Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Building The Collections Stack | Anshuman Panwar @ Creditas Solutions
454 Plays2 years ago
When it comes to innovation, debt collection is a space that has lagged behind. Spotting this gap, Anshuman decided to build a SaaS product to radically disrupt traditional collection methods. By using data analytics, Creditas Solutions helps people recover debt seamlessly on their own terms. Listen on as he talks about how AI and ML are reshaping the debt collection process!
Know about:-
- Challenges of traditional lending practices
- Ethera, world’s first SaaS-based ‘Neo Collections’ platform
- Why AI-driven collection system is useful to banks
- Roadmap for 2025
Recommended
Transcript
Introduction to Anshuman Panwar and Credit Us Solutions
00:00:00
Speaker
Hi, everyone. This is Anshuman Panwar. I am the co-founder of Credit Us Solutions. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour. Take me on a tour.
Transition to Fintech and Innovations in Debt Collection
00:00:35
Speaker
I remember way back when I joined my first job, I started getting calls that would offer me a lifetime free credit card. Today you would hardly get such calls. Instead, you have the new age fintechs like Slice, Unipay etc that offer you a modern user experience on your card and they use online marketing methods to get users to sign up.
00:00:55
Speaker
This evolution in how businesses engage customers is happening across all sectors, be it ordering food or buying groceries. But one space which has been lagging behind is the space of collections, or basically, what happens in case you don't pay your EMI on time? Many companies continue to use old methods to collect debts that are highly people intensive.
00:01:18
Speaker
This is the opportunity that Anshuman Panwar spotted and built Creditas to solve. Anshuman is an IIT-IM alumnus who worked in the collections business for global markets before deciding to build a SaaS product to radically disrupt traditional collection methods and his bet paid off. Today Creditas is a profitable business which has raised very little funding and has some of the biggest banks and NBFCs in India use their product.
Insights into Debt Purchasing and Profit Models
00:01:45
Speaker
Listen on as Akshay talks to Anshuman about the collection space and how Creditas is on the path to be a 500 crore top line business by 2025.
00:01:53
Speaker
I got an opportunity to join this company called Ballador Capital. I am a co-founder at Ballador. So this is way back in 2009. So we know each other from 2009. So this was a business which was essentially purchasing highly distressed consumer assets in the UK market. What that essentially means is they were going to banks and NBFCs or in the UK market and saying, hey, Tomare concert, defaulted, personal loans, credit cards, customers are not paid for six.
00:02:22
Speaker
months to one year, two years, these portfolios are with you, so we'll buy these portfolios at a deep discount, right? So they were essentially a debt purchaser. What is the economics of that business? Like for example, you buy a hundred, let's say there's a bucket of loans which are worth one crore, you'll buy them for what, like 10 lakhs or something like that?
00:02:41
Speaker
10%? No, no. So our average at that point would be about 5%. This was the UK market and the prices were going down. So the beauty is on one side, you have these banks who are looking to sell these assets, troubled assets or toxic assets, so to speak, or these pool of accounts, retail accounts.
Understanding the Yield Curve and Debt Recovery
00:03:00
Speaker
These are not corporate loans. These are individuals like you and I who have deployed, so to speak. Personal loans, mostly.
00:03:07
Speaker
personal loan, credit card, auto loans, the entire gamut of retail loans, trade as such. Everything except housing loan, everything except mortgages, everything with the retail side. So what happens in adverse elections is that you have these banks who have these spoons of accounts. And on the other hand, you'll have these purchases like us, which will come in. Typically what happens is a bank will at any point, let's say has 100 crores of such a house, they themselves will break it up. They'll say, okay, these are accounts which are in the UK market.
00:03:35
Speaker
Let's say they start saying that, okay, we will start selling at a whole stage where the guy has not paid for three months, for example, or let's say six months for ease of this thing. Six months is my first branch, then there'll be 12 months, then there'll be 18 months, and then there'll be two years. So what the bank itself does is it says, okay, I have this 100 crore pool, which has just entered, not paid in six months as it did there.
00:03:59
Speaker
at that point they say okay a part of this pool I will sell the rest I will try to keep collecting on my own then at 12 months again I will send a bit of it again at 18 again at 24 right so that everything either person has paid me if it's not paid then by 24 months I will sell it off to someone else
00:04:15
Speaker
So what happens, the beauty is the debt purchaser, whoever is coming in and say, this guy is only six months old of North payment versus this phone, which is 21 months old. So he will say, okay, the six months phone is far better to collect from than a guy who's not paid for two years. So I will pay a premium one. So I might buy it at 20%, for example. And this phone of 24 months is also available. I'll pay 5% only because obviously there has been squeeze out of it, so to speak.
00:04:42
Speaker
So, that's the sort of idea that has been squeezed on. So, you pay 5%. How much do you recover? Like you paid 5 lakhs for a 1 crore loan bucket. How much would you have recovered from that?
Digital Transformation in Debt Collections
00:04:53
Speaker
What would be the cost of recovery? So, and therefore, what would be left for you? So, from the economics of the business, every portfolio you buy has a yield curve essentially, right? What is equivalent or what
00:05:04
Speaker
to figure it out was that essentially your yield curves don't move much. If you bought at 5% or 20%, a 5% will maybe yield you 20%, a 20% will yield you a 30%. That's it. That's not as if you're getting 20. So 20, you'll get 84X. The movement is not that much. So effectively, the key element to this is the price you have paid. That defines your profit and bottom lines.
00:05:32
Speaker
So it's like a factor analysis, like 20% of those accounts would give you 80% of the revenue. So you would want to figure out which are those 20% and just focus on them and let the other 80% just like not waste too much time. Send a letter, no response, forget it.
