Episode Twenty Eight

Cruising Post-COVID:

Aaron Tan and Aditya Lesmana of Carro

15 December 2020

(past transcripts)

ALAN  0:11  
Welcome back to Indo Tekno. This marks episode number 28 of our series. Selamat datang kembali semuanya! My name is Alan Hellawell. I'm founder of startup consulting firm Gizmo Advisors, and venture partner at Alpha JWC Ventures. We are by many accounts on the cusp of a car ownership boom. A report from Ernst and Young released last month revealed that nearly a third, or 31%, of people without a car intend to buy one in the next six months. Respondents moreover cited the impact of the COVID 19 pandemic as one of the top reasons for their planned purchase. We are very privileged to have the cofounders of Carro, Southeast Asia's largest automotive marketplace, Aaron Tan, and Aditya Lesmana, join us today to discuss this and other trends on Indo Tekno. Welcome to the show, gentlemen. 

AARON TAN  1:00  
Hi, Alan. 

Thank you, Alan. 

ALAN  1:02  
So guys, to begin with, maybe you can give us a sense of: a) where you currently sit and b) what the main focus of your current activities is, maybe starting with Aaron.

AARON TAN  1:15  
My name is Aaron. My role on a day-to-day basis is to look after the overall operations of the business from a very high level strategic standpoint. I don't have any operational responsibilities. My job is to ensure that the investors and shareholders are happy, and to help guide the company from a strategy and a corporate development and M&A standpoint. So that's where I sit and what I do typically on a day-to-day basis. 

ALAN  1:35  
Now Aditya, I believe you're based in Jakarta. Maybe you can kind of give us a sense of what your day-to-day is.

I'm currently based in Jakarta. Thanks to COVID, it's a bit hard to travel. My primary responsibility is growing business in Indonesia. So my role is actually involved in building teams, supervising operations and ensuring everyone is on track to reach a target. And when it's needed, I try to see where there's a gap in the team that is needed to be filled, and try to make sure that those are being taken care of.

ALAN  2:02  
Now, the team uncovered in our prep work that you both went to Carnegie Mellon University. Is this where you guys first met?

AARON TAN  2:09  
Aditya and I first met at a dance floor at Zouk (a Singapore night club), believe it or not. And there was nobody that was dancing there. It was our scholarship ceremony. Both of us were IDA scholars: Infocomm Development Authority fellowship. That's what we were both on. And of course, we went to CMU. And that's where we became even closer friends, because he wasn't a housemate for many years in Pittsburgh. In Pittsburgh, there's nothing much to do, so you rely on each other in that sense. 

ALAN  2:31  
Interesting. So this is not the proverbial "cofounders meeting in a garage", but actually on a dance floor. My next question for you guys is where was the concept for Carro first hatched?

AARON TAN  2:45  
The long-and-short of this was that both of us were at the end of our scholarship life. So in that sense, we were five-plus years into the scholarship. We finished our time with the Singapore government. Adit was a Capital Mall, looking after innovation, and I was in the venture fund of Singtel. And we thought to ourselves, maybe it's time to start our own company. So at that point in time, I got him interested to look into automotive marketplaces, because as I was doing venture investment in the US, I met a lot of auto-related companies, Uber amongst many. And that's where I thought there is going to be a change in the way people move around from a mobility standpoint. And the net of this is that we like cars, ourselves and properties as well. And one thing led to another. We wanted to bring about that experience of buying cars in the US. When you buy your first vehicle, you tend to get information on the car very regularly through services such as Carfax as well as KBB (Kelly Blue Book.) We then decided that there was a white space in Southeast Asia. Why don't we do something like this in Southeast Asia. That was where we first started the inkling of what to do in Southeast Asia: create an auto marketplace. And that was how Carro was started.

ALAN  3:42  
Gotcha. Fascinating origin story. Now Adit, for those of our listeners who aren't aware of this, Carnegie Mellon was where the field of artificial intelligence was invented by Alan Newell and other CMU faculty in the 1950s. It in fact today boasts the largest number of entrepreneurs in the autonomous driving and machine vision segments. How are AI and big data part of Carro's longer term strategy?

