Product: Payment Gateway
Founded: November 1, 2014
Team: Moses Lo (CEO), Bo Chen (CTO), Juan Gonzalez (Co-Founder)
Experience and Education
UNSW (Sydney, Aus) – Information Systems and Finance
University of California (Berkeley) – MBA
The Boston Consulting Group – Senior Associate
Amazon – Product management and Marketing
University of California (Berkeley) – EECS
Brandcast – Engineer
Ripple – Engineer
University of California (Berkeley) – CS
Expedia – Engineer
- Product: A payment gateway, designed for Indonesia, that allows consumers and businesses to receive and send payments. Product improvements are determined by asking customers for their feedback and then the feedback is implemented.
- Initial Capital: Bootstrapped and won hackathons (including hackathons at UC Berkeley and A16z) and then entered Y Combinator (YC).
- Getting Investment: Having traction and growth is very important—don’t let your fear of people “taking your idea” get in the way of getting your product/service out there. Most startups that try to get investment without traction struggle. Also, investors tend to look for startup founders with good chemistry because founders breaking up is a main reason for great startups at YC dying.
- Gaining Initial Traction: Hackathons helped with some initial traction. Currently focusing on targeting startups first and reaching out to them in hopes of growing with the startups while they are still small. Building personal relationships is extremely important.
- Overcoming Competition: Understanding one’s target market and customer service are extremely important. By understanding the problems in Indonesia’s banking infrastructure better than potential competitors that are not yet in Indonesia, Xendit holds a significant advantage over foreign companies that enter Indonesia, even if they are much bigger. Great customer service is also a reason for winning deals over competitors.
- Hiring: Competency is important, but work ethic trumps all. The candidate’s resume or background is not nearly as important as work ethic.
Table of Contents
A. What is Xendit?
B. How Does it Work?
D. Gaining Traction
E. Business Model
H. Further Reading: Making the App
A. What is Xendit?
- What does Xendit do?
Moses: We have two parts to the business. One part is the consumer application, which is kind of like Venmo, and our other part, which is growing fastest, is a payment gateway product for other startups. Startups can use our APIs to receive money and send money.
- Where are your customers?
Moses: We operate in South East Asia, specifically Indonesia, where there are issues associated with the use of physical tokens used for bank transfers. To do a bank transfer, you need a physical token that you have to carry around. Also, there is a uniquely Indonesian problem that if you send money to someone, you cannot tell who it is from or from which bank. So if I send you money and I use Bank of America, you would not know that it was I that sent you money or that it came from Bank of America. It would show you the description of the transaction and the amount.
- How does your service help your customers?
Moses: Our goal is to solve these basic payment problems. We have automated the process with APIs, meaning that no humans are needed.
One of my favorite use cases is one of the companies that joined us early, Peer to Peer Lenders. They need to get loans out to borrowers and need to do that on mass. They started with us when they were doing about ten transactions a day and they are probably now doing hundreds of thousands of transactions a day. We can help them scale at this high rate without them hiring a finance team.
- Currently, how do most people transfer money South Eastern Asia?
Moses: Most of the transactions in Indonesia are still done in cash (it is something like 60%-80%). However, online commerce is growing at about 150% each year. We are kind of riding that wave as it grows. Most of that online commerce is done by bank transfers. In the US, everyone has either a credit card or debit card. In Indonesia, credit card penetration is about 3% and when you think about how those that can afford credit cards have about three of them, then real penetration is about 1%. Debit card penetration is at about 8%-9%, but many banks refuse to allow their debit cards to be used online. This means that people have to use cash or bank transfers and we are helping out on the bank transfer side.
- Why did you decide to target Indonesia?
Moses: I grew up in Malaysia, then Australia, and then the US. I get the culture here more than I would another country, say Brazil.
Indonesia is the fourth most populous country on Earth. I don’t think most people realize that there are so many people living in Indonesia. However, it is an Archipelago nation, which means that infrastructure is poor. At the same time, you have a huge GDP growth with a huge population, so you have a rising middle class. The difficulty in making bank transfers poses a great opportunity for us, especially given the low market penetration of credit cards.
- Who are the majority of your users (i.e. business or consumer)?