00:05:49
Speaker
Yeah, absolutely. The idea is just brilliant because in collections, really what happens is most people view collections as a problem of morality. Unfortunately, that is not the case. And these are all collections teams everywhere in the world. We look at collections as an issue of morality, whereas it is essentially, or they'll try to solve for morality of the customer, so to speak. What they need to solve for, in our view, is solve for friction as such. Make it as frictionless as possible.
00:06:17
Speaker
Even in an Indian context, if somebody has taken a loan, if you owe the bank 1 crore, it is your problem. If you owe the bank 100 crore, it's the bank's problem. The idea is if somebody has taken a loan with an express intent to run away from it, then it's going to run away from it anyways, from hire, help, or hell or I, so to speak. Second also, at the time of lending, that's how lending works.
00:06:42
Speaker
After about four years of working in valid or what we realized, so to speak, was that two or three things. First of all, we realized this is a humongous market, extremely big market, which unfortunately, because it is unsexy, nobody focuses on.
00:06:57
Speaker
And I will say it's a repute or reputation is also well-founded. Those guys behave like that, so to speak. And it's a huge market once. Second, it's also, despite being a huge market, it's a market wherein the ad width and use of technology has been extremely low.
00:07:14
Speaker
Right. They still rely on calling and having customer interaction, so to speak. Whereas even seven years back, you would have all these concepts of business models coming on. How do you lend to customers completely digitally and on? Whereas on collections, it is always understood that in the time we have not
Efficiency of Digital Interfaces in Collections
00:07:35
Speaker
basically mouth of two choices of we are invectives to the customer. That is how the collections is done. And maybe little part was thrown there. So at some point about back in 2014, let's start something of our own. We explored a lot of ideas and obviously because we had this experience of even if it was for about four years in the UK market,
00:08:02
Speaker
You say let's do this in the Indian market and with the technology kind of a thing, we can build out technology for banks and we can go and pitch it to them that instead of doing this entire piece using the manual, field force or telephony, you can do this with
00:08:20
Speaker
system, so to speak, which is interacting directly with the customer, it's far better. And then the idea of that initial state was very simply that from a system perspective, first of all, that entire conversation, right? If, for example, there's a debtor owes money to the bank, if he's talking to a call center agent or a collection agent on one side,
00:08:45
Speaker
versus if it's interacting with a sort of impersonal sort of sanitized screen, the conversation is completely different, right? On one side there is obviously a human emotion somewhere to play and all kinds of things happen, on the other side you can have a much easier this thing and it's not as if Akshay will break his laptop or his phone, right? He gets that angry, it is what it is and you can have a much better this thing. Second,
00:09:11
Speaker
is the fact that if you look at collections as such, ultimately, it is all about or traditionally has been about putting people right, whether it is field boots on the ground on a field force or going and knocking on people's doors to collect money, or the call center agent, which is again putting a homebody on a telephone on the seat with a telephone attached to call up there, right.
00:09:33
Speaker
That has an issue with scale. Whereas if you have a system which automatically lets you send through brain, they're sending out millions of SMSes, a billion of robocalling, emails, Facebook, etc. Right? You're not limited. You can essentially reach out to your entire base within a minute. Whereas with people, you just cannot do that.
00:09:53
Speaker
Right, so there are efficiencies which get built in terms of scale. And third, finally, I don't know, actually third idea there would be that because it's a system, right, as opposed to people, their cost should also reduce that simple from a digital perspective.
Challenges and Strategies in Digitizing Collections
00:10:08
Speaker
The fourth way, which is the most important is if you look at collections,
00:10:14
Speaker
the world over. Okay, like I want to give an analogy here. There was a time when you and me had just come out of college into our first jobs, we would be getting these calls for credit cards that do you want a credit card? There was that whole boom period of companies trying to give credit cards to everyone, which now today's world the way companies sell credit cards through micro targeting
00:10:37
Speaker
and there are different cards being sold to different kind of audiences and there is email marketing and SMS marketing and it is basically the process of selling a credit card is no longer human driven, it is digital driven basically. So that same approach is what you wanted to bring into the collections industry.
00:10:58
Speaker
Yes, you can say that, except fundamentally, the difference being if you're trying to sell a personal loan to someone, digitally versus trying to collect money digitally are two different propositions altogether. The latter is far tougher, right? Absolutely. I'm not trivializing the challenge here, but just a
00:11:18
Speaker
analogy to show the evolution basically from that human driven cold calling based approach to a digital approach. Yes, absolutely. I think the biggest obviously challenge there was for people to look at that even from selling stuff.
00:11:36
Speaker
digitally or banking digitally, right? There were always presidents in the West who were doing that. And then we had players in India. I'm pretty much follow them. Or even before that, you always had the likes of Amazon and all these sort of... That was the promise of digital that you can sell stuff better. Whereas talking about collecting stuff better digitally was something that was somehow jarring to people. As always, if you have to collect digitally and from telling your own customers, then the guy would not be telling quite upfront.
Building Partnerships and Market Entry Strategies
00:12:06
Speaker
But yes, and then we can talk a bit more about how that helps, how that journey evolved and what all we learned. From day one, was it this exact thesis to be pure digital, not have human telecalling process and all that, or did it evolve over the years? So I think as with any business, there was an evolution. But yes, I think the fundamental concept was pretty much, as you said, to build it as digital as possible.
