facebook
| Oct 14, 2022

Friendly Fire: Fighting Consumer Fraud With Machine Learning

With consumer activism on the rise, a chargeback culture in the U.S. is costing businesses a staggering $125 billion every year. AI-powered platforms are leveling the playing field.

So-called “friendly fraud” has surged since the pandemic began with more than 44% of Britons and 66% of Americans admitting to filing chargebacks over the last 12 months. The problem is that friendly fraud rarely leads to consequences for consumers; instead, it forces business owners to foot the bill. 

As of this year, an astounding 94% of merchants say that friendly fraud is an issue for their business, but fewer than three in 10 have found a successful strategy for dealing with it. Friendly fraud doesn’t come cheap, either, with recent estimates pinning the annual cost at $125 billion.

In response to the chargeback crisis, brands like Mastercard and Visa are investing in artificial-intelligence solutions that aim to fill the information gap between card issuers and merchants to eliminate the problem for good.

What is Friendly Fraud? 

The mechanism for requesting a chargeback exists to protect consumers from unauthorized transactions, like fraudulent purchases made by someone who stole their card information or recurring charges for a subscription that they previously canceled. However, roughly 63% of chargebacks fall under the category of “friendly fraud” because the disputes involve transactions that the cardholder did authorize. 

A consumer might dispute a legitimate transaction on their credit card or bank statement for any number of reasons, such as if they think it’s fraudulent or feel they are entitled to a refund. For instance, a consumer might initiate a chargeback if they don’t remember making a purchase, missed a return window, or simply didn’t like an item.

However, a shocking number of people are now using chargebacks as a way to “punish” merchants whose values differ from their own. An astonishing 44% of U.S. consumers have admitted to doing just that, dubbing it “chargeback activism.” Friendly fraud is also surging for crypto and buy now, pay later (BNPL) transactions for far less noble reasons — simply because consumers don’t want to pay. 

“Consumers on both sides of the Atlantic increasingly see chargebacks as simply part and parcel of the retail process — a protection they’re entitled to avail themselves of if they feel in any way disappointed by their shopping experience,” says Roenen Ben-Ami, co-founder of tech startup Justt. Unfortunately, the impact on merchants is anything but trivial.

How Much Does Friendly Fraud Cost Businesses?

Merchants generally have 45 to 60 days to work with a card issuer to prove that a transaction was authorized, but this process adds to a business’s overhead costs. What’s more, merchants must pay a fee for every dispute, which can range from $20 to $100 or more per chargeback. Plus, even if the business is successful in canceling the chargeback, this fee is not refunded.

Once fees, labor, and lost merchandise are factored in, it’s estimated that merchants lose an average of $240 for every $100 in chargebacks.

Now, with friendly fraud increasing at such an astonishing rate, disputes are putting greater financial strain on businesses that have already had to overcome two trying years. 

Regardless of why a consumer files a chargeback, merchants are rarely given the option to make things right. Even though more than half of consumers say that generous return policies would make them change their chargeback habits, nearly 60% of consumers didn’t even reach out to the merchant before filing a claim.

To consumers, chargebacks are better than going to the merchant because the filing process is quick and painless: Their money is credited immediately and they don’t have to deal with returning faulty or unwanted items. This lure of convenience combined with new ulterior motives like chargeback activism have driven friendly fraud to new highs. Now, businesses are desperate for a means to fight back. 

MORE FOR YOU

Fighting Fraud With Machine Learning

Almost half of all instances of friendly fraud are the result of mere misunderstandings, like a consumer not recognizing the billing descriptor on their credit card statement. For this reason, fraud-prevention platforms are investing heavily in sophisticated products backed by machine learning. The biggest players include Mastercard’s Ethoca, Visa’s Verifi, and tech innovators like Justt

Using collaborative networks, fraud-prevention platforms pull data from card issuers, partners, and merchants to provide real-time insights into potential fraud. By identifying the merchants associated with transactions, pinpointing when and where purchases were made, and even breaking down fees that may have created confusion over a price discrepancy, businesses are armed with the information they need to handle transactions more effectively.

These platforms are also attempting to tackle the issue of chargeback activism and instances where consumers just don’t want to pay. As more data is added to these networks, merchants can be warned of serial chargeback filers to deny or cancel transactions before friendly fraudsters have the chance to cost them time and money.

Everyone Benefits, Except Fraudsters  

Honest consumers stand to benefit from these solutions, too. Information on confirmed fraud and recent customer disputes is shared to notify merchants if a purchase was likely unauthorized. This allows the business to issue an immediate refund before the cardholder ever gets involved. In turn, the consumer gets their money back sooner while the business can avoid complaints and fees altogether. 

While solutions are still evolving, these technological advancements are finally helping businesses gain the upper hand when it comes to fighting the friendly fraud crisis.

Sydney Chamberlain

Opinion Contributor, Strixus

Sydney specializes in informational, research-driven projects that often tie into her passion for travel, wellness, and technology. view profile

OTHER ARTICLES

Related Posts