
So, and and all this needs to happen a matter of five to 10 minutes. Otherwise you'll start losing money like anything. So that is essentially how we kind of, um um I mean, that's that's an example ah of of of making this work, but there are multiple pattern matching algorithms that we have at the back end to kind of make some of these things.
00:32:28
Speaker
and Fascinating. ah So when you onboard a merchant, ah do you do some sort of a API connection to get access to all this data, like and the IP address and all of that? Or does this happen when the SnapMint window opens up at the time when somebody at the point of checkout is selecting interest-free EMI through SnapMint?
00:32:47
Speaker
It typically happens when the consumer lands up upon SnapMint. Okay. Okay. From the merchant, you what kind of data are you taking from? Like purchase history and things like that also? Or like what what data do you... we leave that to the merchant. ah So, you know, so a lot of merchants, when they want ah more consumers to buy with them or they want to increase the repeat rates, then then they have to take the necessary consent from the consumer and pass that data across to us during this same entire process.
00:33:18
Speaker
So it really depends upon the merchant as to how he wants to kind of deal with this. Obviously, the more data that we would get from the merchant after the consumer's consent,
00:33:32
Speaker
is we can we can give higher credit limits to the consumer for him to make transactions of higher purchase value. Or in certain cases where the ticket says you're small, let the consumers make more more frequent purchases.
00:33:49
Speaker
okay okay Okay, understood. right So there's a friction in taking too much data. So only if it's really worth it would the data transfer happen like because you have to get the consent of the consumer. Okay, interesting. um Who are the major merchants that you're working with? like you know who Who are like your flagship merchants who give you a lot of business? Just a few examples.
00:34:12
Speaker
um I have quite a bit of merchants that we that he work with. There's no one single merchant to who does like more than two or 3% of a business. But i think some the key, marquee ones that a lot of the listeners probably would have heard of is like we work with Exego and we are working a lot more with them, especially in the trains and the flights categories.
00:34:35
Speaker
um We work with brands like Titan. ah where on their end on on a lot of their products, we we actually power interest-free EMIs.
00:34:48
Speaker
um we um Then there are companies like HealthCart, we work with them. um And some of the brands like Sleep Company, right? These are all the new brands, especially in the furnishing space.
00:35:02
Speaker
There's WakeFit, some of these brands that we work with them. so A lot of these brands, close to about 1500 brands and OEMs is what you're tied up with now. And growing that base.
00:35:15
Speaker
And you are primarily online, ah like you would work with online merchants. You're primarily online, yes. So ah what's your moat now? Like Bajaj Finance is like a giant, right? It's like a massive company in this space of POS financing or end use financing. um I imagine it wouldn't be hard for them to also start working with online merchants. They probably have, like you said, you have data of 5 million consumers for underwriting or for enhanced underwriting besides the credit bureau. They would probably have 50 million. I'm just throwing numbers out, but I'm sure it's orders of magnitude more in terms of the access to data and things like that. So what's your motive?
00:36:01
Speaker
Sure. I mean, Bajaj Finance is an extremely large company. And in fact, I respect the company for what they have done essentially in the offline space. It reached 114 million companies today.
00:36:12
Speaker
um I think go um if to to make this business model work, right, it is it is way too data and machine learning dependent.
00:36:25
Speaker
Taking that decision in that instance second is extremely critical and especially the fraud management piece, right? For a lot of offline incumbents, not just, I mean, not just Pajaj, but any offline incumbent, right?
00:36:39
Speaker
The entire fraud management piece and some of the underwriting pieces are dependent on the individual who is there at the stores, who's kind of helping the consumer take that purchase finance credit.
00:36:52
Speaker
In the online world, we do not have that facility. So, um, So for anyone to get into this space, they'll have to kind of you know make the entire fraud science and the data science engine work for the online world, ah which is which is very different from the offline world. And that's one of the reasons why we're not there in the offline world, because offline, the um the data points, the fraud signals are extremely different from what you would see them in the online world.
00:37:21
Speaker
Interesting. ah I mean, it still doesn't sound like a tough problem for Bajaj to solve, like hiring a bunch of engineers to build a fraud engine.
00:37:32
Speaker
But you're saying that there's a cost of learning. like like It'll cost them to learn what signals to look for. Exactly. It'll cost them to learn. And it takes it takes patience. right It takes two to three years for these machine learning models to actually come up. um the But on the other side, what kind of lands up happening is that while if while this is happening, um merchants, they keep, but I mean, we keep improving our algorithm. So in the end for us, it's about ah giving the highest possible approval rates to the merchant.
00:38:07
Speaker
So you work on approval rates that could go as high as 70%, 80% in some cases, while keeping your credit losses extremely low. So for anyone to get into the market, they would first have to beat this threshold of 70, 80% of Google rates and do this while making money.
00:38:26
Speaker
right Because if you land up losing more than 2, 2.53% in this market, you you'll will make huge losses. So I think marrying that is is is is is the key to the business.
00:38:41
Speaker
And if anyone can do that, yes, they can come ahead and compete with someone like Strappoint. Yeah, okay. So essentially your mote is confidence and I'm using confidence as a statistical term. like Like there is a certain confidence because of which you're able to give that 80% approval rate because your algorithms are able to give a yes, no decision with a high level of confidence.
00:39:03
Speaker
yeah And that is hard to beat. ah Okay, very interesting. ah What kind of ah collection intensity does this business need? Is it like just sending a reminder, SMS, and that's enough? Or do you also need a like a team of callers who are calling and reminding and like field agents, et cetera? How collection intensive is this business?
00:39:29
Speaker
Sure. see I mean, when we're dealing with almost 5 million to 10 million loans, yeah right? the collection intensity is quite high. and But at that volume, um and especially with the margins that we mentioned, right you need to make it work again in a very technology machine learning way.
00:39:52
Speaker
So that is extremely important in this business. So number one, if you're able to acquire the right set of consumers with the right credit scores, you're not going after and un i mean over-leveraged or subprime consumers, right? So first of all, that's the best filter, right? Because these consumers will pay you back.
00:40:15
Speaker
Then when consumers set up, say, their auto pays, right? The UPS scanning instructions are auto pays. Because now it is the right set of consumers who setting the auto pays, the money keeps coming back to you ah at the right time.
00:40:31
Speaker
and And, I mean, that is really what the essence is. Getting the right consumers, reminding them on time, ensuring that your messaging goes on time to them.
00:40:42
Speaker
ah So, I mean, just that lets us collect most of our money, right? 98% of our money just comes from this. Now, whatever that we are unable to get, we obviously then remind them, we tell them about how it could affect the credit scores and the rest of the collections then happen.
00:41:03
Speaker
And what then does not come obviously goes bad after that. But that's what it is. Okay, okay, okay, understood. Okay. um yeah I want to understand a bit about the innovations in India around FinTech space. so I believe just like there is a UPI there is something similar being planned for lending. Can you tell me about that?
00:41:28
Speaker
Sure. So think, I mean, just to understand India, right, it's it ah I think we're at very, very unique juncture. um if we If you look at the entire, I would say, the pyramid, the socioeconomic pyramid of this country, ah most of the wealth is currently being owned by the top 1 to 3% of the people.
00:41:49
Speaker
of the people Right. This is where you would find all the mutual funds, all the wealth companies going after them. You would find, I mean, just those top 3% of those consumers, which is 20, 30 million consumers, you'll find 30 credit card companies just competing for them.
00:42:08
Speaker
Right. ah Now, the entire challenge that is there, there's there's obviously the bottom 50% pyramid, right? um Over there, as as obviously the country progresses in terms of GDP, we would see a lot of these consumers from that segment moving up to the consumption segment.
00:42:28
Speaker
But at the bottom, that's so that's a very difficult piece to work. And microfinances in industry has has has been trying to do as much as possible ah over there. it's It's not an easy industry to be in ah because the because a small mishap like a flood can can kind of wipe out a good amount of business oh for that consumer segment, right?
00:42:55
Speaker
So it's essentially the entire middle income consumer segment, the 3% to 50%. That is where the real value lies. And that is where the underpenetration is huge.
00:43:09
Speaker
And that's not just in lending, right? that's that's That's true with insurance. That's true with wealth. That's true with... the Purchase financing, let's do ah Personal loans, business loans, any any category you look at, it is essentially that segment where if you go and execute it correctly, um there's a lot more value that can be generated. So yeah know we have examples AU Small Finance Bank, who actually was able to go after that entire segment.
00:43:39
Speaker
and and and and and build a great business from their business model as compared to probably a lot of other small finance banks. um So um I think there are these pockets, multiple consumer segments that tend need to be dealt very differently.
00:43:57
Speaker
And only if you understand the consumers in your consumer segment well, can you actually design new products for them. ah Now, where...
00:44:09
Speaker
where you have this new interface called as the unified lending interface, which which the teams are looking at, or there is the ONDC interface for building up a more transparent marketplace, right?
00:44:24
Speaker
These are some of the great initiatives that are coming in now, ah because the key is that um the cost of distribution,
Offline Engagement and Expansion Strategies
00:44:32
Speaker
right? When you get to those next 30-40% consumers, cost of distribution is extremely high. How do you actually distribute your product to a consumer in undermining? It's extremely expensive.
00:44:44
Speaker
Now, what of these interfaces are able to do is that they're able to get the distributors on one hand who have access to those consumers and then they are able to get the regulated entities like, say, the NBFCs or oh other regulated entities on the other side and ah give them a common language to talk.
00:45:04
Speaker
so that ah you give the best possible solution to the consumer. ah Consumers are able to access almost every bank or NVFC ah that could work with them and and and try to create that kind of a marketplace.
00:45:19
Speaker
I think that is what the unified lending interface is is kind of trying to do or ONDC is trying to do. ah Get that access ah of credit credit products to those next 30, 40 million consumers.
00:45:35
Speaker
And because these consumers are young, they're all used to the digital way of working. The hope is that in the next five or seven years, digital is going to become the number one way of making it work for these consumers.
00:45:47
Speaker
ah Can you give me an example? So you're saying cost of distribution is high and ah these ONDC and ULI companies platforms are bringing distributors and lenders on a single platform what what does the distributor mean here what do you mean by cost of distribution is high sure so uh let me give you an example right so we today serve 23 000 pin codes in in India and we are able to serve the 23 000 pin codes because of the internet infrastructure right so any any consumer who wants to buy with the merchant
00:46:21
Speaker
ah Anywhere he sees anywhere in India does not make a difference to us. right We can actually go and give him credit. now ah Now, this is true for those 200, 300 million consumers in India who are buying online.
00:46:36
Speaker
Okay, what about the 300 million consumers who not buying online? right ah there is At one point, there is say someone like Abhijaj Finance who is helping a lot of those consumers buy those consumer durables in the offline world.
00:46:49
Speaker
But there are also this concept of business correspondence, BCs, as we call them, right? ah They have a network of people or so very small branches up in rural areas where a consumer can approach them.
00:47:04
Speaker
And over there, there is one person who is actually helping them out. And this person can actually help them buy tickets of, I mean, train tickets. This person can actually help them buy a smartphone.
00:47:17
Speaker
This person can... help him get a loan against his property if required, can help him get his Jandan account problem solved. So, ah I mean, so there are these business correspondents out there in the offline world.
00:47:33
Speaker
They're like a virtual bank branch in a way. Like like there will be some retailer or a small merchant in a village who is a business correspondent and who's like a virtual branch of a bank, besides also offering other digital services like ticketing, etc.
00:47:48
Speaker
Exactly. now Now, you can't imagine there are close to about 500 large NBFCs in India, middle-air NBFCs as we i'll call them.
00:47:59
Speaker
Uh, you can't imagine that each of those 500 NBFCs are going open up branches in, in that village or each of the 140 share view banks are going to open up branches out there. You can't, I mean, you can't, uh, do that. Right. And that is where something like a unified lending interface or an ONDC network comes in, where in the business correspondent who sitting there, uh, he has the technology, which is tied up to the unified lending interface.
