Risk Management

Ecommerce Fraud Prevention – How to Avoid Fraudulent Transactions without Turning Away Good Customers

Online fashion retailers, especially those selling luxury goods, are often targeted by fraudsters. Naturally, merchants strive to avoid accepting fraudulent transactions and incurring the related chargebacks. Regrettably, this leads to risk-averse policies that result in good customers being rejected along with fraudsters. We have seen many cases of retailers missing out on expansion opportunities due to fear of fraud. In this post, we share some of the classic fraud patterns or things to watch out for, in the fashion vertical, as well as best practices to avoid fraud without turning away good customers.

Payment Frauds in Fashion Ecommere

So how grave is state of fraud in the online Digital Marketing in Fashion Industry? To estimate fraud rates,  we use two central metrics:

  • chargeback rