The Warren Buffett Playbook From Rs1 Insurance To Rs800 Crore Acko Founder Reveals Insurance Secrets

You are making 800 crores by selling a one rupee insurance. Acko is like the Berkshire Hathaway of India in 2026. The first rule of an investment is don’t lose. And the second rule of an investment is don’t forget the first rule, and that’s all the rules there are. I mean, that if you buy things [music] for far below what they’re worth and you buy a group of them, you basically don’t lose money. You know, in the 1960s, Warren Buffett made an investment that everybody thought was a boring investment. He didn’t buy a factory, he didn’t buy an oil company, he bought an insurance company. An insurance company that was bleeding with 127 million dollars in losses. That insurance company until today has made him 19.8 billion dollars because Warren Buffett spotted a very unique gap and capitalized on it through a game-changing playbook. And Varun Dua today is building an insurance company in India by taking inspiration from Warren Buffett’s playbook. And in this podcast, he breaks down all the inefficiencies in the boring industry of insurance. He also speaks about why the insurance industry is a boring but a gold mine industry to build a business in, and he also speaks about how did he sell one rupee insurance to make hundreds of crores. So, although insurance is a boring subject, you must watch this podcast because it is absolutely game-changing. Most people will ignore it, but that’s precisely why this podcast is game-changing for you. Most insurance companies in India selling to consumers [music] in the garb of life insurance is crap. If you were to pay one lakh rupees on insurance not expecting anything back, you could probably get 15-20 crores worth of coverage. How did you solve this gatekeeping problem? And win your customers with very less acquisition cost.

What experience we can deliver to the customer really depend on who owns the customer. We’ve done ads with octopuses, with Javed Jaffrey talking nonsense. We had a Munna Bhai trying to explain health insurance. Culturally, we’ve changed the nation only. I’ve used it only in investor meetings. An Apple user is 20% more likely to take a claim than an Android user. Take a claim, why? Tata, Bajaj, ICICI HDFC. Do you have a preference between the top 10? No. Therefore, if I have to influence the distributor to sell my product, I have to pay him more. The rise of AI, are there gaps that you see opening up for new entrepreneurs who can sell to you today? We need cyber [music] insurance. Nobody in India, including us, knows how to price it. Eventually, the core business is going to be car, health, and life. So, are you telling me the insurance industry is broken? Before we jump in, I want to thank Odoo for supporting this podcast. Now, listen to this very carefully, especially if you’re a business owner, founder, freelancer, or just someone with big dreams because what I’m about to tell you could be the difference between closing your next big deal or watching it slip through your fingers. Let’s be real, guys. In today’s fast-paced world, a single missed follow-up can cost you lakhs. That is why you need to know about Odoo CRM, the tool that’s changing the way India is doing business. So, whether you’re managing 10 leads or 10,000 leads, Odoo helps you track every lead, every deal, every call all in just one place. With sleek Kanban view, you will always know what’s moving, what’s stuck, and what needs your attention instantly. And with Odoo’s chatter system, you will never lose an activity or an email or a client update ever again. But wait, it gets better. Once the deal is closed, you can send a beautiful quotation right from the opportunity tab itself. And with custom dashboards, revenue forecasts, and real-time insights at your fingertips, Odoo isn’t just your CRM, it’s your business command center. So, if you’re building the next big thing or even just getting started, this is your sign to stop winging it and start winning with Odoo CRM. And here’s the best part, your first Odoo app is free for life along with hosting and support included. So, hit the link in the description, sign up for Odoo, and take full control of your customer game starting from today. Varun, in this crowded market, why should I build an insurance company? You keep saying that insurance is a gold mine business, but where exactly is the gap in the market to build an insurance company? Super interesting. I think I I think we I’ll try to first simplify the math because people once people understand the math behind it, you’ll you’ll realize that it’s a it’s a beautiful business, right? All of us understand return on my money, right? Like itna banega. Maine ek mutual fund mein dala, 12% ban raha hai, 25% Wherever I deploy my capital, what is the return on that capital that I’m generating? So, let’s do simple math. Um because it’s a risk business that tomorrow you have to pay somebody back in an eventuality that may or may not happen. Uh there are black swan events possible, right? Pandemic was a black swan event. So many people in the hospital. Or you take Every year, we have a mini black swan in Acko also. For example, every city in India once some big metro has a flood every year. Thousands of cars go down together. We have to pay for all those cars. So, that means your balance sheet has to be strong to sustain. Roughly in India, the regulator asks you to keep 25 to 30 rupees for every 100 rupees that you want to collect. So, if I take I I sell a policy to you and you give me 100 rupees, mere paas 30 rupaye alag se reserve mein hona chahiye. This 30 rupees I can’t spend because it’s reserve capital. You know. So, what do I do with that money? I It’s not like I have to keep it with the government, I have to keep it with me only. So, this 30 rupees I invest. So, it’s largely debt funds, government securities, FDs, you know, stuff like that. So, if you look at that, you will get a return profile roughly of 8%. So, 30 rupees pe 8% 2.4 Okay, 2.4 rupees the company has generated. Okay. Now, 100 rupees is your premium that you’ve collected. You have three costs. Okay, acquisition cost. Yeah, commission doge ya marketing doge jo bhi doge, there’s some cost to acquire the customer. Then you have claims cost, what you will pay. Okay. And then you have G&A or your operational overheads, salaries, office, AWS, technology cost, whatever it is. Your general operating expenses. Running cost. Running cost. Okay, let’s assume 100 comes in, 100 goes out. Maan lo 30 rupaye acquisition mein chala gaya. 60 rupaye claim mein chala gaya, 10 rupaye mera operating mein chala gaya. Okay, 30 60 90. 