Embedded Finance Is Quietly Eating Banking's Lunch
The Real Story: Non-Financial Companies Are Now Banks
Square's merchant segment generated $801 million in gross profits in Q4 2022—a 22% jump—not from processing payments, but from embedded banking. Shopify merchants don't just sell products anymore. They get paid instantly into Shopify bank accounts. They access business loans directly from the dashboard. They manage cash flow without leaving the platform.
This isn't a fintech trend. This is the wholesale restructuring of how commerce works.
Embedded finance—when non-financial companies integrate lending, banking, payments, and insurance directly into their platforms—is now the fastest-growing segment of financial services. The global BNPL market alone is projected to expand from $10.87 billion in 2025 to $14.09 billion in 2026, with some forecasts suggesting the broader embedded finance market could hit $230 billion by next year.
But the real story isn't about growth rates. It's about who's winning, why banks are losing control, and what this means for commerce.
Why Non-Financial Companies Are Better at Finance Than Banks
The advantage isn't technology. It's context.
When you need working capital, you don't want to log into a bank portal and fill out forms. You want to solve it where you're already working—in your Shopify dashboard, your Square terminal, or your Amazon seller account.
Shopify Balance solved this by making financing feel native to the platform. Merchants see their cash position in real time. They can access loans based on actual sales data the system already has. The friction that made traditional lending slow—underwriting delays, document collection, manual verification—disappears when the lender already has your operational data.
Square Checking works the same way. Drivers using Lyft Direct get paid within moments of completing a ride, directly into a branded debit card with cash back on gas and groceries. The product isn't a bank account. It's a solution to the actual problem: gig workers need immediate access to earnings.
Amazon Pay Later crossed 10 million signups by 2025. Not because it's better than Affirm or Klarna—it's because it's already there. When you're checking out on Amazon, why go to a different app for financing?
This is the embedded finance advantage: the financial product isn't a separate thing you go get. It's part of what you're already doing.
The BNPL Reckoning
Buy-now-pay-later exploded because it solved a real problem: consumers wanted installment payments without the stigma of credit cards or the complexity of traditional lending.
But BNPL is also where embedded finance's fragility becomes visible.
PayPal dominates BNPL adoption at 56% of users, followed by Klarna and Affirm at 38% each. The market is consolidating. Smaller BNPL players that couldn't achieve scale or profitability are disappearing. The ones surviving are either integrated into existing platforms (PayPal, Amazon) or have become embedded themselves—Klarna now partners with retailers to offer financing directly at checkout, not as a standalone app.
The lesson: standalone embedded finance products don't win long-term. Embedded finance wins when it's truly embedded—when it's so integrated into the core product that removing it would break the experience.
Insurance Is Next
If embedded lending is proven, embedded insurance is the frontier.
Apple and Asurion partnered to offer device insurance directly in the Apple ecosystem. You don't buy insurance from Asurion. You buy protection when you're buying an iPhone. The insurance is invisible until you need it.
This model is spreading. Retailers are embedding product protection. Travel platforms are embedding trip insurance. The pattern is identical to lending: insurance becomes a feature, not a separate purchase.
The challenge: underwriting and claims. Lending can be automated. Insurance requires judgment calls. But as AI improves risk assessment, this barrier collapses.
What Banks Are Losing
Traditional banks built their business on being the intermediary. They held deposits, made loans, issued cards. They owned the relationship.
Embedded finance inverts this. The tech platform owns the relationship. The bank becomes infrastructure—the licensed entity that provides regulatory compliance and capital, but stays invisible.
This is why Stripe Treasury and Unit exist. They're not trying to be banks. They're trying to be the plumbing that lets every SaaS platform, marketplace, and e-commerce site offer financial products without building a bank from scratch.
The problem for traditional banks: they're not equipped to be invisible. Their entire business model depends on direct customer relationships. When a merchant's primary financial relationship becomes Shopify, not their bank, the bank loses the ability to cross-sell, to build loyalty, to understand customer behavior.
The Regulatory Question Nobody's Talking About
Embedded finance is growing faster than regulation can keep up. When Amazon offers lending to sellers, who's responsible for fair lending compliance? When Shopify issues business loans, is Shopify a lender or just a platform?
The EU AI Act Phase Two and emerging US regulations are starting to address this, but the frameworks are still fuzzy. This creates both risk and opportunity. Platforms that build compliance infrastructure early will have a moat. Platforms that ignore it will eventually face enforcement action.
The Actual Play for 2026
Embedded finance isn't winning because it's innovative. It's winning because it's convenient and because platforms have the data.
The winners will be:
Platforms with existing user bases and transaction data. Shopify, Square, Amazon, Stripe. They have what lenders crave: real-time visibility into customer behavior. They don't need to guess creditworthiness. They can see it.
Infrastructure providers that stay boring. Unit, Stripe Treasury, and similar platforms that let any company become a financial services provider without the regulatory headache. These are the picks-and-shovels plays.
Vertical-specific solutions. Generic BNPL is commoditized. But embedded finance tailored to specific industries—construction financing, healthcare billing, freelancer payments—still has room to move.
The thing that won't work: Standalone embedded finance products trying to compete with platforms that already have the user base. The market for "better BNPL" is over. The market for "financing that's part of your workflow" is just getting started.
The Uncomfortable Truth
Banks had every advantage: capital, regulation, customer trust, operational infrastructure. They could have owned embedded finance. Instead, they treated it as a distribution channel and let tech companies become the primary relationship.
Now Shopify merchants think of Shopify as their financial partner, not their bank. That's not a technology problem. That's a business model problem. And it's not getting fixed.
The question isn't whether embedded finance will reshape commerce. It already has. The question is whether banks will figure out how to compete in a world where they're infrastructure, not the relationship. The data so far suggests they won't.