00:48:24
Speaker
And then there are these banks and NBFCs, the backend. to whom this information gets broadcasted and then they have the choice, okay, how to actually make this work. The business correspondent, in fact, if the relationships are good, if if you know you start rating the business correspondent himself, then you can also rely on collections with this business correspondent with the right incentives for them. So that business model can actually work for ah financial inclusion for the next 300 million um consumers um as you look at them. and i think I think that is where, otherwise that's it's very difficult to distribute in some of these tier two, tier three locations.
00:49:06
Speaker
Interesting, very interesting. There's also another innovation of the account aggregator framework, which has gone live quite some time back, but I don't hear uh too many people talking about it uh tell me about this what exactly is the account aggregator framework and how does it help and has it actually been ah practically adopted sure it's a beautiful framework i think we were probably the second company in india to adopt the account aggregator framework um the i mean imagine this right a consumer is coming and he's only giving his mobile number and otp
00:49:45
Speaker
and a consent to tell us that, okay, these are the banks that are linked with this mobile number. And you know what, I give you this consent to kind of look at my past six month or 12 month history of transactions.
00:50:00
Speaker
So it's all happening with just a single OTP. and And that is beautiful, right? Because now, um you know, um the biggest challenge earlier was when you used to get these past book statements,
00:50:15
Speaker
Okay. Or where a consumer used to have a net banking account and he used to then get his PDF statements, etc. ah It was very, yeah there was a lot of manual work that was always required to kind of make that work, understand the cash flows of the a consumer.
00:50:33
Speaker
um And then we were also getting lot of frauds again, right? Because you could manipulate the PDFs, you could manipulate the passports. And, uh, then how do you make your credit underwriting system work for some of those things? It was it is really very difficult.
00:50:49
Speaker
What a Account Aggregator did was that even if you do not have access to net banking, right, yet all the records in its original form will actually come down to you, which which ah which helps a lot because now, especially for new to credit consumers, you have a way of giving them ah say iPhones and 0% EMI if they have enough spending power and all that's possible with just one single click without worrying too much about these document frauds.
00:51:22
Speaker
So it's a beautiful interface, I think, with mutual funds, insurance and all of these getting linked to the account aggregator and giving that consent and the power to the consumer to share that.
00:51:35
Speaker
right so The consumer can decide today that, okay, if he wants to buy more or if he wants much more loan or if he wants to get a loan against his mutual funds, all that's possible today with just a single OTP.
00:51:48
Speaker
And that's the beauty of the account aggregator interface. And ah is it working as expected? There were teething challenges earlier. There's still some teething challenges, but I think we're getting there because... um If you look at the public sector banks, right, in India, public sector banks probably have 70% of market share. ah And the infrastructure that they have is so distributed that, you know, expecting someone at the scale of state bank or a Punjab national bank to ah to to to give that service level, okay, in terms of the data coming immediately, at etc., is
00:52:32
Speaker
It's not easy because of the sheer volume of infrastructure that they have. And probably legacy systems and things like that. Yeah, legacy systems and a lot of those things are there. So it's, it's ah I mean, ah there were teething issues earlier, um but now it has improved quite a bit. There's still a lot of room for improvement. I think we're getting there. We should be getting there very, very soon.
00:52:57
Speaker
Okay. ah When do you actually need to look at a bank statement of a consumer? So it's, so I mean, typically when the consumer wants to make a high value purchase, right. And we, I mean, from a standard history, we are unable to figure out what his income levels could be.
00:53:14
Speaker
um Then this, I mean, we would like to look at his, his bank statements or his income levels, et cetera, to, to, to, to take the decision.
00:53:27
Speaker
Okay, okay, okay. Understood, and understood. oh Do you, I'm not sure if you want to answer this question or not, feel free to, we can skip also, but um I'm just wondering, do you have advice for people in terms of how to protect their credit score and how to protect themselves from impersonation and fraud and things like that?
00:53:50
Speaker
Sure. I mean, absolutely. I think and and in terms of protecting from credit scores, obviously paying all your dues in time is extremely critical. um Not taking too many personal loans at the same time.
00:54:05
Speaker
If you have too many personal loans that are going on at the same time, typically more than three or five, right? That's essentially over-leverage. I mean, you don't want that.
00:54:16
Speaker
As long as you're taking a loan and you're closing it, you're absolutely fine. But too many parallel loans, if you have, is is usually a red sign in WGP credit scores. and don listen If someone is buying something on EMI, that is seen as a loan.
00:54:31
Speaker
So if I if i bought 10 different products on yeah EMI, it is seen as 10 loans. So if you buy 10 products on EMI, typically what happens with SnapMint is that ah we generate a single a single bill for you.
00:54:47
Speaker
Okay. So yeah, because it's a single bill, it gets reported in a single line. This is when you're doing checkout at one time. But if you check them out at different times, like this month I bought something or this month I bought first week I bought something, second week I bought something, then these would be separate loans.
00:55:02
Speaker
ah No, there's still typically, as long as it's it's with us, we kind of club everything into a single bill for them. Got it. ah Yeah. So it it usually helps consumers improve the credit scores.
00:55:14
Speaker
oh but But again, even if you're taking a lot of personal loans, for example, ours a consumer loan, it's not a personal loan. So it goes into a very different category altogether. but even if you're taking a lot of personal loans and it gets reported as long as you're closing them and you don't have too many lines right you're absolutely fine okay timely payment not too many loans these are like the two broad things that consumers should not too many active loans not too many active loans okay okay okay and uh what about protecting yourself from fraud
00:55:47
Speaker
And think over there, it's all about ensuring that the information that you share with any third party, right? ah You always, i mean, for example, if you're sharing a photocopy, always write the use as to for whom it is, why it is, before actually sharing that across so that it cannot be used anywhere else.
00:56:09
Speaker
ah Apart from that, i think RBI has done an amazing job in RBI. building KYC guidelines, which kind of if if followed correctly, um um it it it solves for a lot of and fraud issues.
00:56:25
Speaker
so So I think, but on the consumer side, it's all about, you know, um like Aadhaar card, right? There is no necessity, no no one today, no regulated entity in this country will ask you for your Aadhaar card number.
