The Curious Case of the Missing Items: a Refund Abuse Story
How a major US-based retailer got their skyrocketing item-not-received claims under control.
In retail, a relatively quiet Q1 is a good time for merchants to reflect on and assess their holiday results, and also plan for the next holiday season. Before you get started, a word of caution: while promotions, attractive offerings, and strong brand reputation attract new and loyal customers in November and December, they also open the door to a new type of fraud.
Refund and return abuse cost merchants over $25 billion each year. Before you blame career fraudsters, know that much of it is carried out by paying customers using their own credit cards and accounts, thus easily bypassing most fraud prevention protections. Recent years saw a steady increase in the adoption of return and refund abuse not only by cybercriminals, but by regular—and sometimes loyal—customers.
Some of these abusive customers might still generate revenue for eCommerce businesses, and some may even be extremely profitable customers. Merchants’ inability to differentiate between good, paying customers and potential abusers makes this problem uniquely difficult to stop.
Let’s take a look at how this scenario played out for one major US retailer.
*Merchant name withheld for anonymity
The story of Merchant X
Merchant X is a big retailer and online marketplace operator based in the US. In 2019, just before the pandemic, Merchant X detected a rise in refund claims. A few months later, as refund claims continued to climb with the inevitable increase in digital purchases, they began to suspect that they were burdened with a significant, and rising, refund abuse problem.
Merchant X has a large in-house fraud investigation team. More than once, their internal investigations were able to link multiple refunds to the same user email or address. Their suspicion was confirmed when the merchant’s name started popping up in several dark web refund scam publications. Acting on this lead turned out to be easier said than done.
The biggest challenge of curbing refund abuse is that, at the end of the day, most claims are valid. Experts today estimate that only about 7-10% of refunds are actual abuse. Which brings us to the next challenge: how do you define abuse? “Abuse” might mean different things to different merchants, and the boundaries created by this definition must be tailored to each merchant’s needs. In other ways, you have to first detect an abuse problem, track, and assess it before you can put a stop to it.
The first step for Merchant X was to establish what they consider refund abuse.
When analyzing their refund claims for “suspicious” activity, which they decided meant three or more claims submitted using one account, email, or home address, Merchant X discovered that around 12% of all refund claims met the criteria.
The question remained: what do they do with this information? On the one hand they could start declining suspicious claims, and risk losing loyal customers who might simply have several refund claims due to their high order volume. On the other hand, they could be more lenient with their refund policy, and continue to knowingly pay unjustified claims in the process. Merchant X was at a crossroads.
That’s where we stepped in. Riskified’s powerful linking technology, coupled with our unmatched merchant network, offers merchants richer insights. We can determine the true identity behind each online action, including refunds and returns, giving you the benefit of seeing the full picture. We also attach a clear abuse rate to each identity, so you’re not only able to refuse abusers instantly and anywhere along the shopping journey, but you can also provide better service to good customers.
Here’s what we found
By analyzing Merchant X’s data, we were able to link 35% of the fake identities that submitted claims (in many cases, while using several accounts) to real, non-profitable repeat abusers. Based on this, we were able to determine that, on average, these abusers accounted for more than half of Merchant X’s refund costs, but only contributed about 5% of the revenue. After segmenting these high cost/low value accounts, Merchant X was able to apply a tailored policy that actioned refunds with the necessary precision and accuracy. This approach enabled Merchant X to cut down abuse loss by more than 50% while still securing at least 95% of the revenue driven by the valuable customers they initially labeled “suspicious.”
Riskified’s technology helps uncover users’ true claims history, so you can tell good customers from bad actors and put an end to policy abuse. Learn more about Policy Protect here.