00:12:33
Speaker
The idea was obviously to understand the market as such as well and try to pitch. And then obviously, it wasn't the easiest thing to do. So essentially, we at some point, we pretty much adopted what we call as the hobo strategy. We pretty much parked ourselves at the bank so that the guy will say, okay, in order to get you off our backs, here are some accounts that you can start working with. Obviously, the bank said, hey, what is the bottom of the barrel? And then Dick,
00:13:01
Speaker
a little deeper and give the most of the most accounts to you just to get us off their back, so to speak. And that's where we started. Was it the same deal that you bought that or was it like a service? I told banks that outsourced collections to us for a fixed fees business. Yeah, it was an outsource or it's an outsource sort of a...
00:13:25
Speaker
Purchasing debt requires massive capital invite to get an ARC license. There are norms, 100 crore plus you need to have balance sheet. So the idea was that, okay, fine. Essentially what we're trying to do is effect the technology change and the change as such. So here's a piece of technology. We will use it. We go to banks. The banks will say, okay, there is a bunch of accounts. Please go collect on them using your method. I don't think
00:13:50
Speaker
care, whether you use field force or all on them, or if you're saying you have this new piece of technology and magically people will start saying, okay, let's see what happens, what doesn't happen. So we pretty much started with our first client was HDFC bank. And as usual, they will, as I said, say, okay, what are the worst of the accounts because that has the least risk, even if it doesn't collect, it's okay.
00:14:14
Speaker
And then they will keep digging further and say, okay, there's something even worse than that. And they will give us those accounts, except that we started performing, obviously, right? Again, it comes out to liquidation curves, et cetera, right? If you get a six-month-old account or a pool of six-month-old account versus if you get 10-year-old account, the liquidation rates are going to be different.
00:14:35
Speaker
But as long as the banks know what the liquidation rates are, as long as you perform, you perform, right? And then you try to go up the value chain. So in these initial days, the sort of business operation was simply banks would give us a bunch of accounts, we would work like any agency. So one of the things is banks on collection sites rely on agencies quite a bit, right?
00:14:55
Speaker
They outsource a lot of it because banks also are regulated by RBI. They have their norms. If they have all these really low cost resources on their own books, there are issues. A lot of the things are outsourced. Collections tend to be big outsourced things.
Personalized Messaging and Customer Engagement
00:15:12
Speaker
So sales are spent by PSA's that don't exist. So we paneled ourselves like an agency because that was easiest. They said this already exists.
00:15:22
Speaker
is a phone sign-in. After that, I don't care what you do, what you don't do, as long as you're not abusing people or breaking the kneecap, so to speak, which anyways was not the model. So what would happen is the bank would give me, let's say, 100 accounts. I will get those accounts. We will do a bit of branching. We'd figure out that here is a customer A, who is, let's say, 30 years old, took a loan, now the credit card at 22, defaulted by 25, 26. It's been four years into default.
00:15:50
Speaker
I can figure out that he's in a job, might be looking for a, has just got married, might be looking for a house or a car. So what I would do is here's a segment to which I will say or send out an SMS and keep it very simple that hey, Mr. Customer, do you know a poor credit score will impact your chances of buying that dream house or buying that dream car, so to speak.
00:16:12
Speaker
And the messaging will be very, very neutral. It'll be very sealed and sealed messaging and speaking to a customer's particular need. Nowhere am I saying that, hey, Mr. Customer, I know that you owe money to HDFC buy, come here and resolve it. The idea is not that. The idea is too for the person that some innate and inherent need or some trigger has been triggered that he clicks on that SMS, lands on a webpage. That way he would sort of have a journey there and will show him
00:16:41
Speaker
that hey guess what this is your credit report he will log in take out his credit report and say hey guess what this is an HDFC account which has not been paid for five years and it is bringing down your credit score and your eligibility to get any access to any new credit line so to speak.
00:16:58
Speaker
We were, by the way, the first ones in all of India to tie up with the bureau. So we were the first company and now I'm sure everybody does it. But way back in 2015, we were the first ones to the deal with Experian and Bohan. So we were the first ones to open this credit report directly to customers, which a lot of other businesses do. Pretty much every other business does it.
00:17:26
Speaker
If you like to hear stories of founders, then we have tons of great stories from entrepreneurs who have built billion dollar businesses. Just search for the founder thesis podcast on any audio streaming app like Spotify, Ghana, Apple podcasts, and subscribe to the show. These other businesses which are doing it right now are doing it for risk underwriting or as a sales funnel approach.
00:17:55
Speaker
Lastly, on a sales funnel approach, you will see the likes of Pessa Bazaar, Back Bazaar, like get your free credit score and when someone fills that form, then they sell them stuff. Yeah, on that and then you also have grade and all which is also somewhere.
00:18:12
Speaker
accessing your credit report because you tick on a thing. So that has evolved over a period of time, obviously. The interesting bit was that we go to the bureau and say that essentially I only want to get credit scores of those customers or give credit reports to those customers who have a delinquency. If the guy is absolutely clean, he's absolutely useless to me. Whereas for all these businesses, that is the one that they want. I'm actually looking for the garbage of it. The bad customers is what I want.
00:18:41
Speaker
which was interesting at some point. A particular website went viral as well. We weren't even marketing. People were coming in to check out their credit reports. A lot of good customers were coming and we had to go to the building and say, please stop this functionality or at least don't let us wait because this is an absolute waste of resources for us.
00:19:01
Speaker
So that's where we started. We started with one bank slowly, slowly, increased to two banks, three banks. Initially, banks would give us the worst of the accounts, then moved up the value chain, built out technology, got confidence on this that, hey, this is working perfectly
Financial Growth and SaaS Solutions
00:19:19
Speaker
well. But one aspect of that business was still
00:19:23
Speaker
something similar to even I will say digital lending, so to speak, even on digital lending, if you click on it, etc. What you will see is 50% of the customers will come and avail of that loan or credit card completely on its own, but they are always drop off
00:19:38
Speaker
and stuff because people find it difficult. So all your PESA bazaars, bank bazaars, all these companies will have a call center, we'll reach out to them and guide them through the process, so to speak. So we had to build that also up on the site. So we will have customers who will come. So if a bank gives me let's say 100 customers of which 20
00:19:59
Speaker
let's say log in there and complete the journey and have not. So for those days, that's an opportunity loss for us and the bank as well, right? So we built out a team of what we call as ironists, so to speak, who are essentially guiding the customer or I will say counselors who will counsel the customer that hey, this is an issue.
