B2B Embedded Finance in 2026: Where the Revenue Actually Is

B2B embedded finance is the practice of building banking products — payments, lending, accounts, cards, insurance — directly into the software that businesses already use to run their operations. A construction-management platform that lets a contractor get paid the day the job is signed off, rather than ninety days later, is doing embedded finance. So is the restaurant point-of-sale system that offers a cash advance against next month's covers. The thesis stopped being speculative some time ago. The open question in 2026 is narrower and more interesting: who captures the margin, and who is just carrying the risk.

That distinction matters more than the breathless market-size forecasts suggest. Plenty of software companies have switched on a payments feature, booked a slice of interchange, and declared themselves a fintech. Far fewer have understood what they actually signed up for when they did.

What is B2B embedded finance?

Embedded finance describes financial services delivered at the point of need, inside a non-financial product, rather than through a separate trip to a bank or a standalone fintech app. The "B2B" qualifier matters because the economics, the customers, and the risks differ sharply from the consumer version most people picture.

Consumer embedded finance is the buy-now-pay-later button at checkout. B2B embedded finance is the vertical software platform — for dentists, freight brokers, law firms, gyms — that adds a business bank account, a corporate card, invoice financing, or payouts to the workflow its customers are already living in. The financial product is not the destination. It is a feature of the thing the business came to do anyway.

In 2019 the Andreessen Horowitz partner Angela Strange gave a talk with a title since quoted to the point of cliché: "Every company will be a fintech company." The line was roughly correct. The instructive word, though, was "company," not "fintech." Strange's argument was that financial infrastructure would become a set of APIs any software business could assemble, the way Stripe had made card acceptance a few lines of code. That has largely happened. What the slogan glossed over is that assembling the APIs is the easy part. Owning the consequences is not.

Why B2B beats B2C for embedded finance

The consumer side gets the headlines; the business side has the better unit economics. Three structural reasons explain why.

First, transaction sizes are larger and recurring. A dentist's practice-management platform processes meaningfully bigger payments, more predictably, than a consumer app nudging someone to split a £40 dinner. Larger flows mean more interchange and more financeable receivables per customer.

Second, the relationships are sticky to the point of being adhesive. A business that runs its scheduling, invoicing, payroll, and now its banking through one platform is not switching on a whim. Embedding finance deepens a lock-in that vertical software already enjoys, and it does so while raising the revenue the platform earns from each account.

Third, the data is better. A platform that already sees a contractor's job pipeline, their payment history, and their customer concentration can underwrite a loan to that contractor far more cheaply and accurately than a bank starting from a credit-bureau file. This is the same advantage reshaping consumer credit, where AI underwriting trained on cash-flow data is displacing the traditional scorecard — except in B2B the platform owns the data outright.

Research from a16z, Bain, and others has put the per-customer revenue uplift from embedded finance at several multiples of the underlying software fee. The precise figure varies by vertical and by who is doing the estimating, and it is worth treating directionally. The direction, however, is not in dispute: financial products tend to earn more per customer than the SaaS subscription they ride on top of.

The four revenue streams

When a software platform "embeds finance," it is usually reaching for one or more of four distinct income lines. They are not equally easy, and they are not equally safe.

Interchange. Issue a card to your business customers and you earn a share of the fee merchants pay when that card is used. This is the most common entry point because it is the least demanding to stand up. It is also the thinnest, and it depends on a quirk of US regulation discussed below.

Lending margin. Offer working capital, invoice financing, or a charge card with a revolving balance, and you earn the spread between your cost of capital and what the borrower pays. This is where the real money sits, and also the real risk: lending margin is only profitable if your loss rates are lower than your spread, and loss rates are a function of an economic cycle you do not control.

Deposit float. Hold your customers' balances and you earn interest on the float, particularly in a higher-rate environment. Shopify, Toast, and others have leaned into this. It is attractive precisely because it requires no credit risk — but it does require customers to keep meaningful balances on the platform, which not every vertical supports.

SaaS uplift. The subtlest of the four. Embedding finance can raise the value of the core subscription itself, reduce churn, and justify a higher price. This revenue never appears on a "fintech" line in the accounts, but it is often the most durable, because it compounds the platform's existing moat rather than bolting on a new and riskier one.

The platforms that struggle are usually the ones that chased interchange because it was easy, mistook a thin and regulation-dependent fee for a business model, and never built toward the lending or deposit lines where the economics actually work.

The infrastructure layer: who provides the rails

Almost no software company holds a banking licence. They rent one. The Banking-as-a-Service, or BaaS, layer is the set of providers that sit between a software platform and a chartered bank, turning regulatory permissions into API calls. Stripe Treasury, Unit, Synctera, Treasury Prime, and the card-issuing specialist Marqeta are among the better-known names in the US; Swan, Solaris, and Weavr serve the European market under a different regulatory regime.

Behind every one of these middleware providers is an actual bank holding the actual deposits. In the US these are frequently small, community-scale institutions — Evolve Bank & Trust, Cross River Bank, Coastal Community Bank, Lincoln Savings — rather than the household-name giants. There is a specific reason for that, and it is one of the more revealing details in the whole sector.

Under the Durbin Amendment, banks with under $10 billion in assets are exempt from the cap on debit interchange that applies to larger institutions. A small partner bank can therefore earn — and share — materially more interchange than a money-centre bank could on the same transaction. The entire economics of card-based embedded finance in the United States rests partly on this threshold. When you read that a SaaS platform has "partnered with a bank you've never heard of," that is usually not an accident of obscurity. It is the point.

The Synapse warning: what the 2024 collapse taught the industry

For most of the 2010s the BaaS story was told as pure upside. Then Synapse happened, and the industry got a hard lesson in where the risk had been hiding all along.