Moses: We do have a consumer side but our clients are mostly businesses like start-ups that need to make large volume bank transfers. Unlike in other countries, such as the in US, in Indonesia, there is friction because the technical infrastructure isn’t there to connect your wallet directly to a bank account. We want to be the first to do that.
There is a whole infrastructure that we have had to build on top of the banking system, which we are now selling to other businesses. We are more like Stripe (another payment gateway company that also came from Y Combinator) than anything else. Stripe does credit cards because that is what people do in the US and we do bank transfers because that is what people do in Indonesia. We also do the money out which is something that Stripe does but is not known for as of now.
B. How Does it Work?
7. What is involved in setting up the API and infrastructure on top of the banks in Indonesia?
Moses: It is amazing how much of a painful process it is to setup the infrastructure here compared to the US. To better explain, I will use the internet here in Indonesia as an example. To get the internet that we have here, we have four different internet lines. The reason is that at any point, one of these internet lines can go down. If it rains, the internet fails. If someone is working on a project requiring digging, due to bad rules around digging, people can actually dig up an internet line. Or perhaps there is an exposed fiber optic cable in the road and someone accidentally steps on it or cuts it.
Apply that same logic to every infrastructure, including banks. For instance, we work with this one bank that, every day, goes down from 11:30-2am. We have other banks where their servers go down for one, two or three days at a time. What that means for us is that we need a bunch of redundant capability to keep one product reliable and stable. The complexity that goes into making something that “just works” is amazing. The technology involved is relatively simple, such as the RESTful API that we have written. Yet, making something that is reliable on the poor infrastructure here that is able to scale is really difficult.
- So how do you meet this challenge?
Moses: The software development challenge is that we need to get money from A to B. There are a number of default routes, which are cheapest, but at any point of time those routes might get cut off. Then we have to re-route through something else. So the software challenge is to manage the complex networks that the different banks have. However, by aggregating all our banking infrastructure, we can get money from one point to another more quickly than banks can by determining which banks are functioning.
- So how do you send money to a bank that is not functioning?
Moses: If it is the recipient account, we can’t do that much. However, if someone wants to send money from one bank, say Bank of America, to another bank, say Chase, and the banks in the middle are down, we can simply reroute the transaction to avoid the banks that are down. That’s one of the great advantages for using our service.
Another interesting part is that, in the US, ACH is managed by the FED. Banks agree that they send files to the FED every night and it all gets settled. In Indonesia, it’s not “the FED”, it’s a set of three private companies. So these three companies are the actual plumbing between the banks, which is why we integrate with them to use their network directly, rather than go through a bank. Here, our knowledge of the Indonesian banking system gives us a big advantage over competitors that assume the banking model is the same here as in, say the US.
- What is a RESTful API?
Moses: A RESTful API is pretty common. It is just the Silicon Valley standard for how APIs should be written. However, banks don’t have these APIs. In fact, they may not even have any APIs. So we integrate with the banking system and serve really nice API to other startups.
- Who are your competitors?
Moses: I think that local payment gateways and some banks present different kinds of competition. I think it is inevitable that global players like Stripe will come to Indonesia—I know Stripe is looking at this country closely.
- What are some of your advantages over your competitors?
Moses: We understand the culture here better than others so we can develop relationships more successfully. Also, when you do business here, you have to understand what the business needs are. For example, when Stripe comes here, they will come with credit cards, but when you look at where the majority of the ecommerce and transactions that are taking place, it is in bank transfers.
When it comes to banks, I think that our advantage is obvious. Rather than having to have a bunch of manual people clicking these physical tokens and manually entering transactions through an internet banking page, you can use a secure, API driven system that is easy to use. Also, as I mentioned before, we understand how banks settle transactions here which is unique to Indonesia.
Compared to local, digital payment gateways, we do both the money and the money out while most digital payment gateways here just do the money in. I would say the equivalent is WePay in the US. WePay continues to grow, even while competing with Stripe, because WePay focuses on platform businesses that receive money and send money out.
The second part is that as these local competitors have gotten bigger, they have had to push towards traditional businesses—meaning that they lost their original focus, which was startups. One reason that I hear we are winning deals is that our competitors have bad customer support, such providing answers to questions that are generic and not specific to the customer’s needs. So our goal is to get startups while they are small and then grow with them and continue to serve them well.
