Belvo: Well Beyond Plaid for Latam

A talk with Ray, VP Product at Belvo

Among emerging markets, Latin Amercia is the only Fintech investment destination with star exits, a low concetration of new giants, and a well-formed entrepreneurial ecolocy. According to Pitchbook, Latin America received $15 billion in VC funding in 2021, surpassing the previous six years combined. 25% of the funding flowed into Fintech. Such proportion kept rising in the first quarter of 2022.

For Fintech investments, we have observed two exciting exits in the last two years: Nubank and dLocal. As for the payment business in Public Fintech companies, 4 out of the 10 fastest growing ventures come from Latin America.

Since Consumer Fintech took off, what's the current stage of Latam Fintech infrastructure? Will there be companies like Plaid in this segmented market? We will try to answer these questions based on contents from Think Further, a Podcast featuring Patrick and Shaw. Patrick is now an analyst at Everything Unicorn and Shaw is a Fintech operator in the bay area.

Our guest Ray is now the VP of Product at Belvo. He is the first and only American employee in this Latam Fintech infrastructure company.

Recognized as "Latin America's response to Plaid", Belvo offers products that center around open banking data in Latin America. It serves clients like Mercadolibre, Rappi, and Jeeves. Founded by Oriol Tintore and Pablo Viguera in 2019, Belvo is now a member of YC W20. In 2021, Belvo raised $43 million in Series A funding, which, according to Crunchbase, was the largest Series A funding in Latam at that time.

Table of Contents

01 Why Are We Interested in Latin America

02 Infrastructure Development of Financial Services in Latin America

03 How Did Belvo Transcend the Definition of "Plaid for Latam"

04 Ties among the Bay Area, Chinese Entrepreneurs, and the Latam Entrepreneurial Ecology

05 Product View from YC and "The Lean Startup"

01. Why Are We Interested in Latin America

Patrick：We can begin with the question: Why Latam? I remember that Shaw became quite excited after talking with Ray about opportunities there.

Shaw: I went on a trip to Mexico. When I got off, I found Jeeves ads all over the airport. To put it simply, this company is the Latam version of Brex, with a business in corporate credit cards. I started shopping around and, from my experience, their Fintech infrastructure was rather backward. I felt like a nerd running into exams. I did some research after the trip and found that Fintech has actually become quite trendy in Latam since 2020. Many people went there to start a business. I then met Ray, who obviously was more confident than me in the market and introduced to me how Fintech was flourishing in Latam.

Ray: There were two reasons for me to join a Fintech company in the Latin American market.

First was my personal reason. I grew up in China but moved to US at the age of 13. I've never worked in China, nor have I interacted with Chinese businesses in my previous company. But China, for sure, has the most advanced Fintech industry. There were numerous innovations that bypassed former systems and technologies. Since I already missed opportunities in China, I need to find a similar emerging market.

During my research, I found that Latam was quite similar to China 10 years ago. Many people didn't have bank accounts. They never borrowed, had no idea what a credit card was, and were still using cash before the pandemic. Apart from Fintech, their behaviors were also very much like those of China - many skipped computers and directly started using cellphones, hence a high penetration rate of mobile devices. These trends were quite similar to that of China in the past.

Another reason were my prediction for the overall market trend. The world is undergoing Fintech innovations. However, I believe that many innovations won't happen in the US. Like today, we are still swiping our credit cars (a most classical function). This is an old but handy technology, one that we are not yet ready to quickly get past. So, I think lots of innovations will happen outside of the US, and it will be much faster than Americans imagine.

Patrick: I looked through the public companies as I have been focusing on Fintech and SaaS in the US. The result was quite unexpected. If we solely look at the payment business of public Fintech companies and rank them in terms of growth rate, 4 out of 10 fastest growing ventures come from Latin America. These include two acquiring companies in Brazil: Stone and PagSeguro, dLocal in cross-border payment, as well as the pan-latam MercodaLibre. Nearly all maintained growth rates above 60%. Stone's growth rate was even above 150%. From a rear-view mirror perspective, Fintech in Latin America is actually an asset segment that has somewhat been validated.

