Web3 Payments: What Businesses Need to Know

Strip the branding off most products sold as web3 payments and you find a stablecoin transfer with a dashboard bolted on top.

That is not an insult.

Sending USDT that settles in seconds for a couple of dollars beats a $43 wire that clears in three days, and the dashboard is usually the part you're paying for. But plenty of CFOs wrote the category off after one bad pilot in 2022. Which is a shame, because the useful half of web3 payments got quietly good while nobody was watching, and the useless half got easier to spot.

What "web3 payments" means once you take out the marketing

There is no committee that defines the term. In vendor materials it stretches to cover at least four different things, and they are not equally serious. First, plain value transfer on a public blockchain: a stablecoin moves address to address and settles on-chain, no correspondent bank in the middle. Second, a gateway or checkout layer that takes crypto from a customer and either keeps it or converts it to fiat. Third, programmable money, meaning payments with conditions written in code (escrow releases, streaming salaries, splits to a revenue-share table). Fourth, the token-flavored stuff: loyalty tokens, NFT receipts, wallet-as-identity checkout.

Category one carries almost all of the real B2B volume. Category three is genuinely interesting and used by a narrow set of companies with the engineering to support it. Two and four are where the marketing budget goes.

So when a vendor pitches you web3 payment solutions, the useful question is boring: which of those four am I buying, and would a plain stablecoin transfer with good bookkeeping do the same job for less? Often the answer is yes.

Never moved a stablecoin? Our breakdown of what a USDT payment actually is covers the mechanics first.

What actually works today

Stablecoin settlement on chains that don't punish you

This is the load-bearing use case. USDT and USDC move 24/7, including on a Sunday in December, and the fee is a function of network congestion rather than a percentage of the amount you're sending. Paying out $180,000 costs the same as paying out $180.

Chain choice matters more than vendor choice.

Tron (TRC-20) runs roughly $1 to $4 per transfer, confirmed in seconds, which is why so much stablecoin volume sits there. Ethereum (ERC-20) is roughly $5 to $15 and climbs hard when the network is busy, but custodians and auditors treat it as the default. Solana and Polygon land under a cent, Arbitrum around $0.05 to $0.30. The tradeoffs between those first two are in our comparison of ERC-20 and TRC-20 for USDT.

What nobody puts in the sales deck: your counterparty's exchange or wallet decides which networks it accepts, and a mismatch means the money is gone. Not delayed. Gone. Every finance team doing this at volume has a story about a $40,000 transfer sent on the wrong chain, and the good ones keep a network column in their address book.

Wallet-to-wallet payouts for people banks are slow to serve

Wallet-to-wallet payments come into their own where the traditional rail is not merely expensive but unreliable. Consider a design studio in Warsaw paying six freelancers in Lagos, Buenos Aires and Karachi. Their choice is not $2 against $43, but "arrives today" against "sometimes bounces back after nine days minus $28 in intermediary fees, at an FX rate 4% off mid-market." That all-in cost is the number people forget. An international wire commonly runs 3% to 8% once you include the FX markup, and the markup is where the money is. We compared the rails in our guide to moving money across borders with crypto, including where wires still win.

Contributor payouts are the other mature case, run for years by protocol teams and DAOs with multisig approvals standing in for a payroll provider, an approach we covered in how DAOs and protocols pay contributors.

Programmable payouts and on-chain settlement

Here the "web3" prefix earns something beyond a stablecoin transfer. Payment and logic live on the same ledger, so conditions get enforced without a trusted intermediary: escrow that releases when a milestone is signed off, revenue splits that fire to eleven addresses, salary that streams by the second. Every one of those needs contract code someone audited, and the failure mode of a bug is not a support ticket; it's a loss. Most companies under 200 people should not be writing payment logic in Solidity. Batch payouts with human approval, reconciled properly, get 90% of the benefit with none of the smart contract risk.

Visibility you didn't have before

Small but real: every payment is verifiable by both sides against a public ledger, immediately. No "let me check with our bank."

A counterparty asking where their money is gets a link.

What does not work yet

Consumer checkout is the big one. After a decade of gateways promising it, the share of shoppers who will pay for an ordinary e-commerce order from a self-custodied wallet stays tiny. Adding crypto checkout will not move your conversion rate. Where it does pay off: high-ticket B2B invoices, marketplaces with international sellers, iGaming, IT outsourcing, affiliate payouts.

Then there's reversibility, or rather its absence. Merchants hate chargebacks, but the chargeback right is why consumers trust cards with strangers.

On-chain settlement has no equivalent. Once confirmed, a transfer is final.

Escrow contracts get offered as the answer; escrow is useful, but it is not a dispute process with an adjudicator, and it does nothing about your AP clerk typing the wrong address. Whitelisting and a second approver are the real mitigations, and both are process, not technology.

Volatility, oddly, is the least interesting objection now. Settle in a fiat-referenced stablecoin, sweep to your bank on a schedule, and the exposure window is minutes. What you keep on-chain is a treasury question, not a payments one.