D. Gaining Traction
- How did Xendit gain initial traction and what was the hardest part?
Moses: We have been following the same formula that Berkley and Y Combinator taught us which is “Make something people want”, which is an easy concept but hard to execute. The way we execute is by incorporating customer feedback into our designs. We always ask our customers and potential customers what problems they are facing. They tell us their problems, we come up with a solution, and then ask for their feedback. So there is this cycle of us asking customers and potential customers “What would you like us to add?” and “What would make you use us over someone else?” and then implementing those features. Each time we add a feature, we try to do so in a way that in scalable so that we can pitch it to other people.
- What is your growth model?
Moses: We have two growth models. On the consumer app side, we got over 200,000 downloads within one year. The key for us is to make such an appealing application that you want to share it with someone else. Venmo is only useful if your friends are on it (and we are in a similar position).
On the business side, it is a little different—Indonesia is very relationship driven. It is us getting products out there by talking with people, like Braintree and Stripe. Braintree famously got Lyft and Airbnb and that basically carried them. So we made similar bets. We think about who are the promising startups and go acquire them so that we grow with them.
- What made your users want to use Xendit in the first place? Did you advertise?
Moses: We have never done advertising—one day we will do that (and I am excited). We spend most of our time on the business side. It is about building relationships with people who we have gotten to use us. I go to other startup founders and ask what are their payment needs, tell them that I think we can help them, and then over time try to convince them to use us. So a very normal, b2b style game.
- Is there a reason that you haven’t done advertising yet?
Moses: In Indonesia, you go to your friends on Whatsapp or another messaging network, and tell them your current problems or thinking. So, it is a different kind of advertising that needs to be done. What we have seen to be more effective is going through Facebook to target people in different companies, and Facebook advertise that way.
- How well has your b2b approach worked?
Moses: We are currently growing 30% every month using this strategy, in terms of our compound annual growth since we have started (we launched in July last year). We will continue have our staff go out and “hunt” until we see our growth slow or we feel that we cannot continue to handle to grow at 30% every month. Eventually, we will go to Facebook advertising and/or Instagram advertising, or whatever works locally. For now, personal relationships are really important for sales. For example, recently I was on a deal where everything was hashed out, but the CEO hesitated in giving us the go ahead. I had breakfast with the CEO to cement our relationship. We knew the pricing, what the integration would look like, and everything. We talked mostly about soccer, life in Jakarta, and previous jobs. We spent about the last ten minutes talking about the deal. Nothing was changed in the deal. It was just talking about it and confirming it. Even if we do some sort of advertising, it ultimately ends up being I or someone else going and meeting someone.
- You mentioned the possibility of using social media, such as Instagram, for advertising. Since. Instagram is very popular in the US, does that mean that you might soon try to expand to the US, or will you continue to focus on Southeast Asia?
Moses: I think we will continue to focus on Asia. I think Southeast Asia is the fastest growing region in the world (in terms of GDP). There is a rising middle class and problems with infrastructure that handles transactions. That is an opportunity because it means you are going from zero to smart phones. People here don’t know what desktops are because they have never seen one before. With technology, everything is mobile first and things are not thought about from the perspective of a credit card world. Furthermore, this is where a significant portion of the world lives, so I think there are advantages in that. Then, I think we will look at countries that are similar to those in Southeast Asia. For example, I think Brazil is really similar to Indonesia. A huge population, bad infrastructure, young people that are driving the country, a rising middle class. The US has one the biggest economies for now, but it won’t be for very long. We want to tap into Asia as it begins to overtake the US.
E. Business Model
- What are your two different bank transfer options “unique amounts” and “virtual accounts”?
Moses: So the problem we are trying to solve is the detection of funds. As I said before, if you send me money from Chase to Bank of America, I have no idea where the money came from or what it was for (for Indonesia).
A local workaround is using unique payment amounts. If you buy a shirt from me worth $30, I will charge you $29.99 and reserve that number just for you. When I receive a payment for $29.99, I know that it was you. When I receive a payment for $30.01, I know that it is Bob and when I receive a payment for $30.02, I know that it is Rom. The reason that it is cheaper is that it is our own software on top of bank accounts.
