UnAPIable

“Americans mythologize competition and credit it with saving us from socialist bread lines. Actually, capitalism and competition are opposites. Capitalism is premised on the accumulation of capital, but under perfect competition all profits get competed away.

The lesson for entrepreneurs is clear: if you want to create and capture lasting value, don’t build an undifferentiated commodity business.”

- Peter Thiel, Zero to One

The life cycle of the average stablecoin neobank goes like this:

1. Integrate Bridge, Rain, Fireblocks, Sumsub, and other vendors.
2. Launch the exact same app that everyone and their mom is launching, but with a UI twist¹.
3. Offer an eye-catching incentive like meaningful cashback and referral rewards.
4. Get a lot of signups from users who want to milk the incentives.
5. Reduce said incentives and maybe start charging for some features. You want to show VCs you have a business.
6. A new app that just got funded starts doing step 3.
7. See all your users churn to that app.
8. Your growth graph is now VC-toxic.
9. You decide to either shut the app down, or do what everyone is doing lately: pivot to business banking.

There are a lot of good reasons to launch a stablecoin neobank. There is real demand for the product, there are established vendors that can sell you almost all the infrastructure you need to ship quickly, and there is a powerful narrative around giving people new ways to access and move money that you, a savvy entrepreneur, can turn into an investable story.

“It’s all about GTM,” someone whispers.

“Shhh. You need to find a niche,” someone says. Stablecoins for freelancers in Latin America. Stablecoins for migrants. Stablecoins for dentists.

“Actually, you should go broad from day one,” someone else says.

Of course, you still need to build a good, usable product. But Bridge already did the heavy lifting of building the banking relationships that turn bank money into stablecoins. Rain already did the incredible work of becoming a Visa Principal Member. Sumsub already built the infrastructure to verify IDs in 170+ countries.

Your job is to put the API puzzle together in a way that looks different from what’s already out there. Claude can help a lean team ship this quickly, and even compliance can be covered surprisingly fast with the right software and a few good hires.

You’ve also already done the work of finding the soundbite that explains why your version is different. You have a wedge, after all. Or at least some made-up version of one that helps you sleep at night and keeps you from looking dumb when an investor asks why this isn’t the same neobank they’ve already been pitched 20 times this month.

So you think about it again and arrive at the obvious conclusion: this is, indeed, a distribution play.

Yet somehow, it’s easier to distribute a stablecoin neobank than it is to distribute water. And people actually need water to live.

On any given day, you can open Twitter and find people sharing screenshots of their Apple Wallets packed with Rain-issued cards. And every time Bridge integrates a new offramp rail, I get 27 push notifications from apps telling me about what they just “built.”

The problem is that everyone reached the same conclusion. Everyone is throwing money at the same users with the same ads, referral bonuses, cashback, and whatever growth hack works this month. And sure, you’ll get signups. You might even get a lot of them. People love to take whatever you’re giving them and move on.

And it doesn’t really matter if your version has a $499 premium membership and is marketed as the high-end one. Users open the other apps and see the same card, the same rails, the same features, and, if they read the small print, the same vendors underneath.

Which is why I believe that if your team isn’t continuously solving the incredibly hard, unAPIable problems that will make users stick with you because nobody else does them as well, while also figuring out how to make money from this in a new way², you most likely won’t survive just by integrating stuff.

But before we talk about surviving, we need to talk about Payoneer and Wise.

Payoneer

If you’ve ever done online gigs from Latin America, Africa, Asia, or even parts of Europe, or paid people who did, you’ve probably heard of Payoneer.

Before Zach Abrams and Sean Yu founded Bridge, Payoneer was pretty much the only way people in those regions could get paid and spend their money through something that resembled a neobank, even though Payoneer never really advertised itself that way.

Payoneer did the genuinely hard work of somehow getting Bank of America to issue virtual accounts to people in pretty much every non-sanctioned country and issuing prepaid Mastercard cards out of Ireland through its own entity after Wirecard collapsed. There was no Bridge. There was no Rain. There was nobody handing them a nice API and wishing them luck.

And oh boy, did that work pay handsomely.

If you were a worker, you already know this: Payoneer's fees sucked. Almost everything had some obscene fee attached to it.

Getting paid? You could easily see 10% of your money disappear in fees.

Using the card? 4%.

Sending money home? Get ready for astronomically high currency conversion fees, and a few other wire fees you only discovered after they had already taken your money.

Inactivity fees, activity fees, account maintenance fees. Fees, fees, fees. Yeah, you get it.