100 in 100 out. Okay. But here’s the catch. You pay me 100 today upfront. Okay. Claim will happen over a period of time. Right? My acquisition expenses are not also not happening today. So, you have given me all the money upfront. Claim you will do after 12 months. So, what do I do with this money because all the money I don’t need on day one. Okay. I invest this again. But because money is going away slowly, you’re not able to make 8% because 100 mein se kuch kuch kharcha hota ja raha hai. Okay. So, how much has the company totally made? 5 rupees plus 2.4 rupees. How much capital did the company put in? 30 rupees. 30 rupees. 7.4 rupees you’ve made. 30 rupees. 25% 25% That’s the business. You’ve made a 25% return on your capital perpetually by doing 100 in 100 out. 25% is crazy. That’s what the business is. That’s what is called return on the return on equity on this business if managed well. If managed well, now here’s the underscore. Let me give you another startling figure. This 100 in 100 out, no nobody in India is able to manage. It’s all 100 in 105 out. 100 in 105 out. Yes. So, are you telling me that the insurance industry is broken? Yes, it is. But why? Because there’s competition [clears throat] on distribution. So, why is So, acquisition cost sabka badhte rehta hai because sab wohi agent ko wohi dealer ko wohi bank ko zyada paisa dete hain. Tu mera maal becho, tu mera product becho, tu mera product becho. So, let’s say if you let let me ask you as a consumer. I don’t know whether you own a car or you don’t own a car, but if you own a car between the top five 10 names in India, if you had to buy car insurance, Tata, Bajaj, ICICI HDFC, do you have a preference between the top 10? No. There’s no pull therefore. You don’t have an opinion why Tata is better than ICICI is better than ABC, right? So, who decides that? The car salesman. The distributor. Therefore, if I have to influence the distributor to sell my product, I have to pay him more. More people because as a consumer, nobody said that, you know, nobody has built a proposition in which Ganesh or any customer saying nahi mujhe toh yahi chahiye. You know, I want this. For every their price is lower or service is better, whatever, there is no unique positioning in my mind. So, what you’re doing is basically you are forced to continuously go to the car dealer, car salesman, whoever it is, and say achha company X itna de raha hai, main 2% zyada commission de dunga. Tu mera maal becho. So, all of these insurance companies together, they keep increasing the commission cost. I’ll give you a startling number. I joined this industry in mid-2000s, early 2000s. I have a 20-year career only in insurance. When I joined, 15% commission milta tha car insurance pe. To car dealer, agent, whatever. Today, it’s 30 to 35%. It is doubled, tripled almost in the last 20 years. Why? Because the industry is commoditized. Industry is commoditized, more players entered. Car dealer is the best place to sell it. Jao commission badhta hi gaya, badhta hi gaya. 15 ka 17 hua, 20 hua, 22 hua, 24 hua. So, today for every 100 rupees that you’re paying, almost 25 to 30 rupees is kept by the car dealer or So, one, it creates pressure on that 100 rupees that we were talking about. Ke 25-30 rupaye toh wahin chale gaye. So, the product also becomes expensive for the consumer because you have to accommodate a fatter uh commission. Then the second part is claim cost kab kam hoga? You want claim cost to be So, you don’t want fraudsters in your system. You don’t want people who have already banged their car to come and buy insurance. You know, who are insurance kharida nahi, thuk gaya toh abhi uh le liya. So, you want this is what is called quality of underwriting. When you’re underwriting something, you’re saying you want your best office there. That’s what you’re trying to do. But we haven’t matured as a market. So up to Maruti Suzuki and Maruti Suzuki if you both of us have a Baleno bought in the same year will get the same price. However, you may be driving from Juhu to Andheri and I may be driving from Vashi to Nariman Point every day. Same car both of us own. I am three times more likely to have an accident because I’m just driving three times, you know, more than more than you are. But if you own the same car the insurance company today gives you the same price. There are also other other issues that happen. People who are generally in financial stress if they have debt, if they have some some issue has happened in their life, also misuse insurance to try and get money back. They create fake accidents, do all of this because it’s a source of getting some large amount of money. So if you don’t underwrite well, which is you don’t select the right customers well, then claim cost is high. A distributor can have distributor about the car dealer. Go. The car dealer up to you some granular company needed. What about the Baleno payment is commission above Baleno match around by criminal channel many money channel area. They have nobody has an idea. So does the cost keep going up? Yeah, it’ll go up because it’s scale is linear. A bank through which you go. Let’s say so bank arms are selling bank branch managers selling. They are also tied up with three four insurance companies. So how do insurance companies actually ensure that the bank sells their product? Good, bad, ugly for the customer that is a second order. More margin. Uh you know, more margin. So more margin and you’ll say a child training killer look Malaga Dunga. I I How do I build preference with my distributor? I have to do something more for them. How do I do something? I’ll say okay, my guy will you know, come and train your people. I will do one, you know, your annual offsite culture I will do. I will do this. Really? Yeah, yeah, all of this. All of this. It’s like FMCG companies giving gifts to retailers. 100%. 