00:20:18
Speaker
It's your choice ultimately what you want to do or not. We don't have a collections approach so it's that you must pay. It's more of educative and you try to basically maximize that funnel in conversion, so to speak. So we did that for about three, four audios and at some point we said, hey, this works. What kind of numbers were you doing? About who were collecting or what kind of revenue were you generating?
00:20:43
Speaker
How many customers were you collecting from year-on-year just to show the growth? Oh, yeah. Okay. So first year was as usual, I would say, I don't know, five. The first year was three months of operation, which is pretty much no revenue. First year would be, what, 30 lakhs? Then pretty much tripling it year-on-year. The thing of the business has been for us that revenue is about collecting. Revenue I'm talking about. So the,
00:21:09
Speaker
Yeah, the way we have run our business has to be focused on revenue, right? The amount collected is the function for it. We've done it very traditionally that our business has to be self-sustainable and we have not raised too much money. I think by Angels, we raised about 5 CR in all. Angels, friends and family. We've been in existence for the last eight years now. Of which the last three years have been profitable. Last year we did about the PCR revenue.
00:21:37
Speaker
with a healthy, I would say almost 40% bottom line. We are growing year on year, hopefully two weeks year on year. So yeah, it's been self-grown, except we took some money initially. Yeah. Okay. So that's the business financials, but coming back to the business and business journey.
00:21:54
Speaker
So about, I will say 3-4 years back, I will say sometime in 2018-19. At this point, we had about 10-20 banks. What they realized was here is a piece of technology that works perfectly. But there was a bit of a problem in the sense that what is happening is I'm still dependent on banks. The banks are giving me sliver of accounts. I'm working on them, reaching out to them digitally.
00:22:17
Speaker
on my own website.
00:22:32
Speaker
enables this from a technology perspective. Remember, as we spoke about personalization, there is a lot of personalization and ML that goes into there. The more accounts that you feed in, the more sort of feedback loops you get that have put 120 have reacted by, have they reacted to which messages, what are they doing? On the website, we have other funding law, what is their behavior?
00:22:53
Speaker
So you're limited by having a small pool of accounts, but the minute I get, and there's an entire piece of the technology is in sort of back-end intelligence, so to
Neo-Collections and Global Scaling Challenges
00:23:01
Speaker
speak. So the idea was, can I give this intelligence directly to the banks? Say, why don't you use it? Why table it? Have the customer. The beauty of that is, two, three things. From a bank's perspective, it will suddenly end up
00:23:13
Speaker
I will say being able to deploy this technology on its entire base, on its entire portfolio, which is cheaper, easier, and from a customer perspective, instead of me sending out an SMS, they might use essentially reintegrate for the banks and also the banks.
00:23:29
Speaker
themselves will be sending out that SMS so the authenticity is better, the person, your engagement with your customers will be higher, they will land on your webpage essentially. So they land on the bank webpage itself. And we said, hey, so instead of keeping this technology with our sim,
00:23:45
Speaker
just white label this box, give it to banks. And this is the concept that we call is neo-collection. So we were pretty much the first ones and from what I can see, the only ones in the world were doing this at scale. So currently we have about, I will say about two crore accounts across banks, which go on this platform. Each bank has
00:24:06
Speaker
its own platform, it's called Vitera, this new collections platform and essentially the idea is the bank holds the platform, it will put each and every account on it. So the idea on new collections is take this piece of technology, give it two banks, it's a platform, goes across, products across,
00:24:25
Speaker
delinquency lines. The beauty of it is that essentially on the background, there's a brain. And because more accounts are coming in there across different banks, across different portfolios, across different customer segments, a PN, a credit card customer versus a two-wheeler customer, there are two different segments altogether, completely.
00:24:46
Speaker
or a consumer durable customer versus a PL customer. And then with these sort of algorithms become much stronger and they can predict and just much more neater than they gain more intelligence, so to speak. So yeah, that's the idea. That is what we're seeing. We're seeing tremendous uplift and tremendous interest with banks. At some of our partner banks, we are seeing collections of
00:25:11
Speaker
the road and a half, two crores daily coming in on the platform, on a daily basis also coming in, which obviously for years back, in a month, five years back is what we would do in a month across high bands. In one mag on a day, if you're able to do that, it's brilliant. So there's the next, I will say, journey and evolution of that concept, what we call SDO collections. I believe it's going to become the standard for us. What we have seen, we're getting a lot of inbound interest also.
00:25:40
Speaker
on the platform with banks wanting to adopt it. I think banks themselves have this entire piece of digitizing. They themselves are also facing a challenge from a lot of, I will say, new businesses or startups, which tend to compete with them in some sense for customer attention, whereas here's the platform that we are essentially giving to them, right? And the customer, credit card's brand name is absolutely in the background, right? Even if you land on the webpage or whatever,
00:26:08
Speaker
In the bottom right-hand corner, it should be better than covered by credit also to speech. The idea is the bank can be providing the technology to them. Got it. Okay. So I'm going to recap my understanding. So in 1415, the approach was, or rather, this is the approach of a service business. Like you have two business lines. One is a SaaS business, where you white label it, and second is a service business, where you charge banks of fixed fees for collecting for them.