And don’t even get me started on support. It was, as many of you probably remember, basically nonexistent.

And the crazy part? None of this really mattered. Payoneer could get away with beating your dog and you’d still have to use them because, for someone like me born in Paraguay, it was the only way to receive an ACH transfer from something like Google AdSense. And that same reality applied to millions of people around the world.

No wonder they went public on Nasdaq and, at their height, were worth almost $4 billion.

Now compare that to today’s market.

If a stablecoin neobank even dares to ask you to pay for a virtual account, nobody signs up. Reduce cashback from 3% to 1%? People leave. Add a charge to cover gas fees? They’ll just move to the next app.

There’s basically no tolerance for friction because there is always another app. And signing up has never been easier. You don’t even need to have your ID with you every time. Sumsub just needs your email and pulls the one you already used for the other 10 neobanks you downloaded today.

That’s a very different world from the one Payoneer grew up in. Payoneer actually had a differentiated product, and on top of that, incredible distribution.

Payoneer understood that distribution only mattered if it reached people actually moving money. They didn’t want millions of dormant accounts. They wanted to be the first option wherever global online work was happening.

If you signed up for platforms like Upwork, Fiverr, Toptal, Shutterstock, Avangate, or even payroll platforms like Deel and Remote, Payoneer was always there. They had co-branded cards with many of those companies. Coca-Cola-level ubiquity.

But guess what? Even that incredible power to be everywhere wasn’t enough to retain users forever. Once Bridge opened the gates to the neobank craze and made it possible for its customers to offer much more affordable products, Payoneer’s usage among workers dropped dramatically. The product got so outdated that earlier this year they had to announce they too were going to start working with Bridge.

Wise

This is a niche story, but it’s important because it shows you what unAPIable actually looks like.

Pretty much everyone working remotely for a company abroad has a Wise account. But since 2023, Wise has not allowed new users in many jurisdictions, including most of Latin America, to open USD account details with them.

And you’d think that with every stablecoin neobank plugged into Bridge now giving you a Lead Bank virtual account, you could finally ditch Wise, right?

Well, you’d be wrong.

Wise accounts became so valuable that people started selling their old accounts. Entire services appeared to help you open one as if you lived in another country, like Spain.

Even if you’re a remote worker who really loves your new neobank, there’s a good chance you’re still forced to use Wise for certain things simply because Wise is so damn good at them.

This is basically the opposite of the Payoneer story, and I’ll show you why.

Let’s say you’re a remote worker in Colombia. The same example applies to pretty much any country where the local currency isn’t USD.

You make $5,000 a month and want to move that salary from Deel or your neobank to your local bank account.

You could use a local offramp, but then you have to convert your dollars into local currency. And they usually get you twice: first with a conversion fee, and then with an exchange rate worse than the one you could get directly from your bank. On a $5,000 salary, what looks like a tiny fee can easily turn into $200 lost by the time the money lands.

So you need SWIFT if you want your money to arrive in the same currency you were paid in and pay only a fixed fee.

SWIFT has become a big part of the neobank conversation this year, and this is roughly how it works with almost all of them:

Your neobank charges you $30 for the SWIFT transfer it gets from a vendor. It gets sent as SHA, which means other entities touching the transfer can take their fees along the way. An intermediary bank might take another $30, and your local bank might take another $40.

That’s $100 gone just to move your own money home.

Wise understood something very basic about its customers: they are extremely price-conscious. So instead of just buying SWIFT from some provider and accepting whatever pricing came with it, they went much closer to the infrastructure themselves. Wise has its own SWIFT/BIC connectivity, works directly with Swift, and will send transfers as OUR, which means the usual intermediary and beneficiary fees can’t be taken out of your money along the way.

So Wise charges you $20, but your money gets there intact.

And because Wise controls much more of its own infrastructure, it can keep pushing prices down in ways that are very hard for a neobank relying on the usual-suspect vendors to match.

And the most beautiful part? They actually make money doing it.

Let’s recap:

Early Payoneer shows you the kind of stuff you can get away with when there is no equivalent alternative.

Payoneer built a powerful moat and had insane distribution, and still eventually lost ground to the frenzy Bridge helped create.

Wise closed the door on millions of people and somehow still has them begging to get in because they got so good at what they do that users choose them even when alternatives exist.

And yes, it is possible to make money in this industry without bribing users with rewards.

Surviving

So where does that leave the average stablecoin neobank?