The web3 payments market: where the money actually is

Follow the volume rather than the press releases and the picture gets simple. Real money sits in three places. Business-to-business stablecoin settlement is the largest by a distance: invoices paid in USDT or USDC because the alternative wire costs 3% to 8% all in. Payouts come second, meaning contractor payments, affiliate commissions, creator withdrawals, iGaming settlements, anything with a long recipient list in countries your bank treats as a compliance project. Third, remittance corridors, where stablecoins compete against money transfer operators charging percentage fees on small amounts, and the economics turn on whether a local off-ramp exists at all.

The adjacent numbers hold up. Crypto payroll alone was a $1.48B market in 2024 and is projected to reach $6.38B by 2033, a 19.2% compound annual growth rate, and payroll is one narrow slice of payouts.

Now the noise. Consumer checkout has absorbed a decade of funding for a share of retail spend that rounds to nothing outside a few verticals.

Token loyalty schemes, NFT receipts and wallet-gated pricing are marketing experiments, not measurable payment volume. The web3 payments market in vendor decks and the same market in settlement data are close to inverses of each other, which is why so much web3 payments innovation went at problems that were not costing anyone money.

Practical read: a provider whose homepage leads with checkout is chasing the smaller half, and a provider whose documentation opens with batch payouts and address screening has looked at where the flows are. Buy from the second group.

Web3 payments vs cards vs wires

Read the first two rows together, because that is the whole argument. Card fees are a percentage, network fees are flat. A $50 payment on card rails is cheap and on Ethereum is absurd.

A $50,000 payment inverts it.

What a web3 payment gateway actually does

Gateways point one direction: money coming in. They generate a fresh address per customer or per invoice so incoming funds are attributable, watch the chain for confirmations, convert to fiat at receipt if you want, and hand your accountant a record with an order reference attached. That is the whole product. Useful, and narrow.

Payout tools point the other way. They connect to a funded wallet, take a list of recipients, screen them, route approvals through more than one person and push a batch out. Different data model, different risk, different failure modes. Nothing overlaps except the word "payment."

What a business gets from a web3 payment gateway: attribution, so you know which invoice a transfer settled; confirmation logic, so you know how many blocks before you mark something paid; an off-ramp in one jurisdiction or two; store plugins. What it does not get: batch sending, an approval chain, an address book with a network column, risk screening on the addresses you pay out to, or anything resembling payroll.

That gap exists because of who these products were built for. Most software marketed as a web3 payment gateway was designed for e-commerce checkout, borrowing its shape from Stripe and its pricing from card acquiring: percentage of volume, plugins for Shopify and WooCommerce. Point that at B2B payouts and the mismatch shows up inside a week. You pay a percentage on outbound transfers where the underlying cost is flat, run a one-address-per-order model against a list of forty contractors, and ask a checkout widget to enforce a dual-approval rule it was never built to hold.

So if the problem is paying people, the gateway shelf is the wrong shelf. Ask for a payout platform, and let a gateway handle whatever inbound crypto revenue you actually have.

Choosing a web3 payment processor

Five questions, ordered by how much they will hurt if you get them wrong.

Custody first. Custodial providers hold the keys, which buys you approval workflows, recovery and a support number, and costs you a counterparty. Non-custodial leaves the keys with you and the operational burden with your team. Neither answer is correct in the abstract. But if you cannot name where the assets sit, which legal entity is on the hook and which regulator supervises it, you have not finished the question.

Chains next, and your counterparties decide this one, not you. Pick a provider that supports six networks you never touch and not the one your Jakarta supplier receives on, and you have bought a web3 payment processor you cannot use.

Fee model third. Percentage-of-volume pricing came from card acquiring and does not fit a rail where cost is flat: 1% on a $200,000 payout is $2,000 to move a file that costs a few dollars in gas. Flat per-transaction pricing is the honest shape, and a few vendors quote it that way: VaultNow, for one, charges $0.50 per transaction plus gas, with up to 100 payouts per CSV batch. Some vendors bury the margin in the FX or off-ramp leg instead, so ask for an all-in number on a real example.

Settlement: what lands in your bank account, in which currency, on which day and who eats the spread. "T+1" from a provider with one banking partner in one country is not the same product as T+1 from three.

Compliance posture last, only because it is easiest to check. Screening in both directions, travel-rule handling, licenses you can look up in a public register. Vendors who get vague here are telling you something. We put the category through those filters in our roundup of stablecoin payment processors for B2B.

Compliance after the GENIUS Act and under MiCA

The regulatory story is the main reason web3 payments moved from fringe to defensible in the last two years. In the US, the GENIUS Act was enacted on 18 July 2025 as the first federal framework for payment stablecoins, setting reserve, disclosure and issuer requirements. The practical effect for a finance lead: settling in a compliant dollar stablecoin is now a documented policy choice rather than something you hide from your auditor. Our walkthrough of the GENIUS Act covers what changed downstream.

In the EU, MiCA has already reshaped which assets you can practically use. USDT was delisted from EU-regulated venues; USDC is compliant. If you have European entities or customers, that single fact will drive your stablecoin choice more than any fee comparison.