00:26:35
Speaker
So in the service business, you created a clear my dues as a consumer facing website and anybody with whom you had to collect, you would send them a message and you would try a variety of messages. You would do AB testing and see what kind of message works best. And you would do that personalization probably emails also I'm guessing. And yeah, absolutely. Yeah.
00:26:55
Speaker
and IVR calls also absolutely everything or any digital means of reaching out to customers i was just using a very simple example of this but you would use all these tools yeah so so you used all these digital outreach methods to get him to
00:27:09
Speaker
land on clear my dues and enter some identifying information so that you know who this person is. And then you would present him a civil report and you would also nudge him to clear off that amount which is due. So sorry, there are certain nuances, but I will not go into them. But yeah, you can figure out who the customer is without having to write out stuff, things that we've been out over a period of time, but somebody lands.
00:27:35
Speaker
Okay. The links would be customized. So the link would tell you who the customer is. Yeah. So the connection where it's not, isn't there always like a negotiation that happens? Like if someone says, okay, there's 10,000 rupees, which is outstanding for me at HDFC bank loan, he may want to
00:27:55
Speaker
He may be the kind of customer who can pay it off, but there might be someone else who may not be able to pay it off, and he may be able to pay off only $2000. How do you digitize that negotiation element in collections? Yeah, so you have options presented to the customer and you also give him an option for him to set up his own payment, right? So if it's 10,000 rupees,
00:28:17
Speaker
There might be one option that I always hear 10,000 in one go, or I will set you 5,000 in one go, or I will pay 2,000 in five months, or 1,000 in 10 months. Or you give him the option that you give him basically two, three options. And then if he likes right on the screen, you can have him create his own option, which he finds absolutely easy for him to deal with, which is all good and well. The problem is that essentially from a customer experience perspective,
00:28:45
Speaker
If the customer has landed there, you want to give a resolution term right then and there. This is where it becomes tricky and which is what we've been able to achieve that we've integrated with the bank engines at times. And if those engine settlement engines don't exist, we build the settlement engines. Because unfortunately, as opposed to
00:29:04
Speaker
the sales side or the digitization of sales, which as you said, would have seen as existed for last eight, 10 years, right? Then the customer comes in, types of stuff.
00:29:16
Speaker
Right then and there, there'll be an integration with Bureau to 424. Online KYC, everything online. Everything will be auto-packeted. Within everything done, I say, the thing is approved. Alone is approved. Whereas, unfortunately, the process would be okay. He wants this. It has to then go to the central team, to the concerned manager, who at the end of the day will look at what settlements exist.
00:29:42
Speaker
Then he'll get an exchange sheet or 500 settlements he'll have to basically manually set an accrual, right? And that is a very... So you're not solving for morality, you have to solve for friction. If there's a logic there, just build that logic. The customer is saying 10 months, 1,000 rupees and the bank's logic, and they provide us with the logic. If it matches with it, good. Goes ahead, if it does not. For example, an easier example might be the customer owes 10,000, says, I will pay 3,000.
00:30:10
Speaker
but the logic in the back end is that he can pay 4000 days what the bank wants to go to minimum right so at 3000 then you say sorry this cannot be done but if you pay 4500 this will go in and you give some payment arrangements the whole point right now which i kept saying initially as well you're not trying to solve for morality you have to solve for friction remember
00:30:30
Speaker
Yeah, and I know it's a bit, was the word creepy or I don't know what the word is. That time is money, so to speak. It's an IRA. Why do you want it? The customer, you have customers mind space with you, right? Now close it, then it there, right? Digitally finish it off. If the deal is on the table, just do it.
00:30:46
Speaker
to the best of your this thing and that's what digital tools and technology can enable as opposed to you know someone physically talking to the person then going to his boss seeking an approval it will come in two days then going and saying this is now approved please start paying stuff like that it's just
00:31:03
Speaker
building a lot of inefficiencies and persist. So once you reflect a pipe, which has a lot of inefficiencies, you just clear all those inefficiencies, the amount of water flowing in, flowing out of the pipe suddenly increases, right? Yeah. And that's what we are seeing. How do you make sure you don't leave money on the table? Because every customer would want to pay as low as possible. Now, why would they want to pay off the full amount? They would just want to pay whatever is the least that they can pay and get away with it.
00:31:28
Speaker
Yeah. So one of the things that we also do as a matter of fact is when we integrate with the banks on your collection requires actually quite a lot of integrations. One of the things that we pretty agree on is that the customer will be treated fairly, right? At which point in time, there are a lot of grids that come to us in the settlement in Jilbanda. The major ones for us is
00:31:51
Speaker
This is very peculiar and most people and even your viewers nobody will understand or know about this. Everybody says let's do a settlement. Unfortunately what a settlement does technically is marked as settled on your credit report which is a negative
00:32:07
Speaker
sort of remark. So what happens there is that if you're marking it as a negative remark that you might say, okay, it is settling, it's fine, I'll be able to access new loan. But what we are seeing is a lot of times banks when you come back and say, no, I will not give you that mortgage, especially your housing loan or not to loan until unless you close this completely to settle right now.
00:32:26
Speaker
So we get those settlement credits and we get those closure credits. From where we started, 7-8 years back when it was very simple, straightforward, one web page. What also happens is we'll have about 300 web pages at a bank, like 300 different journeys. Within the same customer base, a different customer, looking at their profile, you will see something completely different from
00:32:48
Speaker
customer B, let's say, right, a customer A's journey in terms of workflows, what he's seeing will be very different from customer B as well. And that's where all the personalization comes in. And it has obviously taken time, you know, as they say, Rome is not built, was not built over the day. And I think Rome is not built in this case, it is an ongoing journey, you have to keep building on it. As one of the beautiful things with technology,
00:33:14
Speaker
is it keeps evolving, whether it is under the hood technology or even something which is visible to you. You keep seeing new modes of communication which keep coming up to interact with customers. How do you change yourself along with that? That's what the Terra or this platform does for the banks, right? It's a SaaS platform. So it's not... See, traditionally how banks have operated, they said, hey, this is a piece of software I will buy one time free. End the story. Buy, buy data.