Ray: What's most interesting is that these four companies are still at early stages in terms of their market shares and penetration rates. For example, Belvo just launched an AP/AR data product, which obtains data from acquiring agencies like Stone and PagSeguro to help lending companies issue credits. While doing business, we found that their market shares did not rank in the top two, but rather, fourth and fifth. Moreover, currently they focus on small and micro businesses. So they have huge growth potential.

Market Shares of PagSeguro and Stone. Source: Investor Deck@PagSeguro

Outside of this market, although companies like MercadoLibre have been doing e-commerce for 20 years, most South American countries still have low e-commerce penetration, showing greater room for growth than the US.

02. Infrastructure Development of Financial Services in Latin America

Patrick: Now let's talk about financial services in Latam. Many studies I read in the past stated that 70% of the payments in Brazil were still made in Cash in 2019. But last month, a16z Fintech organized a field trip to Brazil for founders of its invested companies. One of the entrepreneurs described Brazil as quite similar to China: people go out without cash. They can also use PIX scans to pay for a drink at the beach. I don't know what you guys think.

Shaw: I was at Mexico city around 4 or 5 months ago. Many still used cash. They also used Visa and MasterCard, but those were mostly tourists. Honestly, I felt quite confused by the fact that people all used WhatsApp but still paid in cash. To me, it is such a pity that WhatsApp doesn't have payment functions like Wechat. But I understand there are reasons, like supervision, and dispersity among countries.

Ray： Well, first I think we need to take a step back. Many entrepreneurs and VCs like to phrase Latam as just one single place. During the pandemic, why did VC investment in Latam become 10 times that of 10 years ago? I think it's because founders from Latam had much better story-telling abilities than before. When bundling Latam countries together, one can tell a bigger story. But the fact is, few companies can become pan-Latam like MercadoLibre. Nubank also had a hard time trying to expand outside of Brazil. The most important battlegrounds in Latin America are the three countries with the largest GDP: Brazil, Mexico and Colombia, and these three countries are very different.

Brazil is in fact the most advanced in Latam. Its central bank is really forward-thinking. For example, they were quite eager to learn from global best practices like UPI in India and established the real-time payment system. Brazil had both good ideas and concrete actions. It also limited the power of big players like Meta. That's why Brazil's open banking and real-time payments are the most advanced in Latam.

Source: a16z

Note: PIX's user number reached 124 million 18 months after its launch. Over 75% of Brazilian adults use PIX to receive or transfer money. On the contrary, it took 4 years for UPI in India to reach the same scale. PIX now has an average of 1.6 billion transactions per month.

Apart from PIX, the Brazilian government has also been working on Payment initation in open banking, so as to expand PIX's use case.

Mexico has also been talking about real-time payment and open banking for years. But with only proposals and no implementation, it's just dragging its feet. Colombia is also very different in that its financial infrastructure and systems are controlled by the largest banks. Though Colombia has a real-time ACH payment system, such a system is still controlled by the traditional banks. Now the new president of Colombia is trying to change the situation where big banks are earning effortless money. But there are some political struggles.

These countries also have very different user behaviors and habits. In these three countries with the largest GDP, only Brazilian people have an investment culture and know how to buy stocks and funds, while Mexico and Colombia do not have this habit except for the top upper-class. People in Brazil also like to use cards, and there are two reasons behind this. The first is that they like points and cashback, and the second is that every card issuer in Brazil offers installed payments. Such a practice is engraved in its culture, but absent in other countries. As such, people in each country have different expectations of financial services and products.

Shaw: Now I understand what you mean. The single success of a product in one country can not be replicated in another. Apparently, people have different investment habits. But would infrastructure companies like Belvo be more universal?

