Sacra Logo Sign In

Shamir Karkal, co-founder and CEO of Sila, on the modern payments stack

Jan-Erik Asplund
None

Background

Shamir Karkal is the CEO and co-founder of payments infrastructure as a service business Sila. In 2009, he co-founded Simple, the original neobank, before selling it to BBVA in 2014. We talked to Shamir about working on payments infrastructure, point solutions vs. end-to-end platforms in banking-as-a-service, and how fintechs today think about choosing providers for key banking functions.

Sacra Highlights

  • I think the experience of a lot of customer of  Synapse, at least, was that it seems like a very compelling proposition when you start. You're like, "Hey, everything that I could possibly imagine is built into this platform. It's just one API call away." And then a lot of customers realize that it's really hard to be good at everything.
  • Questions

    1. Let's start with your background and how you got started.
    2. You mentioned not liking the term banking as a service. How do you think about Sila compared to the sort of companies you think of when you hear BaaS, like Unit, Treasury Prime and Bond? What do you see as the big difference?
    3. There are two competing visions in this space—one being the consolidated suite of products in BaaS and the other being this approach where everybody picks individual solutions and builds their own stack. Long-term, which do you see as being better able to consolidate demand?
    4. I'm curious how your customer base looks, both from a startups versus bigger companies perspective and also from the perspective of products that just want to embed some form of money movement versus more pure fintech players.
    5. Does Sila focus on any one industry?
    6. You mentioned that there will be folks who gravitate towards platforms, folks who gravitate towards point solutions, folks who migrate. Are you seeing folks migrating in your own experience?
    7. You mentioned the transaction volume is roughly half/half crypto/fiat. Could you talk about why it's important for Sila to support crypto?
    8. One way of looking at BaaS is that your customers are a venture portfolio, and you're placing bets on which ones will explode. For Marqeta, it's Square and Instacart, and so on. What's your take on that, and does it resonate with how you think?

    Interview

    Let's start with your background and how you got started.

    I'm the co-founder and CEO of Sila. My background is in software engineering -- I used to be a software engineer twenty years ago. I came to the US, went to business school, and worked as a consultant for a while. Then in 2009, I was in Brussels, in Belgium, and my friend and classmate from business school sent me an email saying, "Let's start a retail bank." You'll see how crazy I am that in 2009, I thought it was a good idea to start a bank.

    I left McKinsey and moved back to the US, and we started up a neobank in Josh's basement in Brooklyn. I mean, I call it a neobank now, but that word was still probably ten years in the future. We had no clue what we were doing, but we figured out how to build and launch probably the first neobank anywhere -- Simple. Just sheer determination and big-headed stubbornness was really the answer because it wasn't smart, at least.

    It took us three years to launch it, because the whole partner banking industry, fintech infrastructure -- fintech itself, as a world -- barely existed. But we built it, and it finally launched it in 2012. Then in 2014, it was acquired by BBVA, which is a big Spanish bank. At that point, they were making a big push into the US. Then they just exited, sold everything to PNC and left. C'est la vie. Funny enough, somebody pointed out to me that that is still the biggest neobank exit in the US ever. Of course, Chime and others are massively more valuable now, but they haven't IPO'd and they haven't been acquired. In terms of the traditional definition of an exit, Simple is still the biggest. I was just like, "Really? At $117 million?" That seems almost a joke now, based on today's valuations and round sizes. But at the time, it seemed like a good exit. We'd raised less than $20 million, so everybody made money. And we distributed $14 million to employees.

    A few months after the exit, I was in Madrid talking to the BBVA executives. One of them mentioned this idea of building a platform. I was immediately like, "Yes, please. The world needs API platforms in banking." If this had existed, then Josh and I wouldn't have spent three years just trying to smash our head against the banking system to launch Simple.

    I started advising them on it. Then I eventually left Simple, moved to BBVA, and built and launched two API platforms for them as internal ventures, one in Europe, one in the US. For a couple of years, I lived a big bank exec lifestyle: I was on a plane and in Spain six times a year. And we built and launched it, the BBVA Open Platform. We signed up customers like Google. But we were never able to actually scale it as a business. At least, I was never able to do that, mainly because you could integrate into the Sandbox in a week, but it would take you twelve months to get past risk, compliance, legal and all the internal BBVA processes. That works for a Google or an Apple. It doesn't work for 99% of fintech or crypto startups. Back then, they couldn't even touch it. The BBVA Open Platform never really got traction because it never solved the compliance problems for customers. It was just too big, too slow moving. Though I think it had probably well over a million users and was bigger than the rest of the industry combined when BBVA exited the US and shut it all down -- it just tells you how unpredictable banks can be sometimes.