00:56:42
Speaker
Okay. They always we are supposed to maintain if we ask for al alpha numbers, we're supposed to maintain them in a mask fashion. ah so So if you give it to any entity in a mask fashion, they are supposed to accept it.
00:56:56
Speaker
So I think giving out the clear numbers should typically be be avoided. But if you have to give them the use has to be clearly mentioned. Okay. Okay. Okay. Uh, so, you know, we spoke of how your motive is confidence from a statistical perspective, ah which is essentially an outcome of data science and data science is somewhat of a legacy.
Nalin Agarwal's Entrepreneurial Journey
00:57:20
Speaker
You are a serial entrepreneur, uh, take me through your journey and what led up to starting snapmate and building this moat of confidence.
00:57:31
Speaker
So I think, um, um, I mean, Snapment, we kind of started off in 2017 and we are actually working with an e-commerce company back then. And this e-commerce company... But just go back further, like pre-2017, like what led up to that 2017 moment? Like you started your first business much before that, right? like Yeah, yeah.
00:57:52
Speaker
Sure. I mean, first business was sometime in 2006, 2007. So I graduated from IED Bombay ah with my colleague Anil in 2004. A couple of years in job, then I was with ITC ah with the cigarette division and then with ZS Associates who was into analytics.
00:58:13
Speaker
And then kind of 2006 is where we actually know ah was the first stint towards entrepreneurship. So you and of you were both, I'm assuming, in well-paid jobs with an IIT tag. ah What made you, and this was not really a time when people used to watch Shark Tank and admire entrepreneurs and entrepreneurship. but what What was the trigger for you to quit your well-paying jobs and start something?
00:58:39
Speaker
think this was something that all of us had discussed way back, even when were in IIT, right? So at at some point, we would want to go into entrepreneurship. And somewhere or the other, it got into our heads that that that we have to get into it.
00:58:51
Speaker
Two years into the job, we realized that we keep getting sucked into it. And now is the time to take the risk. And if you don't do it today, we probably will not be able to in the future. So it's more was like a fear of missing out that that we said that it's it's probably the right time to get into this right now because ah it's the early, you know up to the age of 28, 29 is where you can make a lot of mistakes.
00:59:16
Speaker
After that, the I mean, the risk that you would have and for making mistakes starts reducing. so So that was the reason why we went into this much earlier.
00:59:29
Speaker
And what did you start? So the first company we started was actually into patent analytics. So it was it was we are working with a lot of companies globally,
00:59:42
Speaker
um you know companies like Huntsman, Raytheon, some of these guys. to help them understand the entire pattern landscape, what the competitors are doing, are there any opportunities to license in some of the patterns or license out their own patterns.
00:59:59
Speaker
So lot of analytics around that is where we used to work with all these large organizations globally. And but that was essentially what our uh first company was they're like a data business like like you managed to scrape out the pattern data across multiple countries categories and then uh put it into a format where you could drive insights from it and look at patterns and do some recommendations to clients exactly exactly and but it was more of a services business
01:00:32
Speaker
Because after that, i mean you had to kind of build out the analytics, build out the reports, ah talk to the clients about it. so ah i mean That was what the entire business was.
01:00:44
Speaker
a ah Very knowledge intensive, high touch service business is where it was. And i think over there in that company, you probably reach revenues of close to 1.52 croads.
01:00:58
Speaker
A first company in about two to three years. Only to realize that I think, i think it's, a let's think 10X, let's think where is the next 10 crores going to come from.
01:01:09
Speaker
And that's where we realized that this was a very, very knowledge intensive business. And overall the market of the Pantanatics business was not very big. It was very profitable. We had 70, 80% gross margin in in the business and billing us with a hundred, hundred and fifty dollars an hour.
01:01:24
Speaker
But, but it was still, I mean, the constraints really were on the market size. then not too many companies who file lot of patents a year, right? ah So that's where we said that scaling up this business is is not really of our interest. So we kind of, the company got acquired by another KPO in the industry called Netscribes in 2010.
01:01:51
Speaker
um That's where we then got into the software services business after that. Did you get a good exit when the acquisition happened or it was more like that? I was able to make a home in Bombay. Okay. Nice. Nice. Nice. Okay. Yeah. yeah Okay. send it there But after that, I mean, the next thing obviously was that, you know, I mean, we always had this in mind of thinking 10x, 10x, right? So the other next thing was how do you get to 10 to 10 kilos? That was always in your mind.
01:02:22
Speaker
And that's where we started the software. I mean, we had a software services company, on a lot of other friends was running it. So we came together. ah We started working for a couple of media companies globally.
01:02:34
Speaker
Building what? of like The concept was around video streaming. So it was new at that time, 2009 odd. ah YouTube were just coming up and you know video streaming was extremely new, especially the challenge of how do you transcribe videos?
01:02:50
Speaker
right? How do you make video searchable? How do you make recommendation engines around video? So okay those are problems that have been solved at that time. And again, i mean, it was data and it was huge amount of data ah when when the videos used to come in.
01:03:06
Speaker
So we love that and we are building some of those solutions. We started working with YesBank way back then to help them build their entire wallet system, payments platform, etc.
01:03:16
Speaker
So I mean, that is where we kind of started the working in that, Anil and then I were there, we reached a good business size of close to about 10 odd crores of revenue.
01:03:30
Speaker
ah But then after that, again, the entire challenge that came ahead was that, okay, 10 to 100 crores. And somewhere down the line, now at least Anil and I, we So we are not very people-led, service-led,
01:03:46
Speaker
guys right what we realized over time and being with each other across various ventures was that we love transaction-led models okay not not not human-led models so a business model wherein you could multiply something with transactions rather than humans and that is something that we love and and and what we also knew was that data was literally a darling right so transaction and data is something that we said you know we If that is something that we can scale, services business is something that are not very interested in scaling.
01:04:21
Speaker
And that's where you know we started developing what we call is a today programmatic advertising platform and in 2015.