Ray: Of course there's a relation, which was most surprising to me. All US companies are aware of this procedure - using Plaid or MX to obtain bank data, analyzing your personal data with financial planning apps like Mint, and then telling you how much money or how long you still need to financially retire. In the US, it's just impossible to build apps like that without bank data. There are 5000 banks in the country, which is impossible for customers to connect and acquire data one by one. Meanwhile, Plaid has been operating in the US for over a decade. So, both customers and tech companies know how to operate.

In Brazil, where Fintech is most advanced, customers also have a general idea about why and how to obtain bank data, as well as the value that companies like Belvo can bring. But in Mexico, we came across two problems:

1. Both customers and banks are afraid that you may steal the data;

2. Customers believe that you have gained access to a bunch of JSON data with your API, yet have no clue on how to use it.

In that sense, both Belvo and its competitors find their customers in great need of consultation. However, we are not consulting agencies, nor are we SaaS readily available for customers with a monthly subscription. Instead, we need engineers from both sides to connect API, and then let credit analysts interpret data.

There are high touch and low touch SaaS, and we belong to the high touch. Outside of Brazil, we need to educate companies, governments and users on matters like why Belvo exists, and how it can protect data. Such educations are really important.

Patrick: Following our discussion about Belvo infrastructure: Fintech infrastructure in the US is already well-developed today, from Banking-as-a-Service to API, KYC, Anti-fraud and data normalization for all types of payments. When I was at Fintech DevCon, people were discussing questions like which part of the technology stack should be built on their own, which part should be purchased outside, and how to manage these purchased SaaS. I wonder to what degree Latam customers are embracing such infrastructure?

Ray: I would start with market segmentations. There are different practices for big and small companies. Belvo intends to focus on big corporate clients, who are usually more experienced and have a relatively sound system. For example, companies like MercadoLibre and BBVA would sometimes use outsourcing agencies and consult with agencies to help them deploy certain systems. Especially in Brazil, the government requires any financial company that reaches a certain scale to grant access to the open banking infrastructure and provide APIs for their open data. On top of data points, data extension is also required for banks in Brazil. Some banks would carry out this work on their own, while some would buy products from Belvo.

03. How Does Belvo Transcend the Definition of "Plaid for Latam"

Patrick: Belvo called itself "Beyond Plaid". Compared to Plaid, what are the different problems that Belvo faces and what solutions does it have?

Ray: All startups in open banking have overlapping businesses, but some can also be quite local. The most important business for open banking is to acquire data from banks. This, however, is not stealing. They first need to gain consent from customers. A possible scenario is that customer's data and money is stored in banks, and that customer intends to use such data to borrow money or get a new card. We can then obtain data from the bank through customer access, which is the main part of our business.

Our business does overlap significantly with that of Plaid. But we also have many differences. There are many small businesses and self-employed people in South America, who all live in an "economic system" consisting of companies like Didi, Uber and Rappi. These companies plan to build super apps, where drivers' income may not flow into banks, but end up in the app. When they try to borrow money from other platforms without bank accounts, Belvo would help them make use of data in the app.

Shaw: So we can say that Latam apps are a little more advanced and many bypassed banks. Belvo then wants to converge them all, and you guys also have a payment business.

Patrick: When soliciting questions from friends, I found one thing that mostly intrigued them - Plaid is a very clear direction in Latam. Obviously, there must be a company engaging in such businesses. And Belvo was just founded in 2019. Why weren't there earlier players? And you guys didn't really have many homegenous competitors when starting the business. While watching customer interviews, I also noticed that some used both Belvo and Fintoc, as Belvo focused on Brazil and Mexico while Fintoc was for Chile. They were not even competing in the same market. In your opinion, what's the difference between competitive landscapes in Latam and the US?

Ray: I don't really know much about the history of Latam competitors, but I do have several examples that I've seen.