    I got frustrated and left in 2017, thought about life for a while and decided that I still wanted to solve this problem. I just didn't want to do it the way I'd done it at BBVA, definitely not in a large bank. The way we defined the problem is that it's just too hard to program with money. If you wake up anywhere on the planet tomorrow and you decide to program with email or voice over IP, or SMS, there's a SendGrid or a Twilio, there are APIs, STK ecosystems of providers. It's not hard to do. The hard thing is to compete with Google. But most people who are programming with email are not building Superhuman or HEY, they're just plugging in some email process into their app. It's super quick. It's easier to do it than to talk about it.

    The moment you come to money, the world is a very different place. There is no internet protocol for money. A bunch of crypto projects are trying, but nobody has gotten there yet. And whether you're in San Francisco or Shanghai matters a lot. There are almost no APIs or STKs. There are people like us now, and there's more available. But still, compared to anything else in the developer world, the state of the industry is twenty years behind.

    And you still end up having to work with banks. They kind of all suck, they just suck in different ways. Their technology is outdated. They don't understand how to serve developers. They typically don't have API platforms or STKs or anything that a modern developer would expect. And they're extremely heavily regulated. The entire industry is heavily regulated. They don't know how to manage that, so you have to become a compliance expert and go sign a deal with a bank and figure it out. The end result is it doesn't take three years to build a fintech app or a crypto app now, but it still takes 12 or 18 months.

    I'd argue that it's much easier to compete with Bank of America than it is to compete with Google. But most people are not building neobanks. There are 50 of those now or a hundred. There's a growing market. More power to them. But the reality is that the vast majority of the market is businesses and app builders just plugging in a payments flow into their app and not necessarily realizing that it suddenly took them over a line that made it regulated versus unregulated, and they're just struggling to do that. You're like, "Which business needs money?" I'm like, "I think more businesses need money than they need email."

    That's the market, and that's what we focus on. I guess we are a BaaS platform, but I don't like the term because it's banking centric. I'd like to make it developer centric. That's why I like to call it "programmable money." What that also means is that we always look at it in terms of the network and how to support folks moving money and dealing with complex funds flows, as opposed to particular use cases like neobanking. It also means that we've supported crypto from day one. We have our own stable coin on Ethereum. We have customers doing everything from crypto payments to NFTs. I haven't done this scientifically, but our customer base is probably 60/40 or 70/30 fintech/crypto. But again, there are so many fintech apps building in crypto capabilities, and crypto apps doing fintech things. The whole space is rapidly converging, and we like to sit in the middle of it.

    You mentioned not liking the term banking as a service. How do you think about Sila compared to the sort of companies you think of when you hear BaaS, like Unit, Treasury Prime and Bond? What do you see as the big difference?

    I think the granddaddy of that space is Synapse. They've been around for five plus years now. The banking as a service providers typically end up bundling a bunch of services, like KYC, KYB, ACH, card issuing, bill payments, sometimes photo check deposit, wires, and potentially a few more other things as well. I think the experience of a lot of customer of  Synapse, at least, was that it seems like a very compelling proposition when you start. You're like, "Hey, everything that I could possibly imagine is built into this platform. It's just one API call away." And then a lot of customers realize that it's really hard to be good at everything. You might be a really good ACH processor and be terrible at card issuing. I might be a really great card issuer and be awful at ACH or wires or bill payment or something else.

    So a lot of people, as they started scaling on Synapse -- as probably the oldest one -- realized that it just didn't work for them, because at scale you need an amazing card issuing platform, an amazing ACH, an amazing KYC, and an amazing everything. You need best of breed in every single thing, because every single thing is massively complex. There are a bunch of failures and edge cases.

    I think a lot of the early customers of Synapse ended up scaling on them, and then rolled off them and either built their own in-house bespoke platforms, like Chime and Simple did back in the day, or went to a model where they were using five to six different processors. That's how all the large banks operate. Most of the large banks are not built all on FIS or Fiserv. They have their own code, with some Fiserv tech in there, some TSYS, and twenty others.