You can’t charge like Payoneer because users have 20 other options retargeting them, and thanks to the magic of VC money, those apps will make damn sure the user knows they exist. You can’t assume referral rewards, high cashback, and spending a ton of money on ads will save you because switching has never been easier. And whatever you’re thinking of adding from a partner, whether it’s eSIMs, travel insurance, or rails to every country on earth, everyone else gets the same thing.

And there is also the much uglier problem of unit economics. What’s the point of spending $20 to acquire someone if you never make that $20 back before they leave because a newly seeded app prices the exact same rail one cent cheaper or decides to absorb card FX?

That $499 membership will not save you either. You might think users won’t earn enough cashback to make the math work in your favor. Trust me, they absolutely will. You’d be surprised at the ways people will route around your MCC restrictions for the sole purpose of making money off you and teaching your CFO a few lessons about game theory.

Then what the hell are you supposed to do?

Well, despite what is starting to sound increasingly pessimistic, I do believe there is room for winners here. But you need to raise the bar and start doing unAPIable stuff if you actually want to be different. Fast.

Now, that’s easier said than done because it’s incredibly easy to fool yourself. You come up with that soundbite we talked about earlier, and suddenly your wedge makes you feel different without actually being different. You start thinking your shiny, heavy metal Visa Infinite card is somehow going to separate you from the hundreds of other apps doing almost the same thing.

Well, sorry to break it to you, pal, but the sad truth is that you are competing with everyone.

And no, having better support plus cool marketing won’t save you. Small thinking is how you ended up in a commodity business in the first place.

If you want users to behave differently and not churn, you need to give them a very good reason to. There is no room to play it safe. Instead of spending the afternoon in some vendor’s mintlify docs looking for one more thing to plug in and launch so it looks like you’re moving fast, follow the counterintuitive path: go find the problems nobody has built an API to solve.

To do what Payoneer did, you had to convince Bank of America to give virtual accounts to people all around the world. Do you think it was easy for Wise to get this incredibly old institution in Belgium to give them access to infrastructure usually reserved only for banks?

And that’s really the test. Can you build something so good that leaving you starts to feel irrational?

I’ll give you an example.

Over the past few months, I’ve gotten several emails from different neobanks asking me to move the money I receive through Deel into their apps.

I understand why. Someone with recurring income is obviously a much better customer than someone who deposited $20 to get your little token reward.

But why on earth would I move it?

I already told you Wise is more convenient for me. So if you want me to change that behavior, give me something meaningfully better than whatever shenanigans Wise is pulling to save me hundreds of dollars every two weeks.

Otherwise, your attempts at winning me over aren’t going to seriously fly.

Sure, you might get some people to try it. But if there’s no strong reason for them to keep repeating the behavior, they can switch at any moment based on some slack recommendation.

Maybe if you want people getting paid through Deel, you need to do some research about my country and discover that every time I get paid I have to deal with some insanely complicated tax scheme because I’m being paid as a contractor instead of through an EOR. So you call some local accountants, turn my pain into a product, and now I have an actual reason to move my money to you.

That’s the kind of problem I mean by unAPIable.

And this work isn't going to look like building software. It’s calls, contracts, some regulators that don’t work after 1PM, introductions, “just circling back”, waiting, trying again. That’s why almost nobody does it, and also why there is opportunity in it.

Once you do it, though, you need to keep doing it multiple times. Whatever hard thing you build today, someone may be calling it infrastructure in their YC application a couple of years from now.

And I don’t want you to get the wrong lesson from all of this. I’m not saying that instead of using Bridge for your core money flows, you should spend the next two years trying to convince some sleepy community bank in Nebraska to give you your own unique banking relationship.

In fact, that can be incredibly dangerous. Just a few weeks ago, multiple neobanks that relied on card issuance from a lesser-known partner suddenly found themselves with cards that simply stopped working.

Not knowing who to buy from can be just as risky as trying to build something you were much better off buying. Bridge, Rain, Nium, and others have already commoditized certain parts of this business, and there is no point trying to do what they do, unless you want to compete with them, and that’s a whole separate topic.

So just go after the stuff they still don’t solve, may never solve, and that your users actually care about. You’ll end up with a product that’s different from everyone else’s.

If you keep the pace, everyone else will have a real problem. They will either pivot, shut down, or complain about how unfair it is to play against you.

This may require raising a lot of money. Or it may require operating like a cockroach and doing the hard stuff one market at a time. Fintech has produced successful examples of both.

As I finish this, I’ve realized people are usually lazy and probably were expecting to read some actual ideas. Fine. Here are three areas I’ve been thinking about lately.