100%. Including foreign trips. It’s worse than that. The desperation for distribution is so high that all sorts of practices go on to be able to sort of influence distribution to to sell your product. In this sort of an environment where your costs are going up, your underwriting is not strong what happens is this doing this hundred and hundred out that we are speaking about getting cut shy disciplined if you were to be you have acquisition cost, claim cost and operating cost. You could run it under a hundred. But if this is the way we operate, if this is a way of business or an industry operates, it’s very very hard to keep it under a hundred. Then what happens is that 25% return on capital business which we just spoke about in the calculation becomes a 15% 16% profitable have a baby. Because interest income is generating profits but you’re like a 12 15% return on capital business. Up master up technology at you can run this business at supremely high discipline. It’s not a very high growth business because the minute you try to grow very very fast fraud comes in. You have to be disciplined. But if you’re disciplined, you can generate 25 to 30% return on capital perpetually. The market once it matures is very very large. LIC today is about 50 60 70 billion somewhere in that range in terms of the premiums that they collect. To give you a sense like there were no health insurance companies stand alone specialized health insurance companies till about 10 years ago. Star care Niva Bupa all of these companies have come in the last, you know, like 10 12 10 12 years. There is aviation insurance companies that only write aviation risk only planes, ships transport of goods offices, human lives, health sporting events. So most most people don’t know Wimbledon used to buy pandemic insurance for like maybe 20 years. Can you imagine that somebody thought that pandemic will happen and it’s a one in a hundred year event. I guess they must be paying about one one and a half million euros a year, but they got a payout of over a hundred million euros. To answer your question, India can take care of another 50 insurance company that believe in the in in in in the medium to long run. What is the penetration of insurance in India like right? Typically it’s measured as percentage of GDP. So 3% odd is premiums to GDP. That is for life insurance and it’s about sub 1% or about 1% for general insurance which is non-life insurance. LIC has taught India to look at insurance as a savings product. Correct. Uh so a lot of the premiums that life insurance premiums that India pays is not for protection. It is a It is like a between ULIP, FD, mutual fund, insurance, saving schemes, money back scheme. It is an investment bucket rather than a uh insurance bucket. So actually our people genuinely protected. If you look at that protection gap which is a different measure that how many people can say today in India that uh you know, if I die my family, their education, mortgage, everything shadi, everything will be well taken care of. That’s why you buy insurance, right? For an untimely, you know, demise. The breadwinner has an untimely uh demise. And most people don’t. It’s the smallest market within life insurance today. You won’t believe it. It’s not even a billion dollar market right now. Everything else, all the big numbers in life insurance that you hear about majority of India is not buying life insurance for protection. Majority of India is buying very wrongly life insurance as one alternative for investment between mutual fund which bank never buy a scheme I have unit link plan something after so many years you’ll get something and it has some coverage. So you’ll pay one lakh rupees you’ll get some 10 lakh rupees of coverage and the rest of money is going into some investment. If you were to pay one lakh rupees on pure pure insurance not expecting anything back because you’re not trying to make it an investment product you could probably get 20 crores worth of 15 20 crores worth of coverage which is massive, right? Like that can actually take care of your family and you should just go invest in mutual funds. So big part of the life insurance numbers of penetration and all that we see are actually not good for India. Like on paper they are good, but they’re they’re terrible. I’m I’m from the industry. Most insurance companies in India selling to consumers in the garb of life insurance is crap. Like I I can go out on a limb and say it’s it’s not good for investment, it’s not good for protection. You can easily buy mutual funds. You can easily invest in debt funds. You can make ample money. You can get enough return and you should use insurance primarily to protect against uncertainty not try to make money out of it. Up car insurance [clears throat] let you out to I bought to do you get accident near water get a bus Malaga. So why do we have that attitude both as consumers that if I don’t die why am I not getting, you know, why am I not getting something back? Warren, now can you help me understand how did you win your customers with very less acquisition cost? I think fundamental design principle for us was that you know, the economics of this business and finally what experience we can deliver to the customer really depend on who owns the customer. So we said we’ll go direct to consumer. Okay. Direct to consumer however the problem is that it’s very expensive to build a new brand, you know, and insurance is not you know, some shampoo or t-shirt which I can try some new brand, right? Like you know, you Who am I competing against? I’m competing against a Tata and ICICI, a Bajaj and HDFC, right? So why would somebody buy an Echo, you know, they never even heard the name, right? Like so D2C had that problem that it’s not a product that people are happy to design a child only layer, you know, let’s see. So it’s not a let’s see situation. Like there’s a little more commitment little more faith, you know, uh required. So that was one problem and of course building a new brand for a complicated product like insurance where the the dichotomy was no no no no. So but that was the faith that was the thesis. This is also not a brand new model. It’s been done in other countries. So I assume that where India’s maturity curve is it will be over. Can we be the ones to, you know, uh do it? And of course famously as the most famous is Warren Buffett’s company Geico, right? Like they did this 20 years ago pre-internet. Uh people had to dial in 1800 and buy insurance. You know, there was no agent. Uh so it is about direct to consumer versus being digital. Digital doesn’t matter. People used to buy on 1800 also. Warren, can you tell me what exactly did Warren Buffett do with Berkshire Hathaway? Yeah. Because what you’re essentially telling me is that Echo is like the Berkshire Hathaway of India in

No, it’s an amazing story. Almost 30 years ago 25 30 years ago they decided to go direct to consumer. And can you imagine in an era where people were not buying anything online? Uh so they used to run these large billboards and newspaper ads and Super Bowl, you know, all the sport, you know, whatever the sporting events was. Call 1800 some toll free kind of number. Their tagline or their ad campaigns I think for donkey’s years have been saying one thing. 15 minutes can save you 15%. Like dead straight simple message that What did that mean? That 15 minutes phone per bottle or up to 15 20% you’ll be able to save on your insurance. Like that’s as simple as that message but they did it very creatively. There are they’ve used, you know, outlandish creatives. There are this famous Geico lizard that they used. They’ve used characters, a bunch of things. They just made it super eye-catching, right? Like for for somebody to not notice it. And people started calling in to find out carry it Malaga. Can I save more money from what I’m getting through my agent? As as simple as that. That created it simply created that awareness. Once that cycle sits in two things start happening. One, every time you’re doing an ad you are really trying to get new people who have not called in before or not bought before. And you’re targeting your ads to people or maybe they were doing it in some cities where they were not big or some targeting that they’re doing to get new people. Old people who are there or to renew are there. On that we commission to you are not paying. So you are now like generating lots of free revenue. Year on year you’re generating lots of free revenue. They give good service, basic good service. So people are also like constantly paying and renewing and you don’t have anybody to pay. So what started to happen was that all the marketing investment started to go to acquiring new consumers and revenue free may occur. What happens to the business at a certain point in time? Let’s say the business is totally doing 100 rupees. Okay, of premiums collected or revenue. 100 rupees is coming in. After 4 5 6 years the cycle so ends up happening that if the company is generating 100 rupees 60 rupees is coming from and only 40 rupees is new because the old has now compounded for 5 years 6 years 7 years right? So the that old becomes larger than the new. Yeah. And when the old becomes larger than the new what happens to the company’s economics? 60 70% of company’s revenue is coming free. Absolutely free. And the economics suddenly change. And then you’re sitting on this big investment income. You’re like it’s just now a virtual cycle. It had a very interesting another concept. So you want good risk right? I want the best customers to come to me. But how do you assess the quality of the customer by asking them questions? You ask questions. So of course how they did it 30 years ago they must have asked questions. Nowadays the way we do it there are hundreds of factors. Like hundreds of factors in the sense that today even we don’t know what all factors we use because now it’s all machine learning. Can you give me an example? Let’s say I give you a call. Okay, or let’s say I interact with your app. You you you come up on our website or app and you try to put in your car number. So for example for the license plate you have to enter your license plate number or car number. We know your last 5 years claims history. Okay, how many accidents you’ve had and so on and so forth. We also check things like your credit score. Good financially responsible people generally are safe with all financial products. We’ve we’ve seen that. This car credit score is less they claim more. We’ve seen that correlation. Uh over a period of time. Then we’ve seen behavioral signals. So I’ll give you an interesting one which is a giveaway. Uh I’ve used it only in investor meetings. iOS users or Apple users same car same job same neighborhood same profile like everything else is the same. Let’s say you and me live in the same building. An Apple user is 20% more likely to take a claim than a Android user. Take a claim? Take a claim. Why? You know even we couldn’t figure this out but we looked at data um you know over you know millions of cars that we’ve insured now. So sometimes what happens is you find the correlation but you don’t know what is what is causing it. So it’s a typically called I know the correlation I don’t know the causation. Is it because let me guess. Is it because we are used to Apple care? No. What we have what we deduced after speaking to some of those customers and just forming a theory data backward was that they are just more vain. Okay, vanity. Like I I will not handle the you know I will not handle a scratch on my car. Do you have access to that data as well? We in fact do the reverse. We try to help consumers. We try to get challan data. We’ve tied up with RTOs and show to our customers or people on the app challan pending. You know. That also helps the consumer also helps us to you know sort of assess the assess the risk. My personal belief around insurance is that good customers should not be paying subsidizing bad customers. You know that. So because it’s unfair right? Like what is insurance? It’s a risk pooling mechanism. Like 100 people have the same risk they’re giving you money. If 20 people have very high responsible behavior it is raising the prices for everyone. Yeah. So an insurance company’s job is to keep it fair. So you should not allow misuse of the system by bad actors. You know whether it’s fraudulent or negligent behavior or whatever it is that will keep prices low uh for everyone. So Berkshire Hathaway also understood the quality of customers because of the visibility that they had. The large visibility data is developing. Now what happens is when 100 people come 50 are good 30 are average 20 are bad. Like that’s a rough ratio I’m giving. But when these 50 people are coming you who who will you offer your best prices to? The 50 who are good? You will not offer your best price to the 30 who are average or the 20 definitely not who are bad. Because you offer them good prices what happens is that 50% of traffic becomes 80% of sales because convert because you’re not giving them good prices right? Why will you? 20 last 20 to you might not even give a price saying I don’t want to it’s like in lending I don’t want to give you a loan. You don’t qualify. So maybe last 10 15 20% I am not even but because I’ve seen the whole market a large part of the market is checking prices with me the best way to understand what it does to the business it gives me the first shot at the best risk in the country. Now what that does is creates a