Today, with just a few clicks people across the globe can order everything from computers, diapers, car parts, and even fishing gear online, and have it shipped straight to their home. Thanks to the power of the Internet, a woman in Bangkok can purchase a high-end purse for herself from a boutique store located only in Beverly Hills. Our global network of 0s and 1s can send words, pictures, sounds and video to an entire planet of potential customers, giving them an approximation of the full fashion retail experience. Close enough, in fact, to close the sale.

But was that really a woman in Bangkok who placed the pricey order?

Fraudsters’ new sense of style

This new global bazaar comes its own international back alleys. Fraud rings are leveraging ecommerce’s rise to grow their revenue too. The stolen identity and credit card information from a female victim in Thailand can be quite useful for making quick money via buying a high-ticket item and reselling it.

Fashion ecommerce is an appealing target for fraudsters since that merchandise is typically high value and easily resold, due to brand recognition, high consumer demand and the role of (particularly high-end) fashion to signify the wearer’s wealth and status. Since a single item can reap lots of cash when resold, profits can be high since shipping costs are low.

Given this appeal to online criminals and the crippling effect of chargeback costs and fees on a merchant’s revenue, many online fashion retailers are very tempted to adopt an overly conservative fraud prevention process which ends up still harming revenue due to the large numbers of legitimate orders which get turned away due to this practice. In this context, exclusivity is not something fashion retailers should be striving for. All legitimate customers who are willing to pay your asking price for your products should be welcomed by you, not just a privileged few who by chance are able to slip by an ineffective fraud filter.

With machine learning, fraud is losing its edge

False declines of online and mobile fashion orders amounted to $9 billion in lost revenue in 2015. That’s a lot of money down the drain. Another startling statistic is the fact that there are ecommerce fraud prevention solutions such as Riskified that are able to approve and a significant amount of transactions that merchants decided to reject.

Riskified feeds a great deal of data about an order into a well-trained, machine-learning algorithm, thereby picking up on the subtle patterns and commonalities of both legitimate and fraudulent orders. Many of its customers are online retailers in the fashion vertical, and here are some of the insights from their 2017 report about Card Not Present (CNP) fraud in online fashion sales:

  • If a faster shipping method is selected for an order, it could be a sign that the order is fraudulent since fraudsters want to make their money fast and the extra charge doesn’t cost them since they’re not shopping with their own money.
  • Email addresses which are two months old or less and don’t include the customer’s name are riskier than others, since it takes extra time and forethought to create an email address for each stolen identity used.
  • Real customers will take their time before purchasing from your site. They’ll compare different items, check available colors and product photos. For legit customers, this purchase is a decision they’ll have to live with. Fraudsters, on the other hand, spend much less time actually shopping and thus exhibit a very different browsing pattern.
  • Some fashion items are riskier than others: shoes, watches and jewelry are higher risk than underwear, scarves or socks.
  • Just because a customer is using a re-shipper doesn’t mean the order is fraudulent. Legitimate customers use them to lower shipping costs or to purchase products from merchants who don’t ship to their country.

Leaving $9 billion dollars on the table is criminal, especially when it’s due to imprecise fraud prevention systems that don’t stop the real criminals. There’s no need to discount your revenue by turning away customers – get smarter about your screening by using a solution with actual smarts.