    I'm fundamentally skeptical of the fact that anybody can build all of that and be good at all of it at scale. So we always took the approach of focusing on the core money movement. We do KYC, KYB, ACH, ledgers, crypto payments, and we're adding in wires and more capabilities around that. But for most other things, we just partner with people. We partner with Lithic for card issuance. We partner with Arcus for bill payment. We partner with Plaid for account aggregation. We try to find the best of breed in each capability and build a seamless network.

    As you build your infrastructure honestly, it is more work. It's more work to integrate Lithic and Sila than it is to just integrate one API. But you'll probably end up finding out that, as you scale, it ends up being a much better solution, because Lithic is probably the best card issuer -- or one of the best at least -- and Sila is probably the best at ACH in the US now. It depends on approach and strategy. I much prefer to build less and be better at it than try to do everything and suck at five things at the same time.

    There are two competing visions in this space—one being the consolidated suite of products in BaaS and the other being this approach where everybody picks individual solutions and builds their own stack. Long-term, which do you see as being better able to consolidate demand?

If I'm being honest, I expect to see another ten consolidated platforms, at least, launch in the next 18 plus months. And I expect to see not quite as many point solutions, but more of those as well. Then I expect all of them to grow rapidly for the next decade.

The best example to look at is -- I don't know what to call it, fintech 1.0? Look at Fiserv, TSYS, Jack Henry, FIS, Visa DPS, and the whole wave of financial technology that got started way back, as early as in the late 1960s. You look at the volumes of those guys: any one of those guys does way more volume than the entire fintech industry combined. Those guys are massive. They're just built on mainframe technology that's 40 years old. You'll find that I just named a few large companies, but there's another 50. And you see the same things across all of them.

What you end up seeing is that a lot of the smaller credit unions and community banks end up being a Fiserv shop or an FIS shop or a Jack Henry shop. They have a Fiserv core, and they just use everything that Fiserv sells them. You look at the larger banks, and they tend to be much more bespoke, because they have massive scale. It's not just about six branches and 200,000 users; it is about 2,000 branches and 10 million users and transacting multiple billions a day. At that point, they're like, "No, an ACH system that is good enough is just not good enough. It has to be the best in class. If I get a better ACH system, I can have 200 less people who are doing ACH jobs, and I will pay for that and source it and install it." That's how they end up doing it. The problem for them is just that their entire tech stack is three generations behind.

I expect that fintech 2.0 will not end up looking exactly the same, but will follow some of the same lines. I don't think consolidated platforms will eclipse everybody, and I don't think point solutions will eclipse everybody either. There are folks who will gravitate towards point solutions, and those who will gravitate towards consolidated. Then there are folks who'll migrate from one to the other.

I do think that for the largest companies and the scale players, they need best of breed. It's hard because you get to this point where return processing, chargeback processing, or these tiny things that you don't think about on day one, become massive. Suddenly, you have 50 people in a room somewhere just solving that problem at scale.

I broadly prefer Bo's approach and align with Bo. Of course, they're partners of ours, so there's no surprise there, right?

I'm curious how your customer base looks, both from a startups versus bigger companies perspective and also from the perspective of products that just want to embed some form of money movement versus more pure fintech players.

We see all of those. We have a hundred plus customers now. Across those, there are probably 25 different use cases. But even when you say use cases, it's like industries versus use cases. We have four customers that I can think of off the top of my head in real estate. Yet none of them is doing the same thing. In fact, we've introduced some of them, and they might end up partnering with each other. You look at real estate, you're like, "That's a multi-trillion dollar industry." Of course, there are a hundred different sub-sectors of that, each of them billions in size. So how much space? There could be a thousand real estate startups in the US. There probably already are. They're doing everything from rent payments, to NFTs for land, to mezzanine debt in commercial real estate. And that's just scratching the surface.

We don't have that many traditional neobank customers, but we have pretty much a little bit of everything. As I said, if you looked at numbers, I'd say probably 70/30 fintech/crypto. If you looked at volumes, it's probably more like 50/50 or 60/40. But that changes pretty rapidly, because new customers, they'll sign up, and then six months later, they're growing explosively. It's like, "Hey, okay." That's all good.

Does Sila focus on any one industry?

We really aren't focused on any one sector. We accept all customers. It does make it a huge scaling challenge because we do a lot of the compliance. We have a partner bank -- it's Evolve Bank & Trust, the same bank that works with Synapse and Stripe and a bunch of others. Most of our customers have never spoken to anybody at Evolve and probably never will need to, because we do the onboarding, we do the compliance, we embed as much of that in the tech as we can. We have a fairly large CX and Ops team. That's part of the value proposition. You don't have to be a compliance expert on day one. Of course, when you get to a massive scale, you probably will become one, and that's part of the growing journey. But we'll support you through that.