Bevlo is definitely not the first one. There must have been many companies who tried in the past decade, like Guiabolso, a Brazilian company who started doing personal financial management in 2012. It needed bank data, but back then Latam didn't have products like Plaid. The company, while all by itself, connected to banks one after another and finally became its own Plaid.

Guiabolso's Product. Source: idinheiro

Still, Guiabolso didn't make it in the end. For one thing, it was acquired by PicPay, the largest digital wallet in Brazil. Meanwhile, it turned out that doing business like Belvo was really hard. Technically, helping customers obtain their bank data is already not easy. They cannot directly download a CSV file from the bank. It is also difficult to retrieve data each time, even with users' permits. So it is really a tough job. To my estimate, there will never be over 100 companies doing business like ours in any country or region. I would say 3 to 4 at most.

In addition, for open banking, big companies already exist in almost every region. There are Plaid and MX in the US. Visa just bought Tink for $2 billion in Europe. In South America, Belvo is not the first player, but it is one of the largest.

We also have many competitors, but they might not be in those three countries with the highest GDP. Or, they focus on segmented data, specializing in areas like B2B bank data, ERP data, AP/AR data, and even tax bill data in government systems.

Shaw: You had good timing and seized the trend of open banking in Latam, which has been promoted by central banks in recent years. The market was still not in its late stage when you entered. The difficulty was kind of moderate for conducting business. And then your company emerged. But was it also because of the demand?

Ray: If you think of us as a platform, a Marketplace like Airbnb and Uber, we are characterized by having both ends of supply and demand. Belvo's supply is banks and their data, while the demand side is still mostly Fintech. Besides, some banks are also starting to buy our products, so they are both suppliers and customers.

Source: Belvo

10 years ago, there were no Fintech companies in Latin America and banks were quite at ease. One of a16z's partners specializes in Latam investments. He then wrote an article to summarize such ecology, mentioning that profits of Latam banks were many times higher than those in the US. They were making effortless money. It wasn't until the advent of many Fintech companies, coupled by governments' initatives for innovation, that these banks started to realize the need for change. (In that sense, 2019 was indeed an appropriate time for doing such business, considering the barrier of entry and demand back then.)

04. Ties among the Bay Area, Chinese Entrepreneurs, and the Latam Entrepreneurial Ecology

Patrick: Talking about entrepreneurial ecology in Latam, I had an exceptionally obvious feeling in Singapore last month. There are two companies doing Brex for SEA, Aspire and Spenmo. Both of them are YC companies. There's also a similar phenomenon in Latam, which is Belvo and its competitor Fintoc, as well as Jeeves, the Brex for Latam. All of them are also YC companies. In that sense, do these top startup owners in Latam have some kind of relationship with the Bay Area today?

Ray: I have a good understanding of the YC ecology, because my last company (LTSE) was also a YC. All of its customers are startups, and we have over 1000 YCs as our users.

Currently YCs have several features. First is that it's growing bigger and bigger, with two batches of companies each year and each batch having a number of around 400. Another is that 50% of the YCs are outside of the US. For one thing, most startups remoted by default during the pandemic. Another reason was that many innovations happened outside of the US. In general, there are many talents all over the world. Latin America and South Africa also witnessed many interesting Fintech innovations. So, Belvo constantly ran into YC companies, who could either be clients or competitors.

You can't really say that YC companies went abroad to do business in Latam. But rather, YCs invested in Latam entrepreneurs. For example, Belvo has two founders, both of whom are Europeans and work in Spain. We have approximately 65% of the employees in Latam, and I am the first and only employee in the US.

Patrick: I remember that Shaw mentioned a trend: some startups have been doing business both in the US and Mexico since their first day of establishment.

Shaw: Like Slope, a YC company doing B2B BNPL. It started in both the US and Mexican markets, partly because Mexico is relatively close to California, and it is more compatible with the US economy.

