Embedded Finance Is Eating B2B Software: Here Is Where the Money Is

The embedded finance thesis — that financial products work better when integrated into the workflows where the underlying activity happens — has been proven. The question now is execution: which platforms capture the most value.

Shopify Balance, Stripe Treasury, and Square Banking have demonstrated the model in the SMB space. Now the same pattern is playing out in B2B software: accounting platforms offering working capital, freight marketplaces providing invoice financing.

Why B2B is the more interesting bet

Consumer embedded finance is increasingly competitive on margin. B2B is different: transaction sizes are larger, relationships are stickier, and the financial needs are more complex. A contractor management platform offering a charge card and invoice factoring has a fundamentally different revenue model to one selling software seats alone.

The structural advantage is informational. A vertical SaaS platform sees its customers' invoices, payment behaviour and seasonality in real time — underwriting data a bank would pay dearly for and still receive late. That is why embedded lending through software platforms consistently reports loss rates that stand-alone SMB lenders struggle to match: the platform knows the borrower's cash position before the borrower's bank does.

The revenue mathematics

Three streams stack on top of subscription revenue. Payments interchange and processing margin is the entry point — modest per transaction, meaningful at volume. Float and deposit economics turn idle balances in embedded accounts into interest income, shared with the sponsor bank. Lending margin is the deepest pool: working capital, invoice factoring and charge cards priced on proprietary data. Platforms that execute all three routinely double or triple per-customer revenue versus their software-only baseline — which is why "every SaaS company is a fintech" went from conference slogan to board agenda item.

The infrastructure layer

None of this requires a banking licence, which is precisely the point. Banking-as-a-Service providers — Unit, Synctera, Treasury Prime, Stripe Treasury for accounts and money movement, Marqeta for card issuing — supply the regulated rails, with a licensed sponsor bank underneath. The platform owns the customer experience and the data; the bank owns the regulatory obligations. That division of labour is the entire economic architecture of embedded finance, and it is also where its fragility lives.

The reckoning the sector already had

The BaaS model took real damage in 2024–25. The collapse of middleware provider Synapse stranded end-customer funds and exposed how poorly some arrangements tracked who actually held whose money, and a wave of enforcement actions against sponsor banks made one thing unambiguous: regulators hold the bank accountable for everything its fintech partners do. The survivors responded with what the industry now calls the flight to quality — direct bank relationships instead of middleware stacks, cleaner ledger reconciliation, and compliance obligations written into commercial contracts. For platforms choosing infrastructure in 2026, the due-diligence question has inverted: it is less "which provider has the best API" and more "which sponsor bank will still want this business in three years".

Where the opportunity sits now

The pattern that wins is consistent: start with payments (lowest risk, fastest integration), use the transaction data to underwrite lending (the margin), add accounts and cards where workflow lock-in justifies them. Vertical platforms with high payment volumes and fragmented banking relationships — construction, logistics, healthcare, field services — remain the richest territory, because that is where the incumbent banking experience is worst.

The incumbent response

Banks are not standing still, but their options are awkward. Some are becoming the sponsor banks powering the trend — a sound fee business that nonetheless cedes the customer relationship to the platform. Others are buying or building their own embedded-finance arms, with mixed results: distribution through someone else's software is a culturally alien motion for institutions built on branches and relationship managers. The sharpest strategic question for a bank in 2026 is not whether to participate but where in the stack to sit — owning the regulated balance-sheet layer is durable and dull; competing with vertical SaaS platforms for the workflow layer is a fight most banks are poorly equipped to win.

For the platforms, meanwhile, the constraint has shifted from technology to risk appetite: the rails are commodity, the sponsor banks are choosier, and the winners are the ones treating compliance as a product feature rather than a procurement line.

For a fuller treatment of the revenue streams, the infrastructure layer, and where the margin actually sits, see our complete guide to B2B embedded finance.