When we tried to build a trusted platform for use cars, in this case, a very expensive item for most people, there are plenty of data that are involved just to perform the transaction. Although at the moment, the market is very traditional, we will see that big data and AI will eventually embed in every part of the Carro process, from deciding on the pricing system, doing inspection, predicting demand and supply and even doing KYC ("know your customer") for financing as far as the underwriting process. This probably will be invisible for most people who are not the industry, but we will increasingly embedded this technology inside our own internal processes, as well as offering it to the customer.

AARON TAN  4:44  
We do a lot very interesting stuff. For instance, we also use AI for things that are more back office, things like inspections. We are able to for instance, take a picture of a vehicle or take a video of vehicle and in real-time automate the annotation of a scratch, automate what are other issues with the vehicle and for that matter, then back into the calculation of how much it costs to repair a vehicle like this. In the early days, we also experimented with things like listening to the engines. If you're able to listen to the engine itself, can you then detect the degree of confidence what is the issue if the car itself? If we are able to do all these things, this negates our need to even have a human physical inspection of a car because all we need to do is to tell the customers or prospective seller, "Hey, put your phone beside the car, take four or five pictures around the vehicle and maybe two more in the vehicle." And then in real time, we will be able to tell you what exactly is the condition of the vehicle. And that is something that we are striving towards. We are trying to push the boundary and push the way we think about how we should change the future of car buying or car selling.

ALAN  5:38  
So Aaron, on the topic of big data, we talk a lot about this "proprietary pricing algorithm." Is it possible to measure how this algorithm actually improves pricing over past conventional methods? Can you give us an example of how this comes into play to remove the frictions or inefficiencies of pricing? 

AARON TAN  5:59  
For us to be very frank in Southeast Asia, before we even came about, there was no means of pricing. Pricing means going on to the classifieds. The information is very asymmetrical. The way we think about pricing is that we control a lot of the supply of the cars at the wholesale level. So that allows us to see a lot of information around the "bid spreads." And as a result, we can infer things like what is the interest in a particular vehicle make and model, or what is the last transacted wholesale price of a particular make and model year, from which we can deduce through the model and through machine learning. As a result, we can infer what is the effective price of the car today. What has changed for us then is that we are then able to make very quick purchase decisions. We can tell within a few seconds what exactly is the pricing of the car based on today's market price. So that is a huge leap from what it was like even just three, four years ago. Everything in the past is based on the dealer's inclination and his experience of selling. And this is a very myopic view of pricing information of a particular car because not every dealer knows the pricing of every car. So they tend to specialize and that makes pricing very, very hard.

ALAN  7:00  
That's fascinating. Wow. Continuing on this train of thought, Aaron, what has been the coolest innovation that you've introduced over the past year to the customer that they could kind of say, "Wow, this is something I don't experience on any other platform." Has it been the Carro car subscription service, which you launched last year in Singapore? Or was it something else?

AARON TAN  7:24  
As a company, we tend to do a lot of very interesting stuff. We tend to focus our time on the transaction piece, making buying and selling more seamless to end consumers. And we spend more time to look into more interesting services that could change the future of mobility. You just mentioned subscription. Most people may not know this. Subsctiption itself is a very common thing in the United States. Because this is, in some way shape or form, similar to something called a "lease-to-own" program. But it's not quite a lease-to-own program because you don't get the ownership of the vehicle at any point in time. But it is a new way of thinking about car ownership because you don't need to own the vehicle itself. You can make do with the fact that you need this car for six months. And it becomes more like a subscription, like how you subscribe for Netflix and any other things that you pay for on a monthly basis. If you do not need it, or you need to upgrade it, you can upgrade it at any point in time. And that's what we are trying to offer the customers from a flexibility standpoint. But if you ask me what is the coolest product we offer the last year itself, it has to be then what we call "usage-based insurance," which is much more interesting, at least in my opinion, because we are changing the way we should price the risks of a driver. Because currently its is a "one-size-fits-all" approach. The way I drive, the way you drive Alan, is the same from a risk standpoint. It's a one-size-fits-all approach. But with the launch of usage-based insurance and soon behavior-based insurance, or BBI, I think it is going to change and revolutionize the way that we pay for auto insurance across Southeast Asia. The biggest benefit is as simple as the fact that we have been able to save our customers premiums, well over 50% of what they paid in the past. So if you pay $1,200 or $1,500 a year, we know your current insurance bill with us is probably closer to about $600 or $700.

ALAN  8:54  
Super impressive. Now a question for you Adit: Carro now operates, if I'm not mistaken, in Singapore, Thailand, Malaysia and Indonesia. Is the value proposition nearly identical, or quite different across these countries? And if there are differences, what are the reasons behind those differences?

Overall, our value proposition is somewhat identical. We plan to become the trusted platform that helps users throughout their journey, buying cars selling cars, anything related to make sure that the car is running well, including after-sales insurance warranty. The difference is always in the implementation because when we enter a certain country, we don't have the entire suite of our product ready to be deployed. And when it's deployed, sometimes it may not be economically feasible in certain countries at the start. But we try to keep things on par, especially on the product side, simply because it will simplify our product development and ensure that once we develop a product, it can be reused across multiple countries with minimum customization. So for example in Indonesia our value provision is enabling people to buy high quality used cars without worry. So in this case, we offer a 90-day engine and transmission warranty, no accident, no flooding, proper document warranty and a seven day buyback. We have a similar thing in Singapore but it is not as prominent simply because most of this offering is deemed to be default in the market. In Singapore, we offer very fast financing. Within one hour, the application can be approved. In Malaysia, when we invested in myTukar, we also started with a B2B model. We help people sell their car very quickly. So in the overall grand scheme of things, the value offering will be the same, but implemetation is different, depending on the country and the stage of development. 

ALAN  10:26  
Now Adit, a follow-up question for you, given that you sit in Indonesia: with Carro, what kind of development or engineering do we have in Indonesia? And what is the potential to build ever more sophisticated development teams in Indo? 

In Indonesia, we maintain a minimum presence of development capacity. We have a number of developers who develop and maintain our Jualo.com platform. Jualo.com is a horizontal classified website that helps people buy and sell cars, C2C style. At the moment, most of the development for product in Carro is currently centralized in Singapore. To be honest, building deployment capacity in Indonesia is not easy. It's not so much about hiring a developer. We need to hire the entire set of a team for the product, for QA, for developers, as well as for product and designers. So it's not easy to build a sophisticated development team in Indonesia. We are going to build a development team in Indonesia: a production house. But it will likely not be as sophisticated as the one in Singapore. It will be to maintain localization and customization that's relevant to the market. And we also do this kind of things in Thailand as well as in Malaysia.

ALAN  11:29  
One follow up question: what were the greatest challenges in setting up Carro's operations in Indonesia? Was it staffing up? Was it forging critical partnerships, fostering adoption? Or was it something else?

I would say all of the above and more. Personally, I think the greatest challenge is always getting the right talent, which as a similar aspiration, the necessary talent to build what we want to build, which is a trusted platform for people to own cars. There is plenty of competition in obtaining good talent. And this industry is very traditional. It's unlike e-commerce where people have been educated by Rocket Internet for many, many years. Used cars tend to be a very obscure industry in Indonesia. And even though I would say that we managed to hire many good people, it is still a balance between art and science to get the right talent.

ALAN  12:17  
Now, Aaron, while we have brought online, automated and streamlined the otherwise often laborious process of car buying; we also do offer offline services such as repair and maintenance. So my question for you, Aaron, is: how do we view our offline and online balance in that regard?

AARON TAN  12:36  
This is a very interesting question. Because the way I look at this is that it is more about balancing investor opinions. Because the truth is, from a consumer standpoint, yes, they care about whether or not they get to "kick the tires." But from an investor's standpoint, you'll look at this and say, "I do not want to invest in companies that are asset-heavy." And we have to be very careful about the line that we shouldn't cross where we become too asset-heavy or too asset-light. But coming back to the question: how do we view our offline and online balance? We can think about it more from a customer experience standpoint, because the net of this is that we need to be able to fulfill certain customer needs. And a lot of it, because of the space we operate in, cannot necessarily be done entirely online. For instance, buying the vehicle itself: sometimes the customers will say "I do want to kick the tires." They just want to see that the car exists, that the color is the same color that is what they thought it is. Sometimes they then have to, as a result, be able to have access to the vehicle. So we just have to allow that to happen, because that is the nature of the business that we run. And similarly, for repair, maintenance and after sales. For us it is a constant balancing act about how much of the transaction should be a marketplace-driven one, IE: we don't take any inventory at all. How much of it should be purely inventory-led, so that we can maximize margins, and for that matter, maximize customer experience. Our view is that we should try to stay as much as possible online. But we recognize that there is a need for an offline presence and  offline operations today. We continue to invest in offline stuff, but with the whole plan of moving customers; buying bids, selling bids, any part of the whole purchase or retail process; online. So our job is to try to continuously optimize it. If you think of the whole offline car purchase industry today, I would dare say that 99.9% is offline. Our plan here is as simple as: how much of those 99.9% of the customers, which is a few million cars a year, can we convert to be online? And that is the balance that we are trying to maximize now.