01:04:32
Speaker
We started developing the programmatic advertising platform, um trying to help advertisers improve the return on ad spends. yeah Can you zoom in on this a bit? Like what's a comparable example of programmatic advertising platform, like a meta Google ads? these are Yeah, something like meta and Google ads, right? So what what Google ads does, for example, is that ah from your browser history, right?
01:04:57
Speaker
It does not know you, but it still knows that this particular browser has gone to these websites. During daytime, he goes to these kinds of websites. Evening, he goes to these websites. and bases those patterns that shows the right set of advertisements to you.
01:05:11
Speaker
right So ah that's exactly what programmatic advertising is. that um So Google operates something like an ad exchange. right So where a lot of this, I mean, imagine you have a Times of India.
01:05:26
Speaker
A Times of India wants to show an ad on its website. Now it needs to tell the advertisers okay that, This browser has come to visit my website and this is the ad slot on this page.
01:05:44
Speaker
How much are you willing to bid for it? Now ah that ah that information comes down to the ad exchange. Ad exchange then broadcasts it to multiple advertisers who want to bid for that.
01:05:55
Speaker
And then you bid for it and the second highest bidder wins. so and And you're typically given 30 to 40 milliseconds to actually bid for it. So it's ah ah it's a massive network. There are companies today like Inmobi and some of these guys who are were pioneers in this space.
01:06:12
Speaker
ah So essentially the idea was we are building up this entire advertising network so that we can help advertisers improve the return on ad spends. So that sounds like a very investment heavy business, right? Because you would need to onboard ah Times of India type of publishers and millions of them for it to be meaningful. And then millions of advertisers and a very... uh, a lot of spend on the tech, I'm sure to build that up for all of that low latency bidding, et cetera, et cetera. Uh, yeah. Uh, I mean, the, we, we, we did not have to run after the publishers because the ad exchanges were the ones who are aggregating the entire inventory from the publishers.
01:06:54
Speaker
Okay. So you were bidding on behalf of brands on ad x exchanges, helping them ah get more bang for the buck. ah exactly Exactly. Got it. So i that was i mean really what the business idea was. And and over there, we were working with one of the large e-commerce companies is where they actually signed up this contract with us.
01:07:12
Speaker
And they had just launched credit card EMIs on the checkout in about 2016 or so. And they asked us to run an advertising campaign for that.
01:07:23
Speaker
And we actually went ahead and ran an advertising campaign for them. Most of the advertising campaigns we ran for that brand, either we saw eight to 9% lift in the consumers who would reach the payment page of the brand.
01:07:35
Speaker
But that particular campaign of EMI, so we didn't write the word credit card EMI in the campaign, we wrote the word EMI. And we launched that campaign out and the lift in the payment pages that we saw was almost 18 to 20%, which massive.
01:07:49
Speaker
which was which was massive ah Unfortunately, only three or four people transacted on on on credit card on EMI. So though we saw lift in the consumers who reached the payment page, but the transactions did not happen.
01:08:04
Speaker
And that is what surprised us. And then we talked to the consumers. Most of the consumers said, you know what, we don't have credit cards. And they all came. And because these are online users, they all came from tier two cities, right?
01:08:17
Speaker
So when we did the geo analytics, what we started seeing was that The previous campaign got, I mean, the previous campaigns were getting a lot of consumers from tier one cities, but this campaign that we said got optimized towards tier two cities automatically because it was getting a good response from there. ah So because that happened, where we then then we started digging into this model more deeply. We talked to, um you know, head of a couple of credit card companies to help us understand what this entire business model is, how EMIs work.
01:08:51
Speaker
what the credit cards model is, why is that these consumers don't have credit cards? We started getting deep into it to realize the huge problem that exists in this country, that that credit cards were not consumer first.
01:09:08
Speaker
It were made for the top 2-3% consumers of this country. um Essentially, ah who people like you and I were looking at rewards, right?
01:09:19
Speaker
and and And that is how the credit card companies were attracting consumers just with rewards and hoping that you would make mistakes. And then I would benefit from that mistake.
01:09:31
Speaker
Credit card companies, minimum time to actually break even in the consumer was 16 to 18 months. You take a huge amount of time to break even in the consumers. which We just did not understand that business model.
01:09:42
Speaker
And we said there's an easier way to do it. yeah EMIs. is Why don't you do that? And banks, we realized, they I mean, know they always work on third party technologies. They never build any technologies ground up.
01:09:54
Speaker
And that's why we said that, okay, banks, there's no third party technology that exists for doing something like this. This needs to get built up from ground. So this is this is something that, you know, it kind of fit in amazingly well in our alley, or a very large market, a huge unsolved problem.
01:10:15
Speaker
ah merchants willing to do something to attract these consumers, consumers eager to get this. Absolutely no bank has been able to serve the segment. So it fit in beautifully well.
01:10:27
Speaker
And we said, let's go ahead and execute on this. And and that's how we kind of got into SnapMind. Was version one then... ah Like a place, like a platform for banks to do end-use financing?
01:10:42
Speaker
Like banks could come in and then, okay. would say number one was that, I mean, there's a lot of, I mean, a lot of investors and community came back and said, you know, you build a marketplace.
01:10:56
Speaker
Marketplace is is something that will work well. You need to put balance, et cetera. So I think that was the version one where we started working with some the banks in the back end, merchants on the other end, started to build a marketplace. And that is where we realized the biggest challenge was because banks were not geared up to take these EMI transactions of 10,000, 15,000 rupees
Regulatory Influences and Strategic Adaptations
01:11:16
Speaker
on the books. They were not very geared for the small value transactions.
01:11:20
Speaker
um the The underwriting practices, et cetera, banks were still doing underwriting like the way they would do it offline. ah there was There was a human element attached to the entire inner writing. It was not instant.
01:11:33
Speaker
So it wasn't kind of it it wasn't working at all. And we being from data and transaction-led backgrounds, ah we were not willing to accept the fact that there's so much of human interaction that's required to build this out. um so and And plus, what we realized very early was that we realized the entire intent of RBI. right um ourbi you know oh RBI understood the intent was that RBI wants the NBFCs or the ones that it has given the license to take a lot of these decisions and work independently. right They don't want to they don't want um technology companies under-regulated to work like a regulated entity and do those kinds of lendings.