Help people protect their money

Before the stablecoin neobank rush, one of the big selling points in global neobanking was actually something super boring: deposit insurance.

Most people outside the US have access to significantly lower levels of government-backed deposit insurance. In Paraguay, for example, the limit is around $20,000. If you’re a well-paid remote worker, you quickly realize that isn’t much.

When Synapse launched its Global Cash program, one of the things entrepreneurs quickly figured out was that they could sell access to a real US bank account with $250,000 of FDIC insurance to people all over the world.

Pretty much everyone who thought of targeting people abroad was marketing the FDIC angle. One of them even got a cease-and-desist letter from the FDIC because of how it was talking about it.

Of course, Synapse later failed spectacularly and users are still fighting to recover their money, so maybe don’t copy that part.

And sure, you may be thinking that giving people access to Bitcoin or some crypto protocol already solves part of this.

But there’s a much more interesting example: Xapo.

Bitcoiners spent years telling everyone to “be your own bank.” Then Wences Casares, of all people, went and built an actual bank that takes bitcoin and stablecoin deposits.

You know how hard becoming a regulated bank in Gibraltar must have been? There is no API for that.

Today Xapo gets away with charging $1,000 a year for membership, and people actually pay it, while attracting the crème de la crème of customers. And they can do it because your fiat gets government-backed protection from the Gibraltar Deposit Guarantee Scheme. Their bitcoin custody is hard to match. You can set up beneficiaries without having to teach your family how cold wallets work. And if something goes wrong, there is an actual regulated institution there.

It just feels like a much safer place to make the grown-up decision of where to keep your money.

And it’s a very different pitch than “here is another dollar account and a Visa card.”

It’s not only crypto bros sitting on millions. Think of our Colombian guy earning $5,000 a month from a remote job. A year goes by and he may have $20,000 saved.

People are willing to pay for peace of mind.

Aggressive local expansion

I think the rideshare market shares a lot of similarities with this industry.

Uber would enter a country and basically figure out the local stuff later. The company, the licenses, whatever the government wanted from them, a lot of that came after. Sounds familiar?

Eventually governments caught up and said: if you want to serve our people, you need to play by our rules.

That obviously made expansion harder. But it also created an opportunity to crush competition.

I think the same thing could eventually happen with stablecoin neobanks. Governments generally don’t love having millions of their citizens banking with some entity incorporated three continents away. Many countries already have rules that make parts of what neobanks are doing questionable at best, and if you log into the Meta Ads Library, you can already see certain apps violating solicitation rules in multiple jurisdictions.

At some point governments are going to make everyone set up locally or ban them. When that happens, whoever is willing to do the hard work might suddenly have a huge advantage

Set up the entity. Get the licenses.

Most of your competitors probably won’t want to deal with any of that.

And there’s another upside: once you actually become part of the local financial system, you can start doing things that make a lot more money than charging people for transfers, like lending³.

Companies like ARQ (DolarApp) are already taking this approach. Their expansion might look painfully slow next to the neobank launching in 40 countries, but this might be one of those tortoise-and-hare situations.

Network effects

I did say I think this market shares a lot with rideshare, didn’t I?

Anyone can vibecode an Uber or Lyft competitor today. Their moat isn’t the software. It’s the marketplace dynamics. Nobody wants to use a rideshare app with no drivers, and nobody wants to drive for an app with no users. Solve that problem and boom, it doesn’t really matter if 10 other people launch rideshare apps tomorrow. Breaking into your network becomes incredibly hard.

The obvious attempt at doing this with neobanks was the “global Venmo” idea. Transfers between users are free, so surely everyone will invite their friends and eventually we’ll have this giant global payment network.

Except the dynamics are completely different.

Venmo and Cash App work because everyone around you already has them. Your friends have them. Your roommates have them. The guy you owe $20 to probably has one too.

That’s not really what happens with a stablecoin neobank. I may use one of these apps because I get paid in dollars. But who am I sending money to there? My sister doesn’t have it. My friends don’t have it. Most people I know locally don’t even know what USDT is.

So maybe copying Venmo was the wrong idea.

The interesting network may be somewhere else. A freelancer and his clients. A company and its remote workers. A family sending money back and forth. People in some online industry that already transact with each other all the time.

Félix Pago is interesting here. They didn’t try to build another network from scratch. They built on top of WhatsApp, where millions of these relationships already existed.

There are probably really interesting ways to do more of this.

We just haven’t seen enough people trying.