Six Tell-Tale Behaviors that Predict Retail Fraud
May 17, 2022
The Bolt Team
Advances in technology have improved how retailers create memorable shopping experiences. Today, retailers can launch new shoppable surfaces on social media or convert customers with one click. Each of these is only possible because of innovation.
But behind tech’s growing influence on retail linger some concerns, notably the surge in retail fraud. As technology has evolved, so have cybercriminals’ facilities to exploit retailers. The answer to this problem is not more technology; instead, technology designed with an understanding of human behavior.
Here will look at common behaviors that motivate fraud and how retailers can stamp it out.
Fraud: 5 Letters that Keep Retailers Up at Night
Fraud leaves a bad taste in shoppers’ mouths, but it doesn’t have to be that way. Learn how retailers can protect their customers and sales from fraud.Download Report
What’s the actual cost of fraud
This next wave of commerce has birthed what seems like an endless supply of fraud. A perfunctory Google Search for “types of retail fraud” surfaces phrases like BOPIS fraud, chargeback fraud, friendly fraud, and much more.
And even though something like friendly fraud sounds amiable in name, it, along with other types of fraud, costs retailers billions in lost sales each year. A Juniper Research study found that companies lose $20 billion to fraudulent transactions every year.
That number may be misleading, though. Retailers don’t lose $20 billion of goods and services to fraud each year. Instead, retailers pay up to $3.60 for every dollar stolen.
This number includes operational expenses, acquiring bank fees, and credit card processing fees—many of which retailers never recover. What’s not included is the long-term reputational damage retailers incur after a fraudulent transaction.
Our recent study with YouGov found that 71% of customers would view a brand unfavorably if their information were compromised. It’s difficult to assign a dollar amount to an agitated customer, but if it were possible, the number would be high.
Six common behaviors of retail fraud
Many retailers have, as a result, scrambled to find a solution to this $20 billion problem, and for a good reason. The monetary damage caused by fraud dents retailers’ bottom line in the short term but reputational damage from a stolen identity or other fraud lasts much longer.
Getting in front of fraud, and its consequences, may seem like a daunting task, but with an understanding of human behavior, retailers can spot and eliminate fraud. Here are six of the most common actions to watch.
Different shipping and mailing addresses
One tell-tale sign of fraud is when a customer lives at one address and ships orders to another. Of course, there are exceptions to the rule. Someone may be sending a gift, or perhaps they’re in the middle of a move. That said, in general, fraudsters won’t use an address that implicates themselves.
Repeat digital window shoppers
Like a shoplifter, an online fraudster may stake out a store before committing a crime. They visit a website to find vulnerabilities that they can later exploit.
Purchases that escalate over time
After doing some surveillance, a fraudster may be ready to act. The first purchase will look benign — just enough not to raise any suspicion. Over time, however, the value of purchases gradually increases to a point where it dents retailers’ bottom lines.
Preference for expedited shipping
Some frauds, like friendly fraud, are predicated on the ability to purchase and return goods. Doing both can exploit breakdowns in communication among a retailer’s disparate systems. For example, some bad actors will receive and return a package before the retailer processes the initial delivery.
Shopping at erratic hours
Not all 3 AM purchases raise a red flag. But when a cardholder accustomed to a routine way of life buys groceries in the middle of the night, it suggests something is awry.
Frequent use of VPNs
A recent study found that one in five retail fraud cases involves a VPN. That is a service that protects users’ internet connection and masks their identity. It’s no wonder why bad actors are fond of VPNs. Having a tool that conceals their identity makes it much harder for retailers to trace the identity of a bad actor.
Each of these behaviors in itself is not always a cause for concern, but when shoppers check a few boxes, retailers should jump into action.
What motivates fraudulent behavior
Psychologists and behavioral economists have spent countless hours understanding what drives fraud, and they found a few common threads among fraudulent transactions.
Most bad actors set a number that they believe puts them at the least risk of getting caught. Usually this acts as an anchor for the fraudster to vacillate around. Some will gradually increase their purchases from the anchor to see how much they can get away with. Others will gradually lower their purchases below the anchor so as not to set off any alarms.
The old saying success breeds confidence, unfortunately, applies to fraudulent transactions. Every successful piece of fraud inspires bad actors to act even more brazenly the next time. In ecommerce, overconfidence can be easy to spot. Retailers see it in the number of purchases that escalate over time or erratic shopping behavior.
The internet can sometimes feel like a visit to the circus. Everywhere you turn, there is a new spectacle to see, sound to hear, or game to play. Sometimes websites feed into this frenzy when they do too much to impress and attract potential customers. But the bells and whistles can backfire when retailers don’t think through the vulnerabilities they’ve created. In other words, sometimes less is more.
Implement technology designed with an understanding of human behavior
Retailers don’t need a psychologist on retainer to determine whether a transaction looks like fraud. Modern technology allows retailers to take a systematic approach in spotting suspicious activity.
At Bolt, our fraud prevention technology analyzes over 200 behavioral signals to ensure retailers approve more good orders while still rejecting fraudulent orders. These invaluable insights fuel our supervised machine learning models that identify risky behavior and let retailers intervene before it’s too late.