new virtual cycle. Good customer convert already. Bad customer convert already because price already. The quality of business just keeps increasing. Good good customers keep seeing good prices. Bad customer so the proportion of good customers in your total portfolio that you’re creating keeps So your claim cost goes down. Claim cost goes down because constantly you’re giving better prices to better people. So this is the model that we wanted to you know sort of follow that hey can we become that place where people come direct they check prices and so on and so forth. So we started off fairly straightforward. There have been campaigns where we said you know save money on your car insurance very basic stuff and then we tried to sort of build it into a habit by saying car insurance like your stuff like that just trying to build that memory structure in people’s minds that just check price you know in in some sense. So that helped and of course we did a lot of deep work on insuring that the claim side we don’t falter. Most insurance companies today have outsourced the claim to the dealer. You know our dealer go to you. Dealer behalf claim file. But dealer incentive claim claim. No dealer is very incentivized to give you a claim because after bumper replacing he’s making money. Oh, got it. There is something called F N O L. Okay. First notification of loss. First notification of loss. Okay. There is a loss. Some accident has happened. Who do you notify first as a customer? Do you call your insurance company or do you call the garage or the car dealer? One of them you have. You call phone call. You call phone call inform call that becomes like what is your first notification of loss? What point? Who is who is your first notification loss? If you go to the car dealer first it is in the interest of the car dealer to inflate the repair bill as much as possible. Okay. Why? Because he’s sitting with bumper parts he’s already bought from the OEM. He will burn money. He will burn money. And you also as a customer sometimes don’t care enough because the insurance money is there. Whatever it is it’s okay. Correct. Now the dealer is filing the claim on your behalf. Only if there’s a problem does the customer gets involved. If the whatever the dealer has filed with the insurance company insurance company you know I will not pay so much I’ll pay 5000 less then you will negotiate three dealer you and insurance company surveyor will negotiate and find an answer. Easiest example I can give is there’s a scratch on the bumper. I can repair it and paint it and it will be as good as new. Or I can put a new bumper. Both are possible. Okay. The cost difference is 5000 versus

Now and you don’t actually need to replace that bumper but the dealer will advise you to do that. You know and so on and so forth. That’s the difference that it can create in your claims cost. So now in case of where does the first notification of loss go? 60% of our customers come to us don’t go to the dealer. Okay. You know and that’s why we are able to control the When they come to what happens? Like I said they press a button. We understand what happened. You know there’s a little bit of form they fill or they talk to someone on the phone. Uh in the vicinity whatever workshops that we have which we are aligned to could be OEMs also not necessarily our own workshops mix of OEM workshops multi-brand workshops own workshops like I said we’ve started doing our own workshops. So we will show some options to the customer depending on distance quality all of that. Customer picks we take it to the workshop we repair it and we bring it back. So you make the primary decision and ensure that the customer has three best options. Yeah three four good options. There are some customers who are sticky about no I want to go to OEM only. Do you also tell the customer what is the repercussion of choosing an expensive repair? Yes. So every time they set up a claim we tell them how much they will lose in no claim bonus what will be their out of pocket expense for this repair at every type of garage. That’s amazing. What you also sold 1 rupee micro insurance policies right? What was that about? So I think when we were looking to own customer relationships there was uh and we said we’ll do ads we’ll go direct to consumer we’ll try and build a new brand. Uh there were two angles that were staring at us. One was of course it’s going to take time you know for people to trust buy experience us. 100 will buy but only if 10 15 will experience you claim I got so the cycle of you know sort of building this experience is going to take take take much longer. That was one angle that there was going to be time taken on the strategy. The second angle was there were there were a lot of sort of these hidden opportunities that we were seeing that small small things were not getting solved by insurance companies or even if they were getting solved they were not getting solved well. Uh so let’s say if you were taking an Ola and Uber and you had a small accident in it, the way Ola and Uber used to operate because of the scale that they operated Rosie Kushner Kushner who gave like it’s it’s going to happen every day, right? Just at the law of averages some small accident will happen every day. There would always be a and especially if the customer got hurt, you know, there’d always be a dispute. You know, who has to pay cab guy’s fault, this one’s fault, and you know, so on and so forth. Imagine an Amazon Diwali sale or a Prime Day sale or whatever. A million, 3 million, 4 million mobile phones are getting sold and we’re selling 4 million insurance policies in that day. You know, so it would happen at like crazy scale. Uh and you have a few of them, you know, maybe 5 10% of them having a claim, actually having a broken screen. So when they got the 3 4,000 bucks easily, you know, the conviction to go and buy the bigger product, you know, sort of started to come. So I think it helped in multiple ways. And that became another distribution hack. So while while on one side we were building out the direct-to-consumer brand, it got accelerated by this huge infusion and exposure, you know, in a in a in a meaningful way. If we were if you were not doing that, also that business makes money and today is my second largest business. Yeah. Yeah, it’s because health we started off a little late, but it’s a fairly large business. It’s a I think it’s over a $100 million business right now. You are making 800 crores by selling a 1 rupee insurance.