00:33:44
Speaker
The beauty of this platform essentially is we say, okay, you buy it, it's fine, take it. Your investment upfront is extremely low. Buy revenue is SaaS-based. Essentially, even in SaaS-based, it is going to be actually success-based. So it's a SaaS success-based fee in the sense that it will be based because collections are all okay, fine.
00:34:04
Speaker
If we are paying an agency 100 rupees for an account, a little bit at 10 rupees is a big deal because that aligns me as a business along with the bank. Otherwise, I can do one size thing, but once the pressure is on us also from the revenue perspective, then you keep evolving and improving the product. That's one side of things. The second side of things is that for me, my upside is not gap. When you do a success base, the better the product has, the more money you make.
00:34:34
Speaker
parties are aligned in the correct right direction as opposed to just selling it one off and then forget about it. So your SaaS pricing is on percentage of amount recovered or like number of recoveries done or like how do you price that? It's a mixture across.
00:34:52
Speaker
different portfolios, different products. So I think it is not fairly standardized. I think as an industry practice or very fairly uncertain, I will say products and vintages that can be fixed on certain, that can be more variable as opposed to a range of collections. So it's something which is pretty standard. But again, even on fixed, the per account fixed kind of a thing as opposed to one account as paid 10 rupees versus one account as paid. And the car paid 1,000 rupees, then maybe I charged 10% and make 100 rupees.
00:35:21
Speaker
That's sort of something like that. For a bank, SaaS would be significantly cheaper, but the only thing they don't get is when your analyst intervenes and does hand holding. That's the only thing we say don't get in SaaS. Everything else they would get, but it would be significantly cheaper.
00:35:42
Speaker
Absolutely, yes, extremely cheaper. So what they have also now said that a lot of customers are coming on this platform, but not making payments. Can you handle them? And we've said, no, this product is not that. So they've now said, hey, we also have our own call center.
00:35:59
Speaker
Can you pass those leads there? That is perfectly fine, because it's a Sashford and I pass those leads to them. But on a Thera side or near-direction store, we do not have to know physical, human touch at all. It's a pure Sash business. There's no service. We have a service business arm on the side that's when CMID use that.
00:36:17
Speaker
operates slightly differently, this as a business is different. But yes, there are drop-offs which happen, banks are concerned about it. They simply say it's a SaaS, it's a technology person. If you know where Naxhe is, lumped in right now. In real time, I know what he has done, not done. So I can actually integrate with the call center and their IVR and just have the IVR place to call, but not us. You would create a workflow for the call center agent also for him to see the
00:36:46
Speaker
what's happening in all of that, like a webpage for him to also see all of that information. Yeah, absolutely. So there is a banking side of things as well in terms of when the customer is interacting, what is happening, how many. For example, again, I will use very simple example that you send out 10,000 SMSes. The entire brain is which SMS is working, not working.
00:37:10
Speaker
So all of that is open and available to the bank to look at. And with intelligence, actually it was a step further. A bank has, we have this proof, proof, templates and all. So the bank can either say, Hey, I want to use this template or the system itself essentially says, but guess what? This template will work best. This is the course of action that we recommend.
00:37:34
Speaker
Do FinTech companies also use this? So as a business model, yes, we have now just started to look at FinTechs, but primarily we are pretty much with all the large banks in this country. Because as I said, we are largely self-funded or angel-funded or bootstrapped. So the idea was, as Einstein will say, I would look at the toughest problem and try to solve for it. And that's what we have done. We have gone up to the biggest banks because the volumes are the largest
00:37:59
Speaker
there. And now subsequently, we are looking at our day fintechs as well. Fintechs, the volume is what the revenue potential tends to be slightly lower, but I think there are certain larger fintechs in there as well, especially with BNBL. BNBL as a concept, if you look at it, is absolutely economically unviable to even put a call center on those small ticket size loans or whatever you call as micro ticket size loans. It just makes no sense.
00:38:27
Speaker
You can do it, but if you're doing it, then there is something fundamentally wrong. If you're putting people, they have bigger issues there. If that's the choice you have made.
00:38:36
Speaker
What is the onboarding journey for a bank? Like if a bank takes this up, then how do they get onboarded? Like you do some integrations with their existing system. So like, how do you make them live? Is it like self-service or is it like a lot of your tech team has to do a lot of integrations and all that? Yeah, unfortunately banks in BFCs or BFCs are not
00:38:59
Speaker
the most technically, I will say advanced in the ecosystem as such. So unfortunately, it does require technical integration or these technical integrations tend to be, I will say, easy. It's a three-week process from beginning to an end kind of thing. So it's easy for the banks, except that it just takes three weeks. Ideally, I would want to
00:39:22
Speaker
to it three days a lot of it can be done but it ends up taking i will say three weeks maximum we have to integrate with their core systems as well because and the reason to integrate with the core system is very simple it is about customer journey right so if an akshay is making a payment right now it has to hit and reflect in this account within next two seconds right and that's a journey we have sort of gone through
00:39:47
Speaker
show me that we have undertaken over a period of time. But yeah, it requires core system, core banking system integration. Some banks have APIs available, a lot of them do not. Which case we have to assist the bank with that as well, to build out that integration, or we can do without the integration as well, as you say, okay, let's go live in three weeks. And as phase two of the project will be
00:40:09
Speaker
If I take two months to do that integration, then the customer experience will only improve over a period of time. But yeah, the idea is three weeks. Most of the... And I'll say, there are some very basic, important hygienic integrations which get done within a week, max two hacks, ideally. There are others which are from more customer-delight perspective, and that can happen over a period of time. So if a bank is using a modern core banking system which has API and all, that would happen in three days, I'm guessing.