ALAN  14:23  
That's a very useful broader context which I think you need to take into account when you think about the value proposition. Now Adit, how have consumer behaviors changed since the onset of COVID in Indonesia? Did you find any behaviors that surprised you?

We observe the customer, whether it is the end user or dealership, slow down his spending or risk taking appetite significantly. The mix of lockdown and uncertainty of the economy, as well as the retrenchment, was preventing them from deciding to buy a car. A customer not buying a car means that the dealership is also not buying cars. And this was affecting much of our operations. At the same time, we also observed that multi-finance companies also started reducing their disbursement ratio. They didn't give loans as much because of the current climate that was becoming uncertain. And this happened for two, three months. Then, when people are more familiar with this pandemic and its toll on the economy, we observed customers getting more and more eager to buy cars. So our sales keep increasing about 5X or 6X every month. And it's hard for us to fulfill demand. And one slightly interesting thing is that, in the past the customer usually preferred to buy cars with a low down payment. But now after COVID, for some reason, customers are more willing to shell out cash in Indonesia.

ALAN  15:37  
So Aaron, we continue to articulate a financial services vision, including an in-house financing solution through sister company Genie Financial Services. How are we uniquely placed to offer these services better than anyone else in FinTech?

AARON TAN  15:53  
Well, we have been doing Genie Financial Services, which is a fully owned subsidiary of Carro, for over four years now. Why are we better placed than anybody else? Well, we have the largest auto loan book in Southeast Asia from a startup standpoint. In fact, the loan book itself well exceeds USD150 million within the last three, four years as we are putting it together. At the same time, we continue to have very, very, very low NPL (nonperforming loan) rates. When we talk about NPL rates in our company, it is typically 0.1%. Even during the COVID times now, when we are experiencing a little bit more default than we expected to, the truth is that while we continue to have headline  NPL rates of just about 1%, most of the default is not happening, or has not happened, primarily because of the way we have been doing collections; and I guess more importantly because of the way, we have been underwriting credit. We have been experts in underwriting what we call "higher purchase" credits, especially for this particular business. We're not underwriting $100 loans. We're underwriting a $20,000, $50,000 or $100,000 loan size. So it's a different form of lending. And that's something that none of the startups are focusing on. And as a result, we have been able to claim "pole position" in the auto financing business itself. And more importantly, since we have this four years of head-start, it allows us to build credit models that are very different from everybody else. And more importantly, internal processes; be it claims be it credit, be it NPL; the way we do provisioning is very sophisticated. And the other broader issue is in a business like ours, because we deal with slightly lower margin loans, it is a lot also about the cost of capital. We have been able to borrow at scale hundreds of millions of dollars. In fact, we just announced that we just closed our debt round of $150 million. We have been able to keep the cost very low. The loan book itself has been profitable. So it is no question that as a result that we are much better and more uniquely placed to offer these services than anyone else in the space. 

ALAN  17:32  
Aaron, can you give us a quick overview of the competitive state of the market at this point in time?