Not Not all of them are 1 rupee. They could be anywhere I I don’t think anything They would range from 1 to 1,000 rupees maybe. Okay. Yeah, like different kinds of products. That’s crazy. Yeah. Yeah, that business has been profitable for 5 years now. Amazing. All of this is based on an assumption that I will price my insurance policy right. Yeah. So how do I crack pricing model? Risk businesses. I think lending is a risk business. Insurance is a risk business. I think these are these are risk businesses are very different businesses than regular businesses. And I think in the insurance industry all the numbers of lenders sometimes that go risk is not very well understood. I think the definition of risk business whatever you think can go wrong will go wrong. Okay, like that’s the definition. You cannot assume you don’t know you have to You know, like this is not possible. Nobody thought a pandemic was possible. Nobody thought that Iran like Dubai will be getting bombed today. Like everything is possible, you know. So from that perspective, not assuming goodness but assuming the worst is a inherent skill that the company needs to have first before you get your algorithm right, you know, and and and all of this right. I’ll give you a best example. When COVID came up, partly the government, partly the regulator, partly the insurance companies said that let’s sell COVID insurance. Make sense. We were a 3-year-old 2 and 1/2 3-year-old company. And all our business had gone top. Because we used to sell only two products at that point in time, car insurance and travel insurance. Okay, nobody was driving a car, nobody was traveling anywhere. So we we had gone to like practically zero or like 20% of the business was left. And we were in a huge stress because costs were there, there was no revenue. To launch COVID insurance because it’s selling like hotcakes, right? And consciously we said we are not going to do it at the cost of not having any revenue. Okay. Okay. And the reason we said is like I don’t know how to model it. I have no idea today whether 20,000 people are going to get sick, 20 million, or 200 million. I have no idea. You know, so what seems like you don’t product banana you to big tiger is the most flawed thing that you can, you know, do it do an insurance. Let’s apply this to car insurance. Mhm. Simply put like there is a frequency of accident that what is the possibility that you’re going to have an accident. I am let’s say I have a 10% chance of having an having an accident. Let’s say you have a 5% chance of having an accident. But how do you figure that out? That’s where the data comes in. That’s where the model like I said, you know, Apple customers iPhone customers likely to claim more, somebody is driving 20 km, somebody has had more claims. I look at your last 5 years of history of owning that car, how many times have you claimed? And so so behavioral data, all of that put together you get to some confidence scores. I don’t think you can like model at the, you know, you cannot be 100% accurate, but you can be 95 98% accurate to get to us in a band. So you will say, “Okay, Ganesh probability of his car having an accident looking at his history, looking at his behavior, looking at how the cars he’s used in the past is probably around 8 to 10%.” Like that’s the But you come to Varun, you know, like I had some claim I had you had You’ll model this guy at 15 to 20%. You know, like roughly. But when you’re starting out, let’s say in 2018, how did you model these data? No, we didn’t have it. We just went with broad averages and watched the data play out. We used some surrogates like we knew from global knowledge from Geico and whatever we had seen play out that credit score is a good indicator. So we started using credit score just to see whether it’s mapping. What does the 750 750 score mean in terms of, you know, accident risk rate and so on and so forth. So this is typically called incidence rate. What is the probability that you’re going to have an incident, you know. So like I said, incident into cost of repair. If you multiply these two, 10% chance that you will have a 20,000 rupees car accident. 20,000 rupees worth of repair that accident will generate. Okay. That is how you start the modeling. This tells me that 10% chance that you will generate a 20 is like your risk price is 2,000 rupees. Okay. Roughly, like there’s a little more complicated math, but this is the simplest way to 10% chance of generating a bill of 20,000 rupees, which means your risk price is 2,000 rupees. Risk price is 2,000 rupees. Yeah. So my policy should be priced at Roughly around 2,000 rupees. Like rough There is one a couple of more maths that is involved in it, but this is this is this is the simplest way to This is if I don’t want to make any margin. I have other costs to pay. This is your let me COGS. COGS. This is your cost of manufacturing the product. Okay. Top of that I have my expenses, I have acquisition costs, and you know, and and so on and so forth. I have buffers for a bad for for floods and and so on and so forth. So you can model a little bit of your more cost on like what your acquisition expenses are, how much you need to, and maybe this goes up to like whatever 3,000, 3 and 1/2 thousand, something like that will go up to. But this is like figuring out what your COGS are going to be. Okay. You know. Or this is a break-even point. Or this is like this This is your gross margin in the product, right? Like if you can figure out what your point is. This is how you typically model it. The more data you have, the more sharper you get over a period of time. Then Ganesh is not 8 to 10, then I’m able to get Ganesh to 8 and 1/2 to 9. Like the The narrower I’m able to get, the sharper I’m able to get, the better it, you know, starts becoming. The more I am have confidence on whether the repair