In the increasingly fast-paced world of online retail, refund policies are becoming somewhat of a double-edged sword in some rather concerning cases—essential for building customer trust but vulnerable to exploitation by fraudsters. A recent study by Riskified, a key player in digital commerce risk management, has shed light on this growing issue.
Examining more than a million refund requests from three large retailers in 2024, the research uncovered that roughly 1% to 2% of overall order values are reclaimed through refunds.
Alarmingly, about one-quarter of these refunded amounts stem from abusive practices, where bad actors manipulate systems for unwarranted gains.
The holiday rush exacerbates the problem, with November and December accounting for nearly a third of all claims.
Many of these spill over into January, creating operational bottlenecks that strain resources.
High-ticket items are particularly susceptible: orders exceeding $2,000 see claim rates 2.5 times higher than those under $100, and transactions over $1,000 are 33% more prone to fraud.
Specific claim types also raise red flags; for instance, “Item Not Received” assertions are 25% more likely to be fraudulent than “Missing Items” reports.
Additionally, requests filed within the first week post-purchase are over 20% more suspicious, highlighting how scammers exploit early windows before thorough verifications can occur.
This surge in abuse not only erodes profits but also hampers the shopping experience for genuine buyers.
Delays in processing refunds—often taking days or weeks—frustrate loyal customers, with 71% indicating they’re less likely to shop again if resolutions drag beyond five days. Such slowdowns rank as the second most common trigger for customer support queries, underscoring the need for smarter approaches.
With Dynamic Returns, Riskified’s newly launched AI-driven tool integrated into its Policy Protect suite.
Designed to curb losses while boosting shopper contentment, this feature shifts away from rigid, one-size-fits-all rules.
Instead, it delivers instant, customized decisions on refunds and exchanges by analyzing real-time data on customer history, behavior, and network signals.
Here’s how it operates: When a return is initiated, the system swiftly evaluates the claimant’s profile. For verified, low-risk users, it can approve immediate refunds, dispatch replacements without awaiting the original item’s return, or even waive the need to send back merchandise.
This precision helps flag dubious activities early, preventing revenue leaks without imposing blanket restrictions like fees or narrowed return periods.
In a real-world example, one merchant using the tool instantly greenlit over half of returns for repeat customers, resulting in a 20% uplift in satisfaction ratings and a 97% compliance rate where approved items were indeed returned.
Merchants benefit from fortified bottom lines and streamlined operations, avoiding the pitfalls of overly cautious policies that alienate buyers.
For consumers, quicker resolutions foster loyalty, turning potential pain points into positive interactions.
As Jeff Otto, Riskified’s CMO, noted, fraudulent refunds don’t just tend to dent retailer earnings—they erode trust among honest shoppers.
By leveraging AI to differentiate risks in moments, Dynamic Returns empowers businesses to reward reliability while thwarting deceit, striking a balance between security and service.
Riskified continues to support / enable fraud detection through data-driven innovations. Their report, “Refunduary is Here,” provides strategies for retailers navigating this challenge, available on their site.
In an era where e-commerce thrives on seamless experiences, tools like this could potentially improve how returns are handled, ensuring profitability and customer satisfaction go hand in hand.