01:12:21
Speaker
And realized that way back in 2017, 2018 when RBI said that, hey, you what, fintechs cannot pull credit bureau data, right? Fintechs cannot do other KYC.
01:12:33
Speaker
So those, those, those indications are coming where we realize that I think, uh, you know, RBI is thinking consumer first. RBI is thinking, how do you build the right set of balance sheets?
01:12:46
Speaker
Uh, and, and if that is there, then let's align ourselves with, with that entire piece before we started investing into this business. So we actually went ahead. We got our NBFC license in 2019.
01:12:59
Speaker
before we started investing more in this business. Okay. And what's been the fundraise journey? When did you raise your first round and the subsequent rounds? How has that gone through?
01:13:12
Speaker
Sure. So our first round was 2017, a seed round, which was K-Capital. So Sasha, Mirjain, Nani, he invested in us because he believed in the entire business model.
01:13:27
Speaker
um Then... 2020, obviously COVID was was bad for the entire lending industry. The funds had completely dried up for a period of two months and we just got an NBSC lesson, we were scaling up.
01:13:39
Speaker
We had a couple of term sheets with us and we, in fact, had term sheet on March 19th. yeah Okay. And March 24th is like, you let's wait. I think yeah yeah this will subside. Let's wait for 15 days. This was going to be a series A. Series A.
01:13:59
Speaker
How much was it? doing That was going to be roughly about 5 million. Okay. And this was after and NVFC license. So this after the NVFC license. Okay.
01:14:10
Speaker
Yeah. But it was, I mean, it was, it was very difficult. I mean, the, I mean, uh, 30, 45 days passed and we realized the money is never going to come in. Um, and, uh, it was, it was a very, very difficult time.
01:14:22
Speaker
I think at that time, we we had, I mean, during that period, right we had met this gentleman, Prashast Seth, who actually um loved our business model.
01:14:35
Speaker
And he came back and said, I love your business model. I can understand you guys are in a bad period right now. But I think if you can come out of this period and you can actually show me that you know the business model works.
01:14:47
Speaker
ah then I'll be happy to double things down. And that's where we actually got a million dollars of, I would say, life.
01:15:01
Speaker
know that And because think we had always been focused on being having the consumer first approach and giving credit to the right set of people. Our collections during COVID were absolutely amazing. So, you know, our overall credit losses that happened during COVID, I mean, for the entire COVID the consumers cohort was just about 4.8%, which was way below any industry standard.
01:15:27
Speaker
So ah I think that that that gave the investors a lot more confidence. So once the entire COVID and the moratorium piece in November 2020 was over,
01:15:39
Speaker
is where we started scaling up the merchant network. We started seeing amazing results. We started contributing to 10, 20% sales of a lot of these merchants.
01:15:49
Speaker
Is where then Precious came back and said that, you know what, I love your business model. Forget going down to a venture capital. um i would I would want to take this ahead with you guys.
01:16:01
Speaker
And that's where, yeah I mean, he came in and we raised about roughly about 10 odd million, but ten to twelve wa million in two thousand and twenty one um, and, and from there, I think, uh, today we have almost, uh, we're doing revenues of close to what four three four crores at that time, probably 360 crores in revenues.
01:16:27
Speaker
And I think our Precious has been kind of, I mean, then as as we kept scaling up, right, um he he kept putting in more money ah with with a lot of his clients within Snapwind because he kept believing in the entire the entire business model.
01:16:48
Speaker
Because the business model was turning money, it was becoming profitable. the the the credit losses where the entire industry's credit losses were going up, our credit losses were actually going down, which was completely opposite of what everyone would see.
01:17:03
Speaker
And because the business model, it's essentially was building out beautifully well. The growth kept happening. um The business was touching profitable numbers.
01:17:16
Speaker
So we we never really had to raise a very large amount of capital because we are very prudent about how to build the business out. So it was then now recently when General Atlantic really came in, right? They actually saw ah how big this opportunity opportunity could be.
01:17:32
Speaker
They believed in what we have built so far and the three founders, which is Anil, Abhameet and me. What we're looking to build in the future is where then this round of funding comes in.
01:17:44
Speaker
to essentially help us grow our balance sheets and invest in our entire EMI technology platform a lot more. So this 125 million round is what you're talking about, the general Atlantic round, which is for ah capitalizing the balance sheet. Essentially for a lending business, funds that you're raising is like an input. It's like a raw material. Like like you need money to lend money, even though you take loans from other banks also, but you still need your own. Like, I think the maximum you can take is something like a 3X or a 4X of the money in your own balance sheet.
01:18:19
Speaker
Yeah, typically 4X, I mean, 4X is a prudent number when you are growing at a very fast pace. So, is if if if you're growing faster than, say, 25-30%, then you would have to raise more capital than because then you'll need more debt.
Future Growth Plans and Market Expansion
01:18:38
Speaker
i so That is one use of funds. But I think more than that, there's also a lot more investment that we need in technology as we scale up our entire merchant network.
01:18:48
Speaker
So we'll be utilizing that capital. We we are actually also building out the credit on UPS stack with a couple of banks. So we'll be investing in that. So how do you make credit available across all the merchants, right?
01:19:05
Speaker
on On UPI rails. That's something that we are working with multiple banks on. We'll be investing quite a bit in that technology too. Interesting. So you are now competing with Abajaj Finance, going offline.
01:19:18
Speaker
The credit on UPA essentially is an offline use case, right? for um So credit on UPA really, I mean, I think it's it's more about the consumer segment. I think that's where the difference is because ah our consumer segment is is is fairly young. Our average age of the consumer is about 27.
01:19:37
Speaker
it's a fairly young consumer segment, which is very different from, say, ah a Bajaj consumer financing segment. So it's essentially giving more power to those young consumers in tier two, tier three cities that when they want to buy on installments, they are able to buy.
01:19:54
Speaker
and and um and that workss And that's what really the intent is. And how will that work? Like there will be a QR code, a SnapMint QR code, which people can scan to pay, and that will automatically convert it into EMI?
