How to set a delivery window for cross-chain payouts
A cross-chain payout window is the time you allow a transfer to reach its destination and be available to the recipient. The key condition is the route’s full completion time, not just how long the source transaction takes to confirm.
For a business payout, keep three points in view:
- Measure source finality, relay or route processing, and destination completion separately.
- Set your window from observed slow cases, then define when staff should investigate.
- Track both transaction hashes so you can distinguish a delay from a failed transfer.
Which parts of a transfer determine arrival time?
Arrival time depends on each stage in the route. First, the source chain must include and finalize your transaction. Then the bridge or route must detect it and act on it. Finally, the destination chain must process the transfer before the recipient can use the funds.
A relay bridge may wait for a required number of source confirmations before relayers carry the message onward. That wait can vary with the source chain’s block time, congestion, and the bridge’s security rules. A route using a liquidity pool may deliver sooner when funds are available, but its timing and fees depend on the pool and route conditions.
For example, suppose your team sends a scheduled stablecoin payout from one chain to another. The source transaction confirms in a minute, but the route requires several more confirmations before relaying; destination processing then takes another minute. Treat those as separate stages in your records. A source transaction marked successful does not, by itself, prove that the recipient has received spendable funds.
Use a consistent start and finish when measuring: start when your payout system submits the source transaction, and finish when the destination balance is available. Keep the source and destination transaction hashes, token, amount, route, submission time, and completion time together. This makes your timing data useful for both operations and reconciliation.
How do you choose a window your team can meet?
Base the window on your own route history. For each route, record normal completion time and a slower measure such as the 95th percentile: the time within which 95 out of 100 observed transfers completed. Separate routes and tokens if they use different chains or mechanisms, and gather enough examples to avoid setting policy from one unusually fast transfer.
Set the promised window above that slower measure, with a buffer for busy periods and your payout cut-off. As an illustrative policy, if 95% of transfers finish within 20 minutes, you might promise delivery within 45 minutes and investigate anything still pending at that point. Label the example as internal guidance; actual times depend on the route and current conditions.
Before sending, check that the destination network and token match the recipient’s instructions, the quote has not expired, and the payout amount fits your approval limits. If the transfer exceeds its window, check the source hash first, then the route status and destination hash. Escalate according to your internal process; avoid resubmitting until you know whether the first transfer can still complete, or you could pay twice.
For your next payout run, choose one route, capture both hashes and timestamps, and compare actual completion times with the window you planned. If your team uses a relay bridge, include confirmation and relayer delays in that measure; compare relay bridge fees with the cost of a route that may complete sooner. Adjust the window after you have enough observations to support it.
Questions teams ask next
When does a cross-chain payout count as complete?
Count it as complete when the destination transaction has succeeded and the intended recipient can use the expected token amount. A successful source transaction only confirms the first stage. Store the destination hash and completion time with the payout record so finance and operations teams use the same definition.
What should we do if the source transaction is confirmed but the payout is pending?
Use the source hash to check whether the transaction finalized, then check the route’s status and any destination transaction hash. A pending relay or destination step may still complete. Follow your team’s escalation process and confirm the original transfer’s state before sending again, to avoid a duplicate payout.
Should we promise the average transfer time?
No. An average hides the slower transfers that cause missed payout deadlines. Use a slower percentile from your own route history, such as the 95th percentile, and add an operational buffer. Revisit the window when route conditions change or your records show that the existing target is no longer realistic.
How often should we review payout timing?
Review it on a set cadence and after a route, token, or chain configuration changes. Compare completion times against your promised window and note how often transfers need investigation. If you do not yet have enough observations, label the window as provisional and monitor each payout until the sample grows.