00:40:39
Speaker
Yeah, I wouldn't say that. I wish there were banks between that. Although there are so many fintechs also targeting banks, like giving them better gold banking systems and also I guess it's just a matter of time before the banks digital infrastructure is at par. Yes, absolutely. I think from a startup perspective, what we are seeing is a lot of new banks come into place.
00:41:05
Speaker
But they are also, they are riding on a banking license. Banks' ability to license, but they're saying the customer is ours. The journey that we're providing is very good. Building a better, beautiful wrapper around it. But ultimately, end of the day, they're saying the customer will be ours. In our case, we're pretty much outsourcing the technology to the bank and saying, yeah, take it. Fine. Zero goes to the bank to take it. They only be on success anyways, right? So it's easier for them also.
00:41:33
Speaker
Zero cost, even if you build them a settlement engine, like you were saying some banks don't have a settlement engine. Absolutely. What is a settlement engine? It's just like an algorithm to decide who gets allowed to settle how much amount. Yes. It's as simple as that. It's like an Excel sheet kind of a thing. In a sense, very simple. On an Excel sheet, you have a hundred. You set some rules.
00:41:57
Speaker
Yeah, I just had some rules that side, which yeah, a lot of them have. Unfortunately, a lot of them have it in macros or actions. Some of them still divide by your mistakes. Some manager has to do something, but generally analytics teams and all these guys have it. And with the bureaus, they have it. We are just making it easier to use, so to speak. So you said that the more, the number of banks who are using it, the smarter it is getting.
00:42:21
Speaker
So how is that data being shared back with you to make your system smarter? What is the way in which banks would have a privacy concern also, right? Yeah, absolutely. So from a data perspective, essentially there is no closely identifiable information which is shared. So basically there is a space of software which is deployed at the bank. All the information resides there.
00:42:46
Speaker
Then there is a sort of small engine which is outside of this. So the only information that goes from here to here is very specific account level and not personally identifiable information. So that will be like your name will not go, your address will go mass rather. The only thing in the address that will go is maybe a penny board.
00:43:09
Speaker
The pin code will go because that sort of makes the engine much more smarter. Your balance will go, but no name, no unique ID will go. There will be a separate unique ID, which will be generated. All the interactions that are happening here in some sense are captured, but all are basically the faces sort of thrown off. You can't see the face. There is just stream of data which is coming in. So you can figure out that these kinds of accounts, which are working like this and back. We have also got those kinds of accounts.
00:43:37
Speaker
So this is the treatment that is working here best. So to start with, let's use the same treatment. So that's what the injury is deciding, right? But there is no information movement between banks that happens, so to speak at all. How automated is this? Like for a bank, do they just need to upload? Not even upload, maybe there'll be some integration.
00:43:57
Speaker
Once the integration is done, then does it automatically run everything, every campaign or what is the amount of human intervention needed to run it? It depends upon the amount of integration that the bank has done, right? Some banks are far ahead of it. Some banks are slightly behind, but in steady state, I will say it's absolutely
00:44:15
Speaker
Now, let's say for a fully integrated bank. Yeah, it's a click of a button thing. You just go live with it and suddenly the manager will see money coming effectively. Okay, okay. Single overview of an intervention. As I said, you click a button and then money comes in. As simple as that.
00:44:33
Speaker
That clicker button is for just that one time going live. You need to just click a button for that. Yeah, essentially, yes. A lot of it is automated. The only thing is... After that, there is no manual intervention. After the one-time go let. Yeah. From banks, there might be a little bit of manual intervention, but from our side, no. So the more they automate the process, it is far easier to click off a button.
00:44:58
Speaker
File transfers, because these are large banks with lacks and lacks of customers, so it's not any DPI, will not work for that. Huge data will not work on DPI, so there's an SFTP automated transfer that happens, but yeah, it is really simple. The bank can stop it, and that's the only limited dimension that they can do. Ideally, they shouldn't, and they don't.
00:45:20
Speaker
The system is becoming better and better. That's how it's working as banks are developing more confidence on it. It's essentially a tick of a button. That's it. After that, then it is essentially okay. I think from a platform perspective, the platform is there. It is about making more use cases. Different use cases beyond collections has been on it, which actually banks themselves are coming off the payments. Can we do this also? Can we do that as well? Which is brilliant for us.
00:45:48
Speaker
For example, you can do sales, cross-sale on it. Right now, banks still cross-sale, upsell, will still use, essentially, a call center. They don't have the ability to target customers. Most of the targeting that we spoke about is done by third-parties, like a pass-up of banks, banks who have profitability. They don't even need to. You end up spending a whole lot of marketing dollars if you're a third-party, because watch who doesn't want.
00:46:13
Speaker
This is for the bank if you think i want to customers and i just want to know now they're saving accounts and i want to sell credit card on that i have any information on that i just have to reach out to them so they would pick the use of all simple missing here is a system which is working on collections across an upset should be even easier than you will.
