Gift card seller cuts payment disputes 45% and nearly ends manual order checks

Raise, a marketplace for buying and selling gift cards, cut its chargeback rate by 45% and nearly ended manual order checks. That is according to a customer case study from fraud-prevention company Sardine. A chargeback happens when a customer disputes a charge and the seller loses the money. Raise's rate fell from 0.53% to 0.29% over three years.

What changed

The bigger shift was in human workload. Manual reviews once covered 15% to 18% of Raise's orders. Now they cover 0.25%. At the old rate, 150 to 180 of every 1,000 orders went to a person. Now fewer than 3 do.

Those reviews were costly. Each took 15 minutes to 24 hours, and together they used more than 10,000 hours a year. Customers waited longer for approvals. The case study says Raise has processed more than $10 billion for about 6 million users across 1,000 brands.

Why gift cards attract fraud

A gift card is digital, easy to pass along and can be spent before a disputed payment is spotted. A scammer buys a card with stolen payment details, spends it, then the real cardholder disputes the charge. The seller is left with the loss. A Forter risk executive told PYMNTS in 2023 that a gift card is the online version of cash.

How it works

Sardine adjusted the fraud rules Raise already used. It then built custom models for gift card scams specific to Raise. A rule fires when one value crosses a set limit. A model weighs several signals together, such as a new device, a hidden IP address and an unusual order size, that no single rule would catch.

Across its platform, Sardine looks at device details, location, proxy use, typing rhythm and mouse movement. The case study does not say which of these Raise uses.

What is not known

This is a vendor's own case study, and it leaves gaps. It does not say how many reviewers were moved to other work, what the saved labor was worth, or how much chargeback loss was avoided. It also does not say whether fewer good orders are wrongly declined, or whether sales rose.

That last point matters. PYMNTS Intelligence found that 47% of merchants estimate up to 5% of legitimate orders are wrongly declined. It also found that 85% say their biggest fraud challenge is reducing friction for good customers while stopping bad ones.

What to do

Sellers who still check many orders by hand can ask their payment or fraud provider about screening that flags fewer orders without loosening protection. Useful questions include:

• How many orders go to manual review today, and how long does each take?
• What is our chargeback rate, and how has it changed?
• How many good customers are being declined by mistake?