Virtual accounts, on the other hand, are actual banking infrastructure, which means that they charge us for that, so we pass on some of that cost to our customers.
- How does pricing work?
Moses: For money coming in for merchants and other startups, it works out to about 30 cents per transaction. For credit card transactions, we charge a percentage. We also charge money coming out at about 40 cents per transaction.
- Why is it slightly more expensive to send money out?
Moses: We provide the ability to send money instantly from point A to point B. The banks don’t provide that. The banks provide you with the ability to send about $200 a day, but most startups are dealing with amounts more than that and they want instant settlement. We are still about 40% cheaper than the banks and they don’t allow you to send more than $200 a day.
- What is the difference between your “merchant” and “business” categories?
Moses: The “merchant” category is meant for those using ecommerce. They tend to be individuals and smaller entities. They often do not use our APIs, instead using our UIs (user interface). Startups tend to use more of our APIs and go deeper into the code.
The way we talk to those two parties is different, which is why we have two different categories on the website and have different sales people to help both parties. Both merchants and businesses, are experiencing the same underlying product, but are having different experiences with it.
- What is your strategy or philosophy when hiring?
Moses: I used to work at Boston Consulting Group, which influenced my hiring thinking a lot. BCG is famous for not caring about where you come from or what you studied, but more about your ability to learn. For me, I look for some basic level of smarts, work ethic (which I think trumps all), and ownership mentality (which I think is the hardest to find).
When you have a task, it is not your task because you are told to do it, it is now your baby to run and determine direction. That is difficult to find even in the US, although I think that the US educational system is somewhat aims to foster this ability. Here in Indonesia, it is even harder to find that third quality. I spend a lot of time determining who has this “ownership” mentality and end up hiring a lot of underdogs.
To illustrate the mentality I have when hiring, I once gave a woman a job offer because she provided me with outstanding customer service for my internet–despite being on her day off. I went through many pains to get someone to fix my internet (one time the mechanic did not show up), and customer service failed to help me fix the problem until finally this woman helped me by making sure a mechanic was at my house within two hours. The competence and work ethic that she displayed made me think she was a great fit for our company. She doesn’t have a traditional university degree and has never had a really big job at a really big company–but that was not important to me, She currently works for us and works very hard for our company, often doing customer support at 2 am in the morning.
- How did you get funding?
Moses: We started out by winning hackathons, which gave us tiny bits of money (relative to what we have now), but it was enough to build something on the side, get some traction. We won a hackathon by Andreessen Horowitz (they are one of the biggest investors in the valley). We then got into Y Combinator and we were very lucky that we were growing quickly enough that we had a lot of demand for what we were doing. YC gives you some money for the summer and then we raised about a typical YC seed round I would say. The investors were from both Asia and the US. The US ones can help us with payment companies and payment companies that have succeeded (they have invested in other payment gateways, such as PayPal, Venmo, Stripe, and Braintree). Then, the Asian ones have a great understanding of the problems in Indonesia and can attest that the problems are real. Lastly, there are local ones who can help us with connections and introductions.
- Do you have any advice for someone who is running a startup and needs help with fundraising?
Moses: One is that the founding team is really important. The number one thing that breaks YC companies is founders breaking apart. This is why YC spends a lot of time making sure that founders have chemistry.
The second component is traction. Growth is really important. Some people say that they are in “stealth mode”, but it is really hard to raise money if you have no traction.
H. Further reading: Making the app
- What is required to make an application like Xendit from the perspective of someone with a software development background and for someone without a software development background?
Moses: This might be a very Silicon Valley opinion, but we are taught that you cannot make good tech without having in-house teams (things cannot be outsourced). There are probably many examples and counter examples for that, but in our situation, the few times that we have used external people we acquire a lot of technical debt (programming mistakes that have to be made up at the cost of more time and/or money). The incentives are not always aligned. Once the price is agreed upon, the team which software development work has been outsourced to will most likely care most about getting the project done quickly rather than completing the project well. On the other hand, an in-house team will more likely care about the long term and focus on making the software work well and make it scalable.
One example is a local unicorn which outsourced software development for about a year and acquired so much technical debt from different software companies building on each other, that in one of their investing rounds an investor said that part of the goal of this round is to bring all development in-house. In fact, they were losing in their industry because their technology was so bad.