Machine Learning Technology: Detect and Prevent Refund Fraud

Refund Fraud: The next hole to whack the mole

Stuart Mann

Mar 29, 2022

During the holiday period I was able to take my kids to a funfair, which I believe is called a carnival in the US – an outdoor travelling fair that has rides, amusements, and arcade games. One such game was whack-a-mole and my children spent quite a lot of time (and my money) insisting they play this game to hit a score that would result in them winning an oversized teddy bear.

Whilst this game was new to them, I feel this is a concept we’ve been dealing with in the eCommerce fraud space for years. I’ve always joked that whenever a new ground-breaking technology or new regulation focused on fraud prevention and detection comes along, do fraudsters look at it and think, “ok, it’s been a good run, but they’ve got me now – time to take an early retirement”?…….OR, do they just look for the next vulnerability? If Accertify’s experience is anything to go by, it’s normally the latter.

So what’s the next hole?

According to the National Retail Federation1, in 2021, US retailers sold over $1 trillion of merchandise online, an almost 100% increase in online sales, compared to their 2020 report2. Many retailers have spent years increasing their layered approach to payment fraud prevention and detection whilst ensuring they minimize false positives and customer insult. Fraudsters have had to a step back and determine the areas of weakness across the eCommerce life cycle. One specific area they found is around the refund process.

This is a massive issue for retailers

Pre-pandemic, places like New York were experiencing 90,000 lost or stolen packages per day3 and in the UK, this amounted to 5 million lost or stolen packages in 20204. There are many reasons why genuine customers have their packages lost or stolen, from porch pirates5 looking for packages left on the doorstep to couriers being overwhelmed with the sheer volume of packages and many getting lost during transit.

The NRF report also showed that 20.8% of all online purchases are returned and many retailers are now issuing the refunds, telling the customer to keep the items6, as they can’t handle the volume of packages coming back to the warehouse and the costs associated with handling returns.

This proved to be an opportunity for fraudsters. Last summer Accertify held a webinar on this very topic – refund fraud. A “customer” calls up a retailer, claiming their package has not been delivered when, in fact, it had.  We uncovered that often this “customer” was a professional refunder, advertising their services across many social media platforms, offering refund guarantees for high ticket items. The refunder will show an individual which companies they can target, the items to buy and the maximum amount they can spend.

They will then make a genuine purchase within those guidelines. There will often be nothing unusual about this order and it will bypass any risk controls. Once delivered (the professional refunder often advises that the customer doesn’t sign for the package – perfect in a contactless COVID environment), the customer will contact the professional refunder and give them all the details about the transaction – order number; shipping address; item; cost; their email, etc.

The professional refunder will then contact the retailer, often via their call center, and using a script they’ve created give the information to the agent and most often result in a refund being issued. They professional refunder gets a fee for their assistance – approx. 10% – 15% of the value of the good from the customer, who effectively gets a $1000 laptop for a $100 fee.

The scam is straightforward due to the easy entry point into the retailer – namely a call center agent, who is often incentivised to minimize the length of time of a call and to ensure the customer leaves happy.

So how do you stay protected?

Accertify has created a solution that helps retailers fight refund fraud. Using state-of-the-art machine learning technology, we work with our clients to ensure claims are made online. This enables us to capture critical data elements that can be fed into a model and compare underlying information from the original order to the refund claim. In addition, by leveraging our rich community data, we can see when a fraudster (or a customer) bounces from retailer to retailer and prevent further attacks.

So, whilst we can’t guarantee winning you a soft cuddly toy, just like my kids, you can enjoy the newest version of Whack-A-Mole with Accertify by your side.

Sources:

  1. NRF National Federation – Consumer Returns in the Retail Industry 2021 https://bit.ly/3LbjufV
  2. NRF National Federation – Consumer Returns in the Retail Industry 2021 https://bit.ly/3LbjufV
  3. New York Times https://nyti.ms/3It93CJ
  4. The Guardian, 30th July 2021, Over 5m people in UK had parcels lost or stolen last year, says Citizens Advice https://bit.ly/3iMpMqx
  5. Wikipedia https://en.wikipedia.org/wiki/Package_theft
  6. CNBC, 25 January 2022, A more than $761 billion dilemma: Retailers’ returns jump as online sales grow https://cnb.cx/3ulZ0uj