It’s no surprise the rise of eCommerce operations over the past several months has coincided with a rise in fraud. Before COVID-19, online fraud transactions were forecast to reach $25.6 billion in 2020 with retail fraud accounting for $16.6 billion*.
While the numbers are still being crunched for fraud during COVID, retailers and consumers have seen an uptick in fraudulent activity and scams. On top of pivoting business strategies, retailers will need to prioritize fraud prevention in order to keep their loyal customer base. This starts with identifying susceptible areas and how to tackle fraud while protecting profitability and customer satisfaction.
Top Fraud Areas: Card Attacks, Account Attacks, and Delivery Attacks
In the beginning of the pandemic, many consumers experienced the surprise of receiving COVID preparedness emails from businesses they didn’t realize had their email addresses. These forgotten or “dormant” accounts are very common and particularly vulnerable targets as they may contain sensitive information such as credit card details. Fraudsters can easily swipe card numbers and takeover accounts thanks to the new card-on-file model which has made it increasingly easy for consumers to shop.
Another area of concern has been the steady rise in new account creations. These raise concerns as merchants try to create a more seamless customer experience and often prioritize that customer experience over thoroughly scrutinizing accounts to weed out the “good” from those set up with bad intentions.
The final fraud area is delivery attacks. These have become popular as buy online, pick up in-store (BOPIS) at curbside has become a standard practice to accommodate social distancing. The point of vulnerability occurs at the last leg of the BOPIS transaction where checking information against who purchased is not always occurring or correctly occurring.
The best method for fraud management is a combined approach of both artificial intelligence (machine learning) and manual reviews.
Combining Man and Machine Fraud Prevention Techniques for Total Coverage
One of the top challenges retailers face is the delicate line between declining potentially fraudulent transactions and declining a legitimate sale. When retailers refuse a legitimate sale, they not only provide a poor customer experience, but they often do not know about the legitimate decline unless a customer tells them.
Fraudsters are also incredibly sophisticated. They’re intelligent, adaptable, and good at figuring out the holes in fraud systems and exploiting them.
Because of this, it’s become best practice to combine manual reviews with machine learning technologies as a multipronged approach to managing fraud. This balance will protect the customer experience while still providing fast and effective fraud prevention, saving time and money.
Here’s how the two can work together:
Machine learning does the heavy lifting of fraud detection. It should be used to identify suspicious looking transactions. For example, transactions will be flagged if the purchase does not look like it belongs to the consumer who owns the card. Flagged transactions should then be checked by a manual review. This allows the company to investigate discrepancies and see if irregularities can be explained (ex. the order could be placed on a company card and shipping to another address that’s not associated with that consumer). Manual research could determine that the consumer is following the shelter in place orders and has had to ship products to their home address rather than their company. This will ease concerns to let the deal pass, saving the sale and keeping the customer.
This checks and balances approach of data and human touch ensures merchants are capturing a majority of legitimate sales while preventing fraudulent ones. But it can be tricky to navigate for those who are performing in-house fraud prevention. Retailers who take the time to invest in their fraud prevention strategies that balance both machine and manual reviews will build strong customer relationships and revenue streams during and after COVID-19.
* ONLINE TRANSACTION FRAUD TO MORE THAN DOUBLE TO $25BN BY 2020, FINDS JUNIPER RESEARCH – Juniper Research