Roughly $125 trillion moves between businesses every year, and about 40% of US B2B payments still settle by paper check. That's the short answer. The longer answer is more interesting.
Consumer payments were digitized a decade ago — Venmo, Apple Pay, Stripe Checkout, tap-to-pay. The plumbing between companies never got the same treatment. An accounts-payable clerk in 2026 still keys invoices by hand, cuts checks, and reconciles bank statements line by line. That friction is exactly why B2B fintech pulled in over $20B of venture capital across 2024–2025. This is the last analog holdout in payments, and it's the biggest one.
B2B Payments Fintech in 2026: What the Market Actually Looks Like
B2B payments fintech in 2026 is a market where annual flows reach about $125 trillion globally and roughly $25–30 trillion in the US, yet nearly 40% of those US transactions still move on paper checks. The opportunity is not inventing a new way to pay — it's digitizing AP and AR workflows, embedding payments into software, and shifting volume off checks onto ACH, cards, and real-time rails. Capturing even a few basis points of that flow is a multi-hundred-billion-dollar prize.
To understand where value accrues, you have to start with the rails. Each one trades off cost, speed, data, and acceptance differently — and that's what determines which startups win which slice of the flow.
The B2B Payment Rails, Compared
| Rail | ~Share of US B2B | Cost per txn | Speed | Best for |
|---|---|---|---|---|
| Paper check | ~40% | $3–$6 all-in | 3–7 days | Universal acceptance, carries remittance |
| ACH | ~35% | $0.20–$1.50 | 1–3 days | Cheap recurring vendor payments |
| Wire transfer | ~10% | $15–$50 | Same day | Large, time-sensitive payments |
| Commercial / virtual card | ~12% | 1.5–3% interchange | 1–2 days | Rebates, controls, rich data |
| Real-time (FedNow / RTP) | <2% | $0.045–$0.25 | Seconds | Instant settlement, 24/7 |
| Cross-border / stablecoin | ~1% | 0.5–3% + FX | Mins–days | International supplier payouts |
The headline: the cheapest, fastest rails (ACH, FedNow) carry the least volume, while the slowest, most expensive rail (the check) carries the most. That inversion is the entire business case for B2B payments fintech in 2026.
Why B2B Payments Stayed Analog So Long
Consumer payments are one-sided — you tap, the merchant accepts whatever you use. B2B payments are two-sided and stateful. A payment is the last step in a long chain: purchase order, invoice, approval, matching, payment, reconciliation. Digitizing the payment without digitizing the workflow around it solves nothing.
Two-sided adoption
Both buyer and supplier must be on a compatible system — checks need neither
Remittance data
Suppliers need to know which 47 invoices a $312,000 payment covers
ERP and AP integration
Payments live inside NetSuite, SAP, QuickBooks — not a standalone app
Reconciliation
Every payment must auto-match to an invoice or it creates manual work
This is why the winners in B2B payments are software companies first and payment companies second. You can't bolt a payment rail onto a broken workflow. You own the workflow — the invoice inbox, the approval chain, the supplier directory — and the payment becomes the monetization layer underneath. It's the same vertical-software logic I wrote about in vertical AI: own the system of record, then earn the flow.
The Startups Digitizing B2B Payments
The B2B payments fintech landscape in 2026 splits into roughly four lanes. Here's who leads each, with real numbers.
AP/AR Automation (SMB & mid-market)
Pays/gets paid for 480K+ businesses; processes ~$300B annual TPV
SMB AP/AR; check-to-digital conversion engine
Global mass-payouts and supplier management
Spend Management + Cards
Corporate cards + AP; fastest-growing in the category
Cards, banking, and global payments for startups & enterprise
Embedded & Platform Payments
Invoicing, ACH, and embedded B2B checkout via API
B2B net-terms and embedded marketplace payments
Startup Banking
Banking + bill pay for 200K+ startups
Treasury, cards, and payments in one stack
Note Melio's exit: Xero paid ~$2.5B in 2025 specifically to own the SMB payments layer underneath its accounting software. That deal told the whole market that AP/AR is strategic infrastructure, not a feature. You can track these private valuations on our unicorns dashboard and the broader software comps on SaaS valuations.
How B2B Payment Startups Actually Make Money
The B2B payments fintech business model in 2026 is rarely pure SaaS. The subscription gets you in the door; the payment flow is where the margin lives. Four stacked revenue lines, ranked by how much they typically contribute:
Where the money comes from
- ✓ Card monetization: 1–3% interchange/spread on card volume
- ✓ Cross-border FX: 0.5–3% margin on international payouts
- ✓ Float: yield on held balances (meaningful at 4–5% rates)
- ✓ SaaS: $20–$100+ per seat/month software fees
Where the model breaks
- ✕ ACH-only flow: pennies of margin, no monetization
- ✕ No workflow lock-in: customers churn to the next tool
- ✕ Stablecoins compressing cross-border FX margins
- ✕ Falling rates erasing float income
The structural risk worth watching: the same forces I covered in agents, stablecoins, and the death of interchange are coming for B2B too. If stablecoins move cross-border B2B settlement to near-zero cost, the FX margin line — one of the most profitable in the stack — compresses hard. The defensible businesses will be the ones that own the workflow, not the ones that only own the rail.
What This Means for Founders and Investors
Three takeaways from where the B2B payments market sits in 2026:
1. The TAM math is real but the take rate is small. $125T of flow sounds infinite, but B2B take rates are measured in basis points, not the 2.9% you see in consumer card processing. The companies that win charge software fees on top of payment monetization — that's how Bill gets to a ~$1.3B run-rate.
2. Distribution beats product. The check survives because of inertia. Beating it requires getting both sides onto your platform, which is a distribution problem. That's why embedded payments — selling through accounting platforms, ERPs, and vertical SaaS — is winning over standalone apps.
3. The check finally dies this decade. Check volume is falling ~4–5% a year and FedNow adoption crossed 1,400 institutions in 2025. The shift from 40% checks to under 20% is the single largest migration of dollar volume in payments — and whoever owns the workflow during that migration captures the flow.
The biggest market in fintech isn't consumer payments or crypto.
It's the $125 trillion still moving on paper between businesses — and the migration off the check is just getting started.
Track fintech and software valuations on the SaaS Valuations Dashboard and Unicorns Tracker at Value Add VC. Originally published in the Trace Cohen newsletter.