AARON TAN  17:37  
Oftentimes, when people talk about competition, they will compare us with classifieds and maybe other "C2B" (consumer-to-business) players. We continue to see of course, certain classifieds players in Indonesia and Thailand. In every country there's at least one. But we don't see ourselves as competition to the classifieds because we don't make a single dollar of advertising revenue. So the ones that we get compared to often are the classifieds. And then sometimes investors compare us to or we call "C2B" companies. There is this particular Malaysian company that they always like to compare us against, because of similarities in our geographic footprint. But as a company, we operate rather differently. That particular company recently claimed that we raised $30 million from some venture funds in Southeast Asia. We are not sure how they got to that amount. There is a bit of difference in terms of what we understood, they raised, which is about half of it. But coming back to the differences between us and them: we focus a lot more on the consumer buying experience than the consumer selling experience. And the former company that I was referring to, we are talking about more an auction service company. For us, while we have an auction service business, it is only part of the company. For us, a lot of the focus is about what we call an "AI-driven retail buying experience," whereby we try to provide our customers with the best buying experience through AI and machine learning; whatever data that we have on the vehicles or on the customer himself. And we further differentiate ourselves in the fact that we own the whole business end-to-end. So we don't just talk about simple things like just selling the car to end-consumers or end-dealers, which is honestly very easy to do. If it is nothing more than just frivolous buying of vehicles and then just flipping it off at a loss; anyone can do that with enough money. For us it is about: how do we ensure that the customer has the best experience? How do we ensure that we continue to make this a functional, profitable and a viable business? This is very important. So a lot of our internal focus is on GPM's (gross profit margins.) And since we own a marketplace, from B2C standpoint, we also own a wholesale standpoint, being the auction house. Then the question is: how do we then further monetize this, which is why we then moved into financing, insurance services, etc. We have our own insurance license. We have our own financial services license across our markets. We also have after-sales services. That is something that again, none of our competition, at least from a startup standpoint, is offering today. We are, and remain, the only person that plays on an end-to-end on a full stack ecosystem basis across Southeast Asia.

ALAN  19:45  
Adit, what are we focusing on in 2021? Is it new market entry, new feature and service line development, or other innovations? 

We just stick to our plans: grow our B2C business faster. We're going to quadruple or triple our wholesale number, and of course continue in our efforts in providing financing in countries that we already have presence in.

ALAN  20:07  
That's very useful. Now Aaron, what do we expect to be our revenue breakdown by, say 2023?

AARON TAN  20:14  
So by then, we expect to be well over $1 billion in terms of revenues. But I think what is more interesting here is the GPM (gross profit margin) standpoint because for us, as a company, we do not aspire to be just another company that is looking at selling cars at a loss so that we can pump up revenue numbers. Last year, we did about $100 million plus in terms of revenues. This year, I think we're probably be close to $300 million to $400 million. Next year, we are definitely targeting well above 700 million. But again, I stress that for us, it is a very different business. At the end of the day, we run a high GPM business. We  try to keep our GPM at well over 10%. But from a revenue standpoint, by 2023, I would say somewhere closer to $1.5 billion to maybe $2 billion. 

ALAN  20:50  
Now Aaron, is our most likely liquidity event an IPO? When would you consider a public event? And what milestones do you want to achieve first before proceeding with one?

AARON TAN  21:00  
This is something that we constantly debate. And to be very frank, the last time we closed our series B round, it was already well above a $200 million valuation. In my opinion, given my venture background, we have IPO'ed a few companies before on the New York Stock Exchange and NASDAQ back in 2014, when I was still with the fund, my experience tells me that if I'm past about $200 million to $300 million in size, from a market cap standpoint, it is very hard for me to try to sell the company from a trade sale standpoint. In fact, it's almost impossible. So an IPO becomes the most likely liquidity event. Of course, for me, the preference is always to do a trade sale. That is always the best versus doing a public equity offering. But for us to do a public offering, there are too many things that need to be ready. Everything including the team itself. We need to get people, the CFO, the CRO, which we do, because there's something that has been in the pipeline for a while now. But more importantly, we need to make sure that we have good revenues, good GPM's, a good story, which is very good now because of the fact that Carvana and Vroom are trading at very good multiples. So I think from a story standpoint, the Street gets it. For us, it was more like how do we showcase a strong revenue growth over the last few years and more importantly, over the projected few years. That will ensure that once we are a public company, we remain a strong public company versus a public company that is trending downwards over the long run. I think that's the first thing: we need to have a good strong growth and at the same time, even stronger growth projections moving forward. The second thing I would do number one is always the team. Number two is that and number three is to continue to look into bettering our GPM's because at the end of the day, if we are going to be a public company, if we go out and we announce we are doing $700 million in revenues, but on a 2% GPM, nobody's going to take us seriously. So that's something that we need to work on, which is to continuously push our GPM from almost 10% today. What we can do is get it to 15% or even 20%. That's something that's constantly on our minds.

ALAN  22:41  
Gentlemen, this has been an absolutely fascinating set of insights into the cutting edge of how the car buying and selling process is evolving. I really appreciate you guys taking the time to join us today. Thanks Aaron and Adit.

AARON TAN  22:53  
Thank you, sir. 

Thank you Alan.

ALAN  22:55  
Terima kasih telah mendengarkan. Sampai jumpa lagi!