Synapse was a BaaS middleware provider — exactly the translating layer described above — sitting between dozens of fintech programmes and their partner banks, chief among them Evolve Bank & Trust. In 2024 Synapse collapsed into bankruptcy. When it did, the ledgers reconciling which end-customer owned which dollar at which bank turned out to be a mess. Funds were frozen. Ordinary people and businesses who believed their money sat safely in an FDIC-insured account found they could not reach it for months, because FDIC insurance protects against a bank failing, not against a middleware company losing track of whose money is whose.

The episode forced a distinction the marketing had blurred. A software platform can embed a "bank account" without the regulatory accountability of being a bank — but someone, somewhere in the chain, still has to keep an accurate, reconciled, real-time record of every customer's balance. When that responsibility falls between a middleware provider and a partner bank, with each assuming the other has it covered, the gap is where customer money disappears.

US regulators noticed. Bank-fintech partnerships have since drawn closer scrutiny from the FDIC, the OCC, and the Federal Reserve, with particular attention to the small partner banks whose interchange advantage made them the backbone of the model. For any software company weighing an embedded-finance push in 2026, the compliance burden is no longer a footnote — it is a first-order design constraint, and it is exactly the kind of obligation a serious RegTech stack exists to carry.

The European picture

Europe arrives at embedded finance from a different regulatory starting point, and the differences are not cosmetic. There is no Durbin-style interchange carve-out; interchange is capped more tightly and uniformly across the bloc. Removing that single subsidy changes the centre of gravity. European embedded finance leans toward accounts, payments, and lending as the lead products, rather than the card-interchange play that anchors so much of the US model.

The licensing route differs too. Where US platforms lean on partner-bank sponsorship, European programmes frequently run through electronic money institutions — EMIs — under the scaffolding of PSD2, the second Payment Services Directive, which mandated open-banking access and strong customer authentication across the union. Providers such as France's Swan, Germany's Solaris, and the UK's Weavr built their businesses inside this framework, offering accounts and cards as regulated e-money rather than as deposits held at a sponsor bank.

That distinction — e-money versus deposit — carries real consequences. E-money is not covered by deposit-guarantee schemes the way bank deposits are. Instead, EMIs must safeguard customer funds, typically by holding them in segregated accounts at a credit institution. The protection is genuine, but it is not identical to deposit insurance, and the Synapse episode has sharpened European regulators' attention on precisely this question: what happens to customer money when an intermediary, rather than a bank, is the thing that fails.

The rules are also moving. PSD3 and the accompanying Payment Services Regulation, working their way through the EU legislative process, will tighten requirements around fraud liability, open-banking data access, and the obligations of payment institutions. For any platform building embedded finance for a European customer base in 2026, the safe assumption is that the compliance bar rises, not falls. The UK, post-Brexit, is charting a partly divergent course under the FCA, adding a third regulatory dialect to a market that already spoke two.

The strategic conclusion mirrors the American one, reached by a different road: the licensed, well-governed balance sheet at the back of the chain is becoming more valuable, not less, as the regulatory cost of doing this properly climbs.

Where the real money is in 2026

Strip away the hype and the durable winners cluster in two places.

The first is vertical SaaS platforms with deep workflow ownership and a customer base that genuinely needs capital. Toast, in restaurant software, offers Toast Capital. Shopify built Shopify Capital and Shopify Balance on top of commerce. Construction, field services, healthcare administration, freight, and professional services all share the same profile: fragmented industries, businesses with lumpy cash flow, and a software platform sitting on the data needed to lend against it intelligently. These are not payments companies that added software. They are software companies whose understanding of their niche lets them price risk better than a generalist bank ever could.

The second is embedded lending delivered through specialist providers — Pipe, Parafin, Wayflyer and their peers — that let a platform offer revenue-based financing or working capital without building a lending operation from scratch. The platform supplies the customer relationship and the data; the provider supplies the capital and much of the risk machinery. It is the BaaS pattern applied to credit rather than to accounts, and in a higher-rate environment the spreads have become genuinely interesting.

What ties both together is that the money is in credit and deposits, not in interchange. Interchange got the platforms in the door. It is rarely what keeps the lights on. The settlement layer underneath all of this is shifting too, as stablecoin settlement starts to compress the cost and time of moving money between the parties in these chains — a development the embedded-finance infrastructure providers are watching closely.

What this means for incumbent banks

The incumbent's instinct is to read all this as disintermediation: software platforms stepping between the bank and its customer, capturing the relationship and the data, leaving the bank as a commoditised balance sheet for hire. That reading is half right.

It is true that the customer relationship increasingly belongs to the software, not the bank. A contractor thinks of their banking as a feature of their job-management platform; they may not know, or care, which chartered institution holds the deposits. For a bank that built its franchise on owning the primary relationship, that is a real erosion.

But the partner-bank role is not nothing, and the post-Synapse environment has quietly made it more valuable, not less. Sponsoring embedded-finance programmes well — with the compliance rigour, the reconciliation discipline, and the regulatory standing the model demands — is turning into a genuine specialism. The banks that treat Banking-as-a-Service as a serious, well-governed line of business rather than a cheap source of deposits are positioned to benefit from every software platform that decides it wants to be a bank without doing the hard parts. The economics of the underlying core systems matter here too; banks still running embedded-finance programmes on legacy infrastructure face the same constraints driving the broader cloud-native core banking migration.

The slogan was that every company would become a fintech company. The reality of 2026 is more textured. Every company can now rent the pieces. Only some will understand what they have rented — and the gap between those two groups is where the next few years of winners and casualties will be decided.