that you’ll generate is 20,000, that is coming from Oh, I know I have solved my paint cost and bumper supply chain cost. I know 20,000 can come may 19 mega sector. Because I started solving for my, you know, garage operations efficiency started to come there. These two sort of keep feeding into each other. On one side you keep selecting better customers, on the other side you keep lowering the cost of repair or at least ensuring that they’re not going crazy with inflation. Understood. And after that it’s just a function of time and the variables that you collect. Yeah. Okay. The insurance industry is being disrupted by players like Echo. And in the next few years because of digitalization and AI, this space is about to get disrupted. At the same time, our per capita GDP is about to explode and we might hit $10,000 per capita GDP in about 5 to 10 years, which presents a very large opportunity for two types of players. Companies like yourself and companies that can actually help companies like Echo embrace technology and embrace efficiency because like you said, your entire business is about being efficient. My question [snorts] to you is today because of the rise of AI, are there gaps that you see opening up for new entrepreneurs who can sell to you today? Sell to Echo today. Yeah. What can they solve for you Yeah. and build a business? So I think one of the hardest things and I’m going to go back to the in the insurance business is getting our risk right, underwriting right. Uh because you’re not taking a risk of you’re not doing a simple calculation of 10% chance of a 20,000 rupee worth accident. You know, the the calculation are much more complex. Uh so and the calculations are um dependent on a lot of variables. Let me Let me give an example and why that’s a hard problem to solve and somebody who could solve that we would adopt their systems, you know, or adopt that solution. So typically health insurance a family comes to buy. Very rarely an individual comes to buy. They do, but more Everybody in the family has different health status. We bought healthy, yeah. Get to go diabetes, yeah. We You know, like it’s just a it’s just a mix, you know, in a family you’ll see different, you know, conditions for for everybody. That’s one problem. Okay, the second problem is you have limited to negligible to limited health data. Which is to say that I cannot scrutinize every person in that family and go and tell them to take very detailed lab tests for everything. You know, I can either ask you some questions [snorts] just to understand your health condition. I can ask for some past doctor reports. Or even if I send you to the lab or do a test at home, I can do limited number of tests because there’s an in in I cannot do an MRI also. I can’t do a cancer test also. Like there is only broad level some quick stuff I can do. Otherwise the cost of actually selling the insurance will become so high that I’m not able to sell only because too much friction. You’re trying to check everything which could take days. That’s the second problem. Third Third issue is this is a long-term product. Long-term product is I don’t buy for 1 year and keep changing my insurer. Health insurance people are buying for 10 years, 15 years, 20 years. So, there’s also medical advancement that is happening. Procedures and treatments that are not possible today, but will be possible in 3-4 years. There’s so many things that were not possible. My insurance has to pay for robotic surgery, which is not even been invented invented yet. Kinds of robotic surgery. That’s what the car cost will be. Like God only knows. Okay. You have to model a little bit of all of that, you know? So, creating using medical LLMs or data or sourcing of data to build in prediction models for something at high levels of accuracy. Even at today, we are developing some of this stuff on our own, but it’s going to take time. And going to take lots of data and lots of complexity to model. Today, the possibilities of doing everything with AI is like manifold different, right? Uh I think that’s one area I think uh at least we are continuing to grapple, continuing to work with. Of course, we are making our own efforts, but if somebody can build, I think it’s a large opportunity. Okay. Okay. Varun, I’ll present the summary of our entire episode and you correct me if my understanding goes wrong anywhere, okay? Firstly, we spoke about why insurance business is a gold mine business. Even though it’s a boring business, it’s a gold mine business. And you broke down the math for me and you said that you need to have 30 rupees of reserved capital to get started if you were to receive 100 rupees from your customers. Now, this 30 rupees is something that you can invest in the market and generate close to 8% interest. And the 100 rupees that you have, you get 100 rupees from your customer, but that is not supposed to be immediately disbursed. So, you can keep that capital and invest that capital and generate another 5%. And here’s where the math becomes interesting. You practically invested 30 rupees and you’re getting 7.5 rupees interest, which essentially means you’re generating a return of 25% on your capital. Yeah. And then we spoke about what is a breakdown of 100 rupees. And you mentioned three important costs, the cost of acquisition, cost of claims, and running cost. The reason why customer acquisition is a big pain for insurance companies is because they’re largely dependent on distributors, distributors like the car salesman or other distributors, because of which the customer acquisition cost has gone up. You mentioned that the customer acquisition cost for car insurance policy has gone up from 15% to 30%. And this is not CAC that you pay once, you have to keep paying CAC every year because the market is commoditized. It’s commoditized because of two reasons. Brands