01:20:10
Speaker
ah So, I mean, not just SnapMint QR code, but it's going to be any QR code. It could be, i mean, so ah say a consumer has SnapMint's app, and there is any QR code, even a phone pay QR code that is there on the merchant.
01:20:27
Speaker
If you scan that, you would be given two options, either paid off now or paid off in installments with certain installment plans for him to select. The consumer can select it and the transaction is done.
01:20:39
Speaker
and he pays our partner bank out with whom we have launched this credit line as a product ah every month. And we help the partner bank collect the money from the consumers.
01:20:52
Speaker
Okay, interesting. So this is more of a consumer play now. So now you want people to download the Snapment app, use it for making payments, especially when they want to convert it to EMIs.
01:21:03
Speaker
Yes. And this would only be for repeat customers or even first time users can download the Snapment app, do the KYC and then? i think it's early days. I think we'll start off with, we obviously will start off with our existing consumers. We have a decent base of existing consumers of close to about 10, 14 million now.
01:21:21
Speaker
ah So we'll be starting off with an existing set of consumers to begin with, but over time we would then open it up to new consumers as well. Okay. Okay. You need some sort of license to allow this, right? Like some some payment?
01:21:36
Speaker
ah for to to have a payment app like which people can scan and pay from. Yeah, yeah. I mean, say you need an approval from NPCI to kind of launch this up, yes. Okay. And ah why not do this from your own book? Why are you working with the bank for this?
01:21:54
Speaker
So RBI does not allow and NBFCs to offer this product right now. Okay. It's only available with the banks. i mean, Credit Line is a very risky product So credit lines as a product is typically that RBI does not encourage and NBFCs because if if consumers you start utilizing the credit lines to a hundred percent rate, then do you have enough capital to actually fund that?
01:22:23
Speaker
So the risks associated with a credit limit kind of a product is quite high. So hence it's typically not, the facility is not available with the NBFCs today.
01:22:36
Speaker
But we hope in the future RBI does allow EMI on UPI, if not credit line on UPI, to actually start working with the NVFC zone. Because over there, then that risk completely goes out.
01:22:51
Speaker
So how a credit line works is that once you do your KYC and submit to your PAN, et cetera, then you are approved for a certain amount, like, say, 1 lakh. So which means that at any point of time, ah that 1 lakh worth of loan is available to you, which either you can transfer directly into your own account as cash, or you can pay for a product using that loan and then pay back in EMI. Is that, that's how it works?
01:23:17
Speaker
ah So, I would say it's like a credit card, but cash cash withdrawals are typically not allowed. It's only for paying it on merchants.
01:23:28
Speaker
Okay. ah So, I mean, you can use it only for paying it on merchants, um not for taking loans against. And over there, it's like, i mean, there are multiple products that are getting launched with this. So one is you either pay it in full next next month or some banks are also building a revolving piece by just paying minimum dues, which which we are not which we are not entering into because we don't believe in that model.
01:23:56
Speaker
ah Some of them are giving features like auto convert your EMI. I mean, auto to convert the entire purchase to EMI. So multiple ways in which this can be dealt with.
01:24:08
Speaker
be I think i think what happens over time, we'll have to see how consumers what what consumers adopt, what this evolves into very early days, but very promising.
01:24:21
Speaker
Okay, okay, interesting. ah Wouldn't banks have asked you to give this as a white label thing within their own app? Like wouldn't ICICI want within their own app this kind of a,
01:24:32
Speaker
oh feature or something? or So it's all about, it's currently about distribution, right? Today, you have the distribution. Yeah, we have the distribution.
01:24:44
Speaker
Okay, okay, okay. Interesting. And how do you earn on this? Like, like you would get some sort of a ah commission on the amount which is disbursed? Or like, how how would it work for you?
01:24:56
Speaker
It's a simple fees. It's a fee-based transaction with the banks out here. like a fixed piece uh per uh per transaction okay okay okay interesting so what are the ah opportunities now so you answered the 10 crore question you answered the 100 crore question i'm guessing now the question in your mind is the thousand crore question right like how do you hit a thousand crores so what are the answers that you have in your mind that these could be ways to hit a thousand crores
01:25:29
Speaker
I think 1,000 crores is something that, as I said, right? So for for hitting 1,000 crores, what we believe is that this model, we just need to continue building out on this model right now. um Acquire a lot more merchants, work with a lot more merchants.
01:25:46
Speaker
There are some segments that we are entering into things like food and grocery segments, et cetera. So there's a lot more marketers remaining. We will start trying our hands out with offline as a segment.
01:25:59
Speaker
So there's I think there is, I mean, in India, we're talking about 300 million consumers, right? ah The market opportunity is is extremely large. And there's so many micro segments that one can work with. And we kind of will continue to build for the consumers across each of those categories and segments.
01:26:22
Speaker
And by when do you think you'll hit 1,000 crores?
01:26:26
Speaker
i think if we do it correctly in another two to three years, three years, we should be able to do it. Wow. And is an IPO on the cards then? So IPO for us is more like, I would say, requirement you know so to
Investment and Capital Raising Strategies
01:26:43
Speaker
scale further. If we need to raise more capital is where we would look at IPO.
01:26:47
Speaker
Because if you need to grow at 50 to 80 then then we would need to raise capital. right as As soon as the growth tapers down to probably about twenty twenty five percent is where a need for capital will go away.
01:27:02
Speaker
So IPO will either be for it will primarily be for raising more capital because most of the investors have come in with us over a period of time. right They want to stay with us for the next 10 years.
01:27:13
Speaker
They believe that the business can be much... I mean, oh i mean from 1000 crores, it should ideally grow it to another ten x
01:27:25
Speaker
So, you know, from a series A fundraise to a series B fundraise, how do the questions of the investors change? Like in a series A fundraise, what questions are you answering? And in a series B fundraise, what questions are you answering?
01:27:39
Speaker
Sure. I think in the series fundraise, it's more about, you know, product market fit, right? Whether, I mean, I think the way we have always looked at the entire business model, right? We look at this is this This was someplace I dread probably about seven, eight years ago.