00:46:31
Speaker
with things. So yeah, stuff like that. Or for that matter, I will say on credit cards, change behavior to say, Hey, I want to increase the usage and spends on my credit card. Can we use this platform to communicate and engage with customers? One of the key things is obviously driving engaged with right on the platform of the customer. It's
00:46:53
Speaker
And that concept at Atonscore can move across different use cases, not just in collections, for sales as well, for cross sale, up sale. So you hit 30 CR top line last year, you said, told me. And this is all like bootstrapped, like you've not raised any funds so far. No, life to date. I think friends, family, and the angel would have been five CR, what, including our own money that we put in. Yeah, so about five CR these two.
00:47:20
Speaker
We did about it. Last year we did about it. We see a lot. Give or take a few. There was a lack. And this year, what do you estimate? That you'll double it. So this year has been more of building on the thera side of business to scale that side of business. What we realized is the integrations that Clinton banks are taking longer than what we anticipated because of the banks' things. So I will say about
00:47:46
Speaker
It should be around 50, sort of a number, not really double. But the beauty is I think the next year should be again, massive growth. Not even 2x, 2.5x, 7c, which is fine in business. They're always upside down. So as long as you're going profitably and sustainably, it's fine. Like you just hit some bumps, which is okay. And you don't need to raise any funds even to achieve that 2.5x next year. You don't need funds.
00:48:11
Speaker
And absolutely not. It's more about working capital management, etc. So you don't really need to raise funds, so to speak. Scale this up completely. My hornet zone is when you can raise money to 2.5x, do 3x, 3.5x. But money deployment also, see, it's a cash generating business, profitable business, right? There are always opportunity to deploy funds and it becomes more of a working capital thing. And even the fund
00:48:56
Speaker
So like by 2025, what kind of
00:48:59
Speaker
revenue do you estimate you'll be doing end of 2025 like that's about two years away three years three years yeah so in three years time i would be looking at anything north of 500 three years annual revenue date i should be there that is i think there's the scope of business is high and double triple i think
00:49:16
Speaker
right now is more of, I will say, deployment phase of the platform, right? To maximize, as I say, sort of land grab for free is for us, right? The more you deploy, revenues follow through it, right? And then you build up efficiencies, et cetera, of the platform across banks. Right now, it's about making as many deployments. The focus, actually, to be honest, has not been so much on revenue also. It's been more about adding these banks
00:49:45
Speaker
It's more so than to focus on driving you at this stage, which is not used to be the case till a year back. As I said, we have different cycles. So this is more of, I will say, a capital cycle in a sense for us. So in that 500 crore in 2025, how much would be subscription? How much would be service? 80-20 or 70-30.
00:50:08
Speaker
Yeah, subscription is a much bigger play because it's cheaper for a bank. Yes. That also, I think the service business is, as I said, there is an element of people there, right? Of having massive call centers. So to scale it up and run it, it's not a very high margin business, but it's a cash cow business, right? So if we keep running, it's a choice you make, how much you want to grow it or not grow it.
00:50:34
Speaker
Ah, it's a choice. Essentially, you can load it to 200 and that's 400 CR here. So 500, 600. These are choices that you make. And these are not choices that you make right now for three years hence. These are always truly right. So you can make the choice for this year. This is what I want. And then next year, we again, revisit and figure out how do you want to grow, how much resources you want to spend inside of the business.
00:50:59
Speaker
And do you also plan to go global? Absolutely. So as a matter of fact, we are in talks with a couple of banks in Southeast Asia and Middle East Asia as well. So hopefully something in the next two, three months should rectify with a couple of banks outside of India. That's our initial foray to understand, to start and then build on it again.
00:51:18
Speaker
And there, obviously, the margins, the revenue potential is much higher because of the currency being stronger and the ticket size of the loans being higher and all of that. That's what we have heard. Once we start with a couple of banks, we'll have a better idea. The general wisdom is that dealing even B2B, anything in India is a very cost conscious, I think, society as such.
00:51:45
Speaker
So even what they pay for a particular service in India versus something that they will pay in Middle East or South East Asia will be different. The difference in India is that, okay, maybe they might pay you 100 rupees, whereas let's say in Singapore, they pay you 500 rupees.
00:52:01
Speaker
But in India, that 100 can be paid across 120 crore customers, which is the population of India. Whereas in Singapore, that 500 is across every 1 crore or whatever the population of Singapore is. So yeah, those are trade-offs. But yeah, we will see what happens.
00:52:17
Speaker
But yeah, we're going to go like a couple of banks soon enough. So would you also be looking to enter US? Yes, at some point, yes. We are out actually in talks with a couple of players there. It is something we are exploring as well. Again, at the very exploratory stage right now, it's a huge market. It's a market where dollar value is also high and the customer value is also high, right? Which is the largest economy in the world. I suspect it will have its own challenges. And I think one of the things
00:52:47
Speaker
that we are realizing is that any market that you enter, you need to have a local presence, so to speak. And so I think there is one other company, which is in this space. And I just wanted to ask you, what is the difference between you and them? There's this company called Crenjenix, right?
00:53:05
Speaker
Like so, so what is the difference in approach between you and CredGenix? Okay. So as far as CredGenix as a business is a bunch of very smart folks have come together to again, try to solve this. There are a lot of similarities and there are a lot of differences as well. They are again, trying to solve the collections problem, but these out there have taken is one of the legal side of things.
00:53:31
Speaker
how to send out notices digitally, et cetera. That's the sort of route that they have taken. It's an interesting business. They are extremely well funded by, I will say, long term patient capital. Bunch of web guys have not met them. They're also tele-based by Alma, neither as well. So, yeah, that's a very similar, also slightly different way of looking at business.
00:53:59
Speaker
If you like the Founded Thesis podcast, then do check out our other shows on subjects like marketing, technology, career advice, books, and drama. Visit the podium.in, that is, t-h-e-p-o-d-i-u-m.in for a complete list of all our shows.