You mentioned that there will be folks who gravitate towards platforms, folks who gravitate towards point solutions, folks who migrate. Are you seeing folks migrating in your own experience?

We've seen folks switch from Synapse to us. We've seen folks switch from Dwolla to us. We haven't seen anybody switch from Unit or Rise, but honestly, those guys are so new, we've barely had time for that to happen. But yeah, we do see folks migrate off to us. It's relatively small because it's a boom right now. There are thousands of new fintech companies being created. Also, it's much easier to build the first version of your app on Sila and stay there than to do a migration, which tends to be complex and has many moving parts.

But the ones who migrate tend to have more scale, because they're already live. They already have more customers and everything else. It's conventionally like, "Hey, sign up somebody who just raised a pre-seed or seed round, or did a token sale, and has the funding and the team. And they spin up and get live in six weeks. That's awesome. Off to the races. Best of luck. Hopefully, you take over the world and fix all the problems." But then, there are the folks who are larger, and they transition. That takes 12 to 18 months sometimes, because you have to scale, and it's tricky to transfer over customers' money and do it seamlessly in the background without missing a beat.

You mentioned the transaction volume is roughly half/half crypto/fiat. Could you talk about why it's important for Sila to support crypto?

That number was a best guess. I'm not sure that it is necessarily right. It even gets hard to be fully consistent-- what is fintech and what is crypto at some point? We have a bunch of crypto customers whose backends look very fintech, and some fintech customers whose backends are becoming more crypto. It's like, which one do you classify where? But we have a lot of both.

Crypto is important because I think it's one vision of the future of finance. It's not necessarily the only vision, but it is a powerful and growing vision. There is a lot of activity happening and beginning to change. Just within the last 12 months, one of the biggest things we've seen is more and more fintechs who were "traditional fintechs" -- whatever that means -- who are now like, "Hey, I'm now interested in offering crypto capabilities." Whether it's very straight forward, like the purchase and sale of cryptocurrencies or NFTs, or doing more smart contracts-- there's a lot of interest, for example, in doing financing through blockchain-based platforms and getting access to liquidity or providing liquidity and earning yield.

All of that is of huge interest because the traditional markets are large and way larger than anything in the crypto space. It's really hard to get access to that. If you're an early stage company in crypto, just go on a blockchain, plug into a smart contract, and there you go.

I don't know if I answered your question there, but I never really viewed fintech and crypto as two different worlds. That's like saying that banking and fintech are two different worlds. No, they're very different people and they solve problems in different ways, but honestly, they are solving the same problems. I think the same way about fintech and crypto: it's different people and different ways of solving problems, but actually many of the same problems.

One way of looking at BaaS is that your customers are a venture portfolio, and you're placing bets on which ones will explode. For Marqeta, it's Square and Instacart, and so on. What's your take on that, and does it resonate with how you think?

It does, but I prefer not to place bets at all. I prefer to just say, "Hey, if it's compliant and you have the basic funding required to put together a dev team and build and ship and pay our fees, we will take all of that." In that way, our approach is probably more like a shortcut. We serve as many customers as we can. Customers do fail. We've had two or three companies that got acquired or shut down after building and shipping. You just get used to it. It's part and parcel of the business. At the same time, we also have customers that took off and grew explosively as well. In that way, it does end up looking like a venture portfolio, if you looked at our transaction data.

On the flip side, that's not how our business model works. We are independent of our customers’ successes. It's not like Marqeta placed a bet on Square. They just signed up everybody, and one of them became Square. Cash App wasn't what it is now back in 2012 when Marqeta signed them.

You will end up looking like that. You will end up having concentration from large customers just because they grow 10x a year for a decade. But I think if you look at Marqeta, they have a long list of smaller customers, too. And so do we.

Disclaimers

This transcript is for information purposes only and does not constitute advice of any type or trade recommendation and should not form the basis of any investment decision. Sacra accepts no liability for the transcript or for any errors, omissions or inaccuracies in respect of it. The views of the experts expressed in the transcript are those of the experts and they are not endorsed by, nor do they represent the opinion of Sacra. Sacra reserves all copyright, intellectual property rights in the transcript. Any modification, copying, displaying, distributing, transmitting, publishing, licensing, creating derivative works from, or selling any transcript is strictly prohibited.

Read more from