Ray: I don't think I've ever met a company like that, but of course the big companies would come to the Latam market, like Twilio's team came to Latam and asked Belvo if they wanted to buy their products. I remember a company which is the Mexican version of Flexport. It does a lot of shipping businesses between the US and Mexico, because there are plenty of transactions and no import/export taxes. Personally, I am very optimistic about Mexico, especially because the US wants to pull more manufacturing back. But the State is too expensive. So, Mexico and Canada are two favorable locations for the US.

Patrick: Some Fintech teams in China are now expanding their businesses to Latam. I wonder what's your opinion about that. Most of them do not have deep personal ties with Latam. It was more out of a commercial motive.

Ray: I think there are many companies like that. For example, Liquido, a company doing payment API. One of its founders once worked on Didi's payment in Latin America. Personally, I think the company's strategy is to focus on cooperative business between China and the US.

Liquido's Official Website

Shaw: It sounds like the barriers to entry are not that high for Chinese entrepreneurs, as they can use engineering resources to do something related to Fintech.

Ray: Barriers surely exist, there are several things to consider:

1. Language and cultural issues. Before joining Belvo, I wanted to know whether I could successfully kick off without speaking Spanish or Portugese, and most people's advice was that there was no problem. After nearly a year in Belvo, my experience is that, for any company doing business in Latam, the default language is still English. Although Spanish and Portugese are similar, you can't expect a Brazilian to speak Spanish, or a Mexican to speak Portugese. But English is not 100% problem-free, we still need to use local languages during some client meetings.

2. The bigger problem is your understanding of that country's national policies, as well as the relationship between its banks and government. Latam is a huge place. The Mexican government alone can be quite difficult to handle. Many people wonder why big companies like Stripe and Adyen have low penetration rates in the Latam market. There is actually a reason behind it.

So you need to know local operations and recruitment. You must have someone who is local and really understands the market. Why isn't there open banking in Mexico yet? Why did PIX succeed in Brazil while the real-time payment system in Mexico didn't work? These questions are quite esoteric. You need to really understand what's happening behind the scenes. It's not something that an outsider can quickly grasp. It requires hands-on help from the local team throughout the process.

05. Product View from YC and "The Lean Startup"

Patrick: Another general question: your product is Latam-based, API First and also data-related. As a VP of Product, what are the new challenges during the process from demand analysis to product design when compared with your previous experience?

Ray: I don't think it's very much related to our business type and our location in Latam. Many of my ideas in this area are universal for companies across the world. I would divide a product into two segments: One is to build it from zero to one, the other is to optimize and extend the product after it gained good momentum. Most of the time, Belvo is still at the stage from zero to one .

While building products at this stage, I was influneced mainly in two aspects. First is the boss from my last company, Eric Ries, who wrote the book "The Lean Startup". In fact, hot words like MVP (minimum viable product) and pivot all came from him. Another is the article "Do Things that Don't Scale", written by Paul Graham in YC.

I was greatly influenced by those ideas, such as "starting from the small". When doing payment business in Latam, we don't need to become Stripe the next year and support all types of payments. Instead, we choose a segmentation or a familiar customer group, ask them what they need, or let them try what you've built. This is what we call "do things that don't scale". Companies with sizes like Stripe cannot always call the same customer. But if your company is small, it would then have an advantage - when the user encounters a bug, you can immediately have the engineer fix it in an hour. In that sense, users would think that you really understand their pain points and want to build a product to solve them.

Another point is experimental thinking. We need to establish assumptions and conduct experiments. The process of proving the assumption is really important. For example, when making payment API between China and the US, one of my assumptions was that bilingual developer documents would be more appealing to Chinese customers. It was a simple assumption. My MVP was to translate the Chinese document and show it to the client, who then said: "Our programmers find it a lot easier to use.", or "our programmers speak English, so they don't really care about that." If you keep repeating these minor procedures, you should be able to make a very successful product.