How to pay out marketplace sellers: schedules and rails

Collecting money in a marketplace is the part everyone builds first — a checkout, a card form, a settled balance. Then you have to give most of it back. Paying sellers is the hard half, and it’s where marketplaces quietly lose margin and goodwill: the wrong schedule, a default wire that eats a seller’s thin margin, a payout that “sent” but never arrived. This is a practical guide to the three decisions that actually matter — when you pay, which rail you pay over, and how you pay sellers in other countries — and how they fit together.

Payout schedules

The first decision is cadence: how often a seller gets paid, and how fast the money lands once you decide to send it. It looks like an operational detail. It’s really a three-way trade between what sellers want, what fraud risk allows, and what each payout costs.

Sellers want their money now. A genuinely instant payout — money landing in seconds, any hour, any day — is the strongest retention lever a marketplace has, especially for people paid for their labour. But “instant” is expensive per transaction and it’s irrevocable: real-time rails don’t claw back, so if your risk checks run after the money moves, there is no after. Fast payouts and loose controls are a bad pairing.

Scheduled payouts sit at the other end. Batching a seller’s earnings into a weekly or daily run is cheaper per payment and gives you a window to catch chargebacks, refunds and fraud before the money is gone. The cost is felt by the seller, who waits.

Most marketplaces land in the middle: a reliable default schedule — weekly is common — plus a faster, sometimes fee-bearing option for sellers who want it sooner. The honest read: there’s no single right cadence, but there is a wrong move, which is selling “instant” and delivering same-day-batch that sleeps on weekends. Whatever you choose, tell the seller which rail their money is on and exactly when it will arrive.

Payout rails and reachability

Once you’ve decided when to pay, you decide how — and “how” is a rail, not a checkbox. The rail sets the cost, the speed, the cut-off times, and whether the money can ever come back. Sending £90 to a seller has several possible routes, and they are not interchangeable.

In the US alone there are four common rails with genuinely different characters. Same-day ACH is cheap and reaches effectively every bank account, but it’s batch processing on business days — miss the Friday cut-off and “instant” lands Monday. Push-to-card (Visa Direct, Mastercard Send) lands on a seller’s debit card in minutes, around the clock, but it’s priced like card traffic. RTP and FedNow settle in seconds, 24/7/365, with the money final on arrival — but each only reaches banks that have joined its network, and plenty of banks are on neither. In the UK, Faster Payments does much of this work in one scheme. Every country has its own map.

Which brings up the concept that decides whether a payout succeeds at all: reachability. A rail only works if it can actually reach that specific recipient. A real-time rail is useless if the seller’s bank doesn’t participate; push-to-card needs a valid debit card. So the rule for any payout system is: pick the best rail that can reach this seller, and keep an explicit fallback for when it can’t — usually a slower but universal rail like ACH or a local scheme. The failure mode to avoid is the silent one, where an “instant” payout quietly downgrades and the seller finds out by not getting paid. Ask any provider the same two questions: what’s the fallback, and who gets told?

Paying international sellers

Now scale that across borders, because most marketplaces do. Your sellers are in different countries holding different currencies, and this is where a single default rail does the most damage. A default SWIFT wire to a seller in Brazil or the Philippines might cost more and take days — and on a small payout, the fee can be a real slice of the seller’s margin. Do that a few thousand times a month and you’ve built an expensive, slow payout stack by accident.

The better model is corridor by corridor. For each seller, ask what the genuinely cheapest compliant way to move money into their country and currency is — and route to that, not to one pipe you happen to have wired up. A cross-border partner payout might go over a local scheme in one country, a faster-payment rail in another, push-to-card where the debit card is the quickest path, and a stablecoin corridor where a wire would cost more than the seller earns. Multi-currency matters too: collecting and holding in the currencies you actually transact in avoids paying to convert twice.

The principle is simple to state and easy to get wrong: there is no single best rail for international payouts — there’s a cheapest compliant rail per corridor, and the job is to pick it every time.

Where Fynex fits

This is the exact shape of money Fynex is built for. Fynex is an AI-native finance-operations layer that sits on top of your accounts and rails — not a bank, not a card network. It owns no rail and earns no spread on your flow, which is the whole point: when it chooses how to pay a seller, it’s optimising your cost, not steering volume onto a network that pays Fynex. The routing is neutral.

The flow runs end to end. You define a split rule as a set of lines — each a percentage or a fixed amount, to a named payee — and Fynex divides each incoming payment along those lines. Then it pays each recipient’s slice over the genuinely cheapest compliant rail for their country and currency, with a fallback when the first choice can’t reach the account. And every payout is reconciled back to the sale that created it, into Xero or QuickBooks, so a settlement is a matched record rather than a mystery deposit. Auto-invoicing, agentic collections and cash forecasting sit on the same layer.

Accounts hold money. Rails move it. Fynex is the layer that thinks — deciding, per seller and per payout, when to pay, which rail to use, and how to tie it all back to the books.

The takeaway

Paying sellers well isn’t one feature — it’s three decisions made on purpose. Choose a schedule that fits your risk and your sellers’ expectations, and never sell “instant” you can’t deliver. Choose a rail per recipient by reachability, with a fallback that’s explicit, not silent. And for international sellers, route each payout to the cheapest compliant corridor instead of defaulting everything to a wire. Do all three and reconcile the result, and payouts stop being the hard half — they become the part sellers stop asking you about.