How to avoid double withdrawal fees

Double commission is when you pay more than once for one withdrawal: for example, a fixed exchange fee and a high network fee; or take it to the wrong network, then bridge it to the right one and pay again; or first convert the asset with a spread, and then another network. The good news: in most cases, this can be avoided by planning a route "where the money is now → where it should be" and removing unnecessary steps.

1) Where does the "double" commission come from?

Network fee + fixed provider fee. The exchange/service takes its "withdrawal" fee on top of the onchain expenses.

The recipient has an invalid network. The output is ERC-20, and the target accepts the TRC-20 → you have to bridge/convert: another commission.

FX/conversion. First USDT→USDC (spread), then network transfer (gas) - pay twice.

Bridges and cross chain. Output to a "foreign" network, then the bridge → the network fee + bridge fee.

Resending due to error. Forgot Memo/Tag, sent the wrong way - you have to make a second payment.

UTXO fragmentation (BTC/LN). Many small inputs add to the output cost in the onchain.

2) Algorithm: choose a route without unnecessary payments (1 minute)

1. Where is the asset now? (exchange X, wallet Y; network and token exact: USDC (Arbitrum), USDT (TRON), etc.)

2. Where should he end up? (service/wallet of the recipient, exact network/token, whether Memo/Tag is needed).

3. Is there a direct output to the desired network/token? If yes, use it (one provider fee).

4. If there is no direct:

Convert within the source site to an asset/network that the recipient accepts directly.
Avoid external bridges/add. transactions.
5. Check the totals: conversion spread + output fee + network fee from the provider.
6. Make a test transfer of $5- $20, then send the principal.

3) Practical templates to avoid paying twice

The exchange → a casino/wallet on the same network.

Keep the stablecoin on the network that the recipient exactly accepts (for example, USDC on Arbitrum). Take immediately there - without bridges.

Crypt → fiat/bank.

Instead of switching to an "exotic" network + a bridge to a fiat provider - convert it on the stock exchange into an asset/network that offramp accepts directly.

BTC in small amounts.

If you saved up with small UTXOs, before a large withdrawal, do consolidation during quiet hours (one onchein-th instead of many expensive ones).

Lightning for small BTC payments.

For frequent small amounts, use LN (if the receiver supports) - do not pay twice: on-chain fee + "something else."

One ticket office is different networks.

If the service's withdrawal commission is very different across networks, choose a cheap network (L2/TRON/Solana/TON) instead of an L1 ERC-20.

4) "Black list" of situations where a double payment is almost guaranteed

Output USDT (ERC-20) and the receiver only accepts TRC-20 bridge/convert →.

Send XRP/XLM/BNB/EOS without Memo/Tag → manual parsing/resubmission.

Do DEX swap + bridge + withdrawal instead of one direct operation.

Unlimited'approve 'and then revoks every time - extra gas costs.

Trying to "save 20 cents" on a dubious bridge - risk paying twice (or losing funds).

5) How to compare routes (quick example)

Task: you have USDT (ERC-20) on the stock exchange, and the casino accepts USDC (Arbitrum).

Route A (expensive): withdraw USDT (ERC-20) → bridge in Arbitrum → swap for USDC → deposit. (terminal gathering + gas ERC-20 + bridge fee + gas in L2 + DEX commission/spread).

Route B (cheap): on the swap exchange USDT (ERC-20) → USDC (Arbitrum) (internal spread) → direct withdrawal of USDC (Arbitrum) to the casino (one withdrawal fee).

In 9 out of 10 cases, route B is cheaper and faster.

6) Subtle, forgettable moments

Withdrawal commission ≠ network fee. The provider can take a fixed fee - compare networks at a specific site.

Lows per withdrawal. Too small amount can "eat up" the collection. Sometimes it is more profitable to save up and withdraw once.

Gas at the receiver. In EVM networks, the recipient needs a native gas token for further action - add a couple of dollars "for gas" so as not to make a second transfer.

UTXO and "dust." In BTC, "dirty" small entrances make any transaction more expensive - consolidation saves in advance.

LN invoices with due date. Late invoice → repeat payment → another fee. Generate a new one.

7) Life hacks of savings

Match the net beforehand. Keep your balance on the network where you most often display.

Choose L2/cheap L1. Arbitrum/Optimism/Base/Polygon, Tron, Solana, TON - usually cheaper than ERC-20 on L1.

Consider all-in-one. Add conversion spreads and fixed fees to the "network fee" - be surprised at the difference in routes.

Test tx and address whitelist. Protects against network/memo error and second send.

Official bridges only when necessary. Otherwise - swap within the exchange and direct withdrawal.

8) Player checklist (30-60 seconds)

Exactly the same asset and network at the output and at the receiver.
Checked the output collection over the networks at my site and chose the minimum for the desired network.
If there is no direct network, I make an internal swap on the site and withdraw directly without bridges.
For XRP/XLM/BNB/EOS, specified Memo/Tag.
Made a test transfer of $5- $20.
The receiver has native gas in the network for the following activities.

9) Operator's checklist (cash desk/payments)

The showcase shows supported networks/tokens and minimums for each.
Hints/validations for Memo/Tag, big show the network.
There are cheap output alternatives (L2/TRON/Solana/TON) and they are visible by default.
For large amounts - tranche breakdown and limit control.
Internal FX snapshot and comparison of routes "svop→vyvod" versus "vyvod→most→svop."
Idempotency and protection against double enrollment/de-enrollment in webhooks.

10) Mini-FAQ

Why is the "network" zero, but they wrote off the fix from me? This is the output provider's collection. Compare networks: in one fix it can be many times less.

Is the bridge always more expensive? Almost always, if it is possible to make a swap and direct withdrawal on the original site.

TRON/TON/Polygon/Arbitrum really cheaper? In most cases, yes; the main thing is that the recipient accepts these networks directly.

Is it worth dividing the conclusion into parts? For very large amounts - yes (limits/risk); for small ones - on the contrary, one translation is more profitable.

Forgot Memo/Tag - money gone? Usually not, but you have to open a ticket and wait - and this is the time and often the second payment "in return."

Avoiding double commission means removing unnecessary steps from the route. Match the network with the receiving point, swap before output at the source site, use cheap networks, carefully specify Memo/Tag and start with a test transaction. This simple discipline reduces costs at times and eliminates repeated shipments and bridges, which make the conclusion "more expensive than it should be."