do not have USP. And number two, insurance in general is such a afterthought that nobody takes it very very seriously. So, it’s very difficult to establish differentiation in this market and establish trust with the customers. Secondly, claim cost is directly proportional to your inefficiency. If you’re extremely inefficient, claim cost will go up. If you’re extremely efficient, then your claim cost can go down. Now, what do we mean by efficiency? Your claim cost is a function of two things. The quality of your customers and secondly, the efficiency of your operation. When we speak about quality of customers, you collected a lot of data and you figured out who are good quality customers and who are bad quality customers. And that’s where counterintuitive insights came out like the iOS insight, wherein Apple user is more likely to claim insurance than an Android user because the thesis says that they like finished stuff and they have very less tolerance towards imperfection. And here’s where we spoke about the Berkshire Hathaway story, where because of the deep visibility that they got, they were able to figure out the right customers. And let’s say 50% of the customers were good quality customers, they eventually became 80% of sales. As a result, they were able to get a very large pool of capital that they could then invest. Because claims were very low and the efficiency was also very high. And Acko is practically the Berkshire Hathaway of India because you are built on the same model, but with new age instruments. For customer acquisition, you have the digital channel. For customer acquisition, you also tried out the 1 rupee insurance model, because of which customers could build trust with you. And as a result of that trust, you don’t have to depend on the dealer. And because of this independence, you unlock several superpowers. Firstly, you don’t have to pay customer acquisition cost. So, the possibility of you becoming profitable with each policy sold is very very high. Number two, every time a car has to be fixed, it doesn’t have to go via dealer, because there the dealer is incentivized to increase the cost so that he can make money out of your the repair cost. Uh the your repair cost. At the same time, you are incentivized to claim insurance because you’ve paid money for it, so you would rather get a new bumper than getting the bumper painted. So, that’s where the claim cost goes up, because of which insurance companies are not profitable. Like you mentioned, most insurance companies, in fact, no insurance company has been able to remain below the 100 rupee point, right? So, just maybe one thing we can add in this this part. Uh so, there are there are two this 100 if if a company is below 100, like 99, it typically is called as the company is generating underwriting profit. Okay. So, so because of that, every insurance company has two income streams. Okay. One is premium minus claims minus If that was 99, that means I’m generating 1 rupee of underwriting profit. The second is my investment income, which is all the interest money that I’m making. So, it has underwriting income and investment income. If I’m over 100, my underwriting income is negative. And I’m but I’m still positive on the uh interest interest income. So, the point that I’m just simply trying to make is we can call this that if you’re below 100, you’re making an underwriting That means your core product. Forget the bad income that you’re getting. Does your core product make money or not make money? That’s the So, no company has been able to crack underwriting income as of now. No, there have been Bajaj and ICICI who on and off have been for one odd quarter 100 100.5. It’s a good score. Consistently near a month. If you see over like a 5-year journey, they might have had one quarter where they are 99.5 100 100.5. But that’s like one odd quarter. Got it. Understood. Then we spoke about the superpowers that get unlocked once the dealer is out of the picture. Like we discussed, the dealer is more likely to take your vehicle to the garage and give you a big fat bill, which you don’t have to pay for, but the insurance company has to pay for it. In your case, because you don’t have the dealer bottleneck, you can partner with workshops and ensure that the vehicle gets the best service and you also incur very low cost. And you solve that through transparency, where you become the primary decision maker. You have the customer walk through the three options that will be beneficial for them as well as your company. And you establish a win-win situation where the customer makes the decision that is in favor of both the customer as well as the company. Then we spoke about how did you manage claim cost? In fact, while we when we went deeper into claim cost, the most important thing to note is the first notification of loss. If it goes to the car dealer, then your prices will be inflated, your claim cost will be very high. But if it comes to you, that is where you can have the advantage. So, if a company has to operate efficiently, they have to ensure that the first notification of loss comes to them and not the dealer. Yeah. After it comes to you, then you can optimize your cost and eventually have the customer make the decision and optimize your back-end operations. And then we also spoke about how did you acquire customers back when you were a new player? And here’s where we spoke about your 1 rupee idea that is now an 800 crore business. And lastly, we spoke about pricing of insurance, where you said you have to assume that everything that can go wrong will go wrong. And it’s a game of patience and discipline with a bit of pessimism with your modeling so that you’re always safe. Does that sum up our conversation? Perfect. You can now do my investor IPO pitches. Thank you so much, Varun. Thank you so much. This was amazing. [music]