Customer Operations
Resetting the Counter After a Card Terminal Goes Offline Mid-Transaction
A card terminal freezes while a customer is standing at the register with a completed sale. The queue is watching, the transaction is half-finished, and the terminal may have captured a payment the store cannot see. A calm reset routine protects the customer from a duplicate charge and gives the counter a defensible record. The routine should begin with the transaction on screen, not with a reboot.
Hold the transaction before touching the terminal
Ask the customer to keep the card and the receipt in hand while the clerk checks the terminal status. Note the transaction number, the amount, and the last four digits shown on the screen. Do not void or retry the payment until the terminal's state is clear. A second attempt can create a duplicate authorization even when the first attempt looked unsuccessful.
Check whether the terminal shows an approval, a decline, or a timeout. A timeout is not a decline, and it may still settle later. Write the terminal message exactly as displayed, including any error code. The exact wording helps the payment provider identify the transaction when the store calls.
Keep the printed receipt in the terminal until the case is closed. A receipt that is handed to the customer and then lost makes the reconciliation harder. If the terminal cannot print, write the transaction number on the store's outage form. The record should survive a terminal reboot.
Tell the customer what is known and what is not
Explain that the payment status is being checked and that the clerk will not ask for a second payment until the first is resolved. Avoid blaming the network or promising that nothing was charged. Give the customer a realistic wait time and an alternative if they cannot stay. The conversation should reduce uncertainty rather than fill it with guesses.
If the customer needs to leave, take a callback number and the transaction details, and give them the store's direct line. Tell them what the store will do next and when. Do not keep the customer's card or identification as collateral for an unresolved payment. A respectful hold is better than a forced wait.
Use the approved fallback payment path
Follow the store's written fallback procedure, which may include a second terminal, a different payment provider, or an approved cash option. Confirm the fallback is allowed for the sale type and the customer's identification check. Do not improvise a workaround that bypasses the store's payment controls. The fallback should be documented before the outage, not invented during it.
If a second terminal is used, label the new transaction as a replacement in the note and keep the original transaction number beside it. Tell the customer which payment method is being used and what will happen to the first attempt. If cash is accepted, count it at the counter and issue the standard receipt. The customer should leave with one clear record.
Tell the next customer in line that the counter is resolving a payment issue before they reach the register. A short warning prevents a second customer from presenting a card at the wrong moment. If the fallback will take several minutes, open a second lane or move the queue. The outage should not become a lobby-wide surprise.
Check for a duplicate authorization before closing
Look for a pending authorization on the terminal and in the payment provider's portal before the customer leaves. A pending amount may disappear without settling, so note the time of the check. If the provider shows a captured payment and the fallback also succeeded, begin the refund process immediately. The customer should not have to discover the duplicate on a statement.
Record the original transaction, the fallback transaction, and the refund reference in one case note. Attach the printed receipts or screenshots. If the refund cannot be completed at the counter, give the customer a case number and a expected resolution date. A duplicate charge is a trust problem, and the record should show that the store acted on it.
Reconcile the terminal at the end of the shift
Compare the terminal batch report with the sales record and the payment provider's settlement. Investigate any transaction that appears in one place but not the others. Check the fallback transactions separately so they are not counted twice. The reconciliation should name the clerk who handled each exception.
If the terminal lost connectivity, confirm that the stored transactions uploaded after the connection returned. A terminal can appear offline while holding transactions that will settle later. Record the reconnect time and the batch number. The closing lead should sign the reconciliation after the exceptions are explained.
Keep the provider portal open on a store device until the batch settles. A portal that is closed too early can hide a pending transaction. If the store cannot access the portal, record the provider reference and call the next morning. The reconciliation is not finished until the settlement is visible.
Write the outage note and assign the follow-up
Write a short note that covers the terminal, the time, the affected transactions, the customer contacts, and the open refunds. Name one owner for the provider call and one for the customer follow-up. Set a date to confirm the refund posted. A note without an owner becomes a story that nobody finishes.
Review the outage for a pattern at the end of the week. Repeated failures at the same terminal may point to a hardware or network issue, while repeated fallback errors may point to a training gap. Test the fallback path during a quiet period. The next outage should be shorter because the routine was practiced.
Test the fallback path during a quiet block once a month. A procedure that has never been practiced tends to fail under pressure. Record the test date and the result. The next outage should be a drill the team has already run.