Why non-custodial settlement is the real unlock for stablecoin billing

When we started building OpenSettle, the easy version was obvious: take custody. Aggregate customer payments, batch them up, sweep them out to merchants on a schedule. That's how the processors do it. It's a known, regulated, surmountable problem.

The easy version is also a worse product. Here's why we went the harder route.

Custody is a compound cost

The bill for taking custody isn't just the legal one. It's an ongoing drag on every dimension of the product: you carry reserves, you pay for insurance, you hold funds long enough to be worth attacking, you operate a treasury function, you owe your merchants payout-latency transparency, you build dunning flows for underfunded withdrawals, you reconcile across bank days that don't match on-chain days. Every one of these creates customer-facing friction.

Non-custodial is architecture, not policy

Here's the version we picked. A deterministic smart contract — the Router — receives the customer payment and atomically forwards 98.5% to the merchant's wallet and 1.5% to ours, in the same transaction. We never have the funds. Not for a block, not for a second. The merchant's accounting is final at the moment of settlement. Our revenue recognition is trivial.

This isn't a policy we follow — it's a property of the system. We could not take custody if we tried.

What this unlocks

• No reserves, no payout delays, no minimums.
• Smaller regulatory footprint (we're a software platform, not a money transmitter).
• Merchants can switch wallets instantly without our involvement.
• One less thing for their counsel to ask about.

Non-custody isn't a marketing phrase. It's the foundation every other decision gets to sit on top of.