The economics haven't changed (but the options have)
If you're processing under $1M annually, you're paying 2.5-3.5% all-in on card transactions. The math is brutal: on $500K revenue, that's $12,500-$17,500 straight to processors. Small businesses have limited negotiating power, so you're typically stuck with flat-rate or tiered pricing from aggregators like Square (2.6% + $0.10) or Stripe (2.9% + $0.30).
The underlying structure hasn't changed: interchange fees (1.5-3.0%) go to issuing banks, assessment fees (~0.14%) go to networks, and processor markup (0.1-1.0%+) is where your leverage lives. What has changed is the growing number of rails available to route around these fees.
Instant payments are now real infrastructure
Instant payments have moved past hype: adoption is measurable and the cost advantages are real money.
RTP Network (The Clearing House, launched 2017): Processed $163 billion across 100 million transactions in Q1 2025 alone. Reaches 70% of U.S. demand deposit accounts through 850+ financial institutions. Transaction limit recently increased to $10M.
FedNow (Federal Reserve, launched July 2023): Hit 1.4 million transactions in Q1 2025, growing 405% quarter-over-quarter. Now reaches ~40% of U.S. demand deposit accounts through 1,400+ institutions. Transaction limit raised to $1M. Cost: approximately $0.045 per transfer.
For context: ACH processed 8.5 billion payments in Q1 2025. Instant payments are still tiny by volume but growing fast in specific use cases.
Why this matters for you: If you're doing B2B payments, paying contractors, or running payroll, instant payment rails cost pennies versus percentage points. A $5,000 vendor payment costs $2.25-$2.50 via card versus $0.045 via FedNow. At scale, this compounds.
73% of businesses surveyed now use instant payments (RTP or FedNow). The limiting factor is whether your bank and your counterparty's bank are both on the rails. Two-thirds of banks still haven't signed up, but the ones that have cover the majority of demand deposit accounts.
Practical implementation: Check if your business bank offers RTP/FedNow sending. Common use cases seeing adoption: instant payroll (employees get paid in seconds), vendor payments outside ACH windows, marketplace disbursements, insurance payouts. Not ideal for high-frequency small transactions due to operational overhead.
BNPL is a conversion lever (with real costs)
Buy Now Pay Later grew to $560.1 billion globally in 2025 (13.7% YoY). In the U.S., BNPL accounts for ~1.4% of retail sales, with transaction values expected to grow 12.5% in 2026.
The merchant economics:
PayPal data shows BNPL increases average order value by 62% for small businesses (91% for enterprises)
Conversion rates increase 20-30%
Cart abandonment drops up to 35%
44% of shoppers say they would've abandoned without installment options
The cost: BNPL providers charge 2-8% per transaction, typically higher than card processing fees. You're effectively subsidizing a 0% loan for your most price-sensitive customers.
Who should use it: E-commerce businesses selling $50-$500 ticket items where conversion lift outweighs the fee delta. Fashion, electronics, furniture see strong results. Gen Z (46%) and Millennials (47%) actively use BNPL. If your customer base skews younger and your AOV is in the sweet spot, the math works.
Who shouldn't: Businesses with tight margins, low AOV, or where the incremental conversion doesn't justify 4-6% fees over standard card processing. Also avoid if you're already converting well: you'd just be paying more for the same sales.
Cash use is collapsing (but not dead)
Cash dropped from 26% of transactions in 2019 to 16% in 2023. Projected to fall below 10% of POS transactions by 2026.
Digital wallets (Apple Pay, Google Pay, Samsung Pay) hit $300 billion in U.S. usage and saw 33% YoY growth in in-store adoption during 2024. Contactless tap-to-pay now represents 79% of in-store card transactions.
What this means: If you're B2C with physical presence, contactless acceptance is table stakes. If you're online-only, wallet integration (especially Apple Pay for iOS users) reduces friction in a meaningful way.
The fraud reality
Fraud is material, and it's on the rise. 55% of small businesses ($5M-$50M revenue) reported fraud in the past year. Checks (still used by 91% of businesses) are 16x more likely to be lost, stolen, or altered than electronic transfers.
Real-time payments like RTP/FedNow are irrevocable with no chargeback window, which cuts both ways: lower fraud rates reported so far (banks screen upfront), but no recovery mechanism after the fact.
Credit card chargebacks remain your enemy. Standard advice applies: use Address Verification System (AVS), require CVV, monitor for velocity patterns.
What you should do
If you're doing $0-$250K ARR:
Stick with flat-rate processors (Square, Stripe, PayPal). Simplicity matters more than optimizing 50 basis points.
Accept digital wallets at minimum. Apple Pay alone is material volume.
Skip instant payments unless you have specific vendor/payroll pain.
Consider BNPL only if you're e-commerce with $100+ AOV and see high cart abandonment.
If you're doing $250K-$1M ARR:
Move to interchange-plus pricing if you haven't made the switch. Providers like Helcim offer 0.2% + interchange with no monthly fee. The savings are real: potentially $5,000 to $7,500 a year.
Implement instant payments for high-value B2B transactions. Even 10 large vendor payments monthly at $0.045 vs. 2.5% saves over $1,200 a month.
Add BNPL if your customer data shows demand (check cart abandonment rates, AOV, customer age demographics).
Use ACH for recurring or subscription payments where speed isn't a priority. Costs ~$0.26-$0.50 per transaction.
If you're crossing $1M ARR:
Negotiate every fee. Your volume gives you leverage.
Build a hybrid payment strategy: ACH for recurring, instant payments for urgent B2B, cards where necessary, BNPL for conversion-sensitive SKUs.
Track your effective rate every month. Aggregate fees, divide by volume. If you're above 2.5% all-in with your volume, you're leaving money on the table.
The meta trend
Payment infrastructure is fragmenting from a card-network duopoly into multiple specialized rails. The technical indie hacker move is to route intelligently: use the cheapest rail that meets your settlement timing and UX requirements for each transaction type.
Most small businesses overpay because they default to cards for every transaction. Cards are expensive convenience. As instant payment adoption grows and tooling improves, the delta between "easiest option" and "optimal option" will become a real cost.
Start building optionality into your stack. The savings scale linearly with revenue.
---
Related reading: Want to prepare for payments PM interviews at Stripe, PayPal, or fintech startups? See Payments product manager interview questions and answers: the complete guide.