01:27:58
Speaker
You always look at four pieces of the model. So you look at the product. I mean, I would say the market, the product, the channel and the business model. ah Now, if you look at this entire quadrant, right, if you look at, say, a market product fit. okay So is this product, um does the market want this product? right That is the first thing we're trying to understand.
01:28:23
Speaker
The second thing that needs to work for a series A is what we also call as a product channel fit. Okay. So your channel of distributions could be different, right? For example, we could be distributing this product online on the checkout with the merchants, or we could be distributing say a con with, with the business correspondent network, right? The BC network that we said, uh, people in the villages who are selling smartphones to the consumers, right? That could also be a channel.
01:28:52
Speaker
Now, both these channels are essentially for the market of you know selling smartphones, but the markets are slightly different and the channels are different. so you need different products for them.
01:29:04
Speaker
So in at the series A stage, you're looking at product, ah whether the market that you're going after is that market clearly defined, is the market big enough, and does the market really want your product.
01:29:16
Speaker
And the other you're looking at is that ah can is Is the channel of distribution large enough so that your product can scale up in that channel? And is there a product channel fit? Is the channel adopting your product or not?
01:29:31
Speaker
So every channel would need a new product. So there's a series, yeah I mean, a a bit of it, right? so Now, once your series A has come in, right? or Once the investor comes in and says, okay, I see that this is working.
01:29:45
Speaker
I need more confidence on whether the business model can scale. That's where now the model comes to the business model. So now we are looking at something called as a channel model fit.
01:29:56
Speaker
Okay. So what do you mean by channel and the model fit? So we have market product channel and now the channel and the model fit, right? What is happening at the channel and the model fit is that, say if your cost of acquisition is extremely high, right?
01:30:11
Speaker
ah Say like a SaaS business where cost of acquisition in some cases where you're dealing with large enterprise clients, it can be high. So are you making enough revenue and enough profit from one single client?
01:30:24
Speaker
And what is the retention for that particular client? Or is it a one-off project that you make enough margin? So it justifies your acquisition costs for that channel. ah Say someone like a Facebook, right? It's, it's, so I mean, you just do a network effect. Your cost of acquisition is very low.
01:30:41
Speaker
You're essentially getting your business model, which is advertising, but your your your cost of acquisition is extremely low. So again, it can scale up. Your, your model can scale up in that channel.
01:30:52
Speaker
So it is, it is typically over here where the big death really lies is where your cost of acquisitions is somewhere in the middle.
01:31:03
Speaker
It's not high. It's not low. Your retention probably is not very high. So that is where the biggest issues then start happening because now what we call is the lifetime value to CAC, right?
01:31:14
Speaker
That starts going for a toss. So once you've hired your Series A, the biggest piece is to see that what is ah the channel of acquisition, your business model or your way in which you're making money. Are you enough making enough money to get get rid of your entire CAC?
01:31:30
Speaker
How fast you can do that. That is the channel model fit that is very important to look at in series B. And the third piece for series B essentially is that, okay, can this model scale up for that market? That's the last piece.
01:31:45
Speaker
Is the market large enough? And will this business model or the way you earn earn money at at all at a large scale, is the market willing to give you that money?
01:31:58
Speaker
So getting that comfort is very important because now everything then if all of these four pieces work, then there's no stopping to the business. And that's when a series B or a series C capital comes in because all these questions are now answered.
01:32:15
Speaker
Amazing. Amazing. That was a super lucid explanation of the Series B lens for a founder. ah By when do you recover your cost of acquisition? Like the first time someone takes an EMI product, does it cover your cost of acquisition or is it when he comes back for the second time or third time?
01:32:35
Speaker
It typically takes three to four months because once the second or a third repeat happens is where we start to recover the money. Do people ah repeat based on the SnapMint name, or is it driven by the merchant? Like if I am regularly buying from, a say, WakeFit, if I'm going on the WakeFit website and buying stuff on a regular basis, and so I'm doing transactions again and again because of WakeFit, I don't necessarily remember that SnapMint was the provider. Is that the case? Or do people know, OK, SnapMint was the provider? at like like What is the brand awareness ah that people have?
01:33:12
Speaker
ah So i think it takes a about a couple of transactions for the consumers to be completely brand aware. But once they are aware of the brand, and they then they do keep looking out for Snapman. And in fact, we give them, now I mean, we also give them a list of brands where they can actually buy with us from, which helps merchants also then increase the sales.
01:33:32
Speaker
Okay. So that flywheel is there. Okay. Okay. And when someone is taking an EMI from Snapman, do they have to download the app or they're just setting up a UPI mandate and...
01:33:43
Speaker
app download is optional? We typically set a UPI mandate up if they want to repay, but okay ah there's no necessity really to download the app. So is is it something that you want to push going forward, like to have a lot more app installs so that the credit on UPI product, ah for that you need a large app install base, right?
01:34:08
Speaker
Sure. I mean, over i mean what we've seen typically is that if you keep giving enough value to the consumers, they land up downloading your app, you don't really have to push the app downloads.
01:34:20
Speaker
um But then that enough value needs to come for the consumer. So i mean, that's the reason why we always go zero, zero percent a by zero processing fees, zero late fees, right?
01:34:31
Speaker
If that value comes into the consumer, then starts working with you. So your marketing spend is essentially your product, right? I believe you don't actually need to do any other marketing spend other than have a solid product.
01:34:44
Speaker
ah We do marketing spend. Solid, we don't do marketing spends. And the end entire idea of marketing spend, obviously, is that the consumers need to know that we also work with a lot of brands, work with them ah different merchants.
01:34:56
Speaker
So there is some marketing, so I mean, there are marketing spends that are done, especially many you times when some of our brands, they want to launch new products. We want to work with the brands to ensure that their launches are successful.
01:35:09
Speaker
So we do work with them on a lot of these campaigns. So like like say an iPhone campaign might have EMI by SnapMint in it and you sponsor part of the cost of that campaign, something like that. Yes.
01:35:21
Speaker
yeah okay Okay. Okay. Okay. Interesting. Amazing. Amazing. Thank you so much for your time, Nalin. I learned a lot in this conversation. Super insightful. right. Thanks. Thanks, Akshay.