Reporting caught up too: Form 1099-DA for digital assets in the US, the 1099-NEC threshold at $2,000 under OBBBA, FASB ASU 2023-08 putting crypto holdings on fair-value accounting, DAC8 extending EU information exchange to crypto.

None of this makes compliance lighter than card acceptance, only knowable, which is a different thing. Budget for address screening in both directions, KYB files on counterparties you pay, travel-rule data where thresholds apply and bookkeeping that captures fiat value at the moment of transfer rather than reconstructing it in March.

How to start without betting the company on it

Run it as a narrow pilot with a real deadline. Six weeks is enough.

1. Pick one payment flow, ideally outbound. Contractor payouts or supplier invoices in one or two countries. You control the timing and the counterparty list, and no checkout conversion is at risk.

2. Choose the asset and network before the vendor. USDC if EU exposure or bank-facing optics matter; USDT if your counterparties sit in Southeast Asia, Latin America or the CIS. Then pick the chain they can actually receive on.

3. Write a one-page policy. Who approves, what the per-transaction and daily limits are, which addresses are whitelisted, how much stays on-chain, who holds the keys. One page. Nobody reads five.

4. Screen every counterparty address before the first send. Risk scoring on receiving addresses takes seconds. Skipping it is how a payout ends up in a bank compliance review a year later.

5. Send a small test transfer to each new address. $10, confirm receipt, then send the real amount. Nothing else here has saved more money.

6. Reconcile weekly from day one. Fiat value at time of transfer, fee, counterparty, invoice reference. If you also bill in crypto, set the paperwork up properly using our guide to invoicing clients in crypto rather than improvising in a spreadsheet.

7. Measure against your old rail honestly. Total cost including staff time, failure rate, days to settle, counterparty complaints. If it doesn't beat wires on your corridors, say so and stop.

Most companies that run this pilot properly land in the same place: stablecoins for cross-border flows and contractors, cards for consumer revenue, a bank for everything domestic and large. A sensible outcome, not a failure of ambition.

The honest summary of web3 payments in 2026 is that the boring part works well and the exciting part mostly doesn't yet. Flat-fee, near-instant settlement to counterparties your bank handles badly is a real win available today, and if you want payouts, invoicing, AML screening and analytics in one place, that is roughly the job VaultNow was built for. Start with the flow costing you most in fees and delay.

Frequently asked questions

What are web3 payments in simple terms?

Web3 payments are transfers that settle directly on a public blockchain between two wallets, without a bank or card network in the middle.

In business practice that almost always means stablecoins such as USDT or USDC moving on Tron, Ethereum, Solana or Polygon. Everything else sold under the label, from checkout widgets to programmable escrow, is built on top of that basic transfer.

What is a web3 payment gateway?

A web3 payment gateway is software that accepts crypto payments from your customers: it issues an address per invoice or order, monitors the chain for confirmation, optionally converts to fiat, and exports records for accounting. It handles money coming in, not money going out, so batch payouts, approval chains and counterparty screening need a payout platform instead.

What is the best web3 payment processor for a business?

There is no single best one, because the right web3 payment processor depends on whether you are collecting or paying.

For outbound B2B flows, judge candidates on flat per-transaction pricing, the networks your counterparties actually use, address screening in both directions, granular permissions and a named off-ramp bank. For consumer collection, checkout-focused providers do that job better than payout platforms.

How big is the web3 payments market?

Nobody publishes one clean number, because the category spans settlement, payouts, gateways and remittance. The measurable slices are informative: crypto payroll alone was worth $1.48B in 2024 and is projected at $6.38B by 2033, a 19.2% compound annual growth rate. Business-to-business stablecoin settlement is larger than payroll, while consumer crypto checkout stays a small share of retail spend.

Are web3 payments cheaper than card processing?

It depends entirely on ticket size.

Card fees are a percentage, typically 1.5% to 3.5% plus a fixed amount, while network fees are flat: under a cent on Solana or Polygon, $1 to $4 on Tron, roughly $5 to $15 on Ethereum. Below about $100 per payment, cards usually win. Above a few thousand dollars, on-chain settlement wins by a wide margin.

Can a crypto payment be reversed if we send it to the wrong address?

No. A confirmed on-chain transfer is final, and no chargeback mechanism or central authority exists to reverse it.

That is the biggest operational difference from cards and, to a lesser degree, from wires. The standard defenses are address whitelisting, a small test transfer to every new address and a second approver above a set threshold.

Is accepting crypto payments for business legal in the US and EU?

Yes, in both, under conditions that are now reasonably clear.

The GENIUS Act, enacted 18 July 2025, created the first US federal framework for payment stablecoins. MiCA governs the EU, where USDT has been delisted from regulated venues while USDC remains compliant. You still carry AML, KYB, sanctions screening and tax reporting duties, including Form 1099-DA and DAC8.

Which stablecoin and network should a business use?

USDC on Ethereum is the conservative default for anything EU-facing or bank-facing, since it is MiCA-compliant and widely accepted by institutions. USDT on Tron is the practical choice for contractor payouts in Asia, Latin America and the CIS, where counterparties ask for it by name and fees run $1 to $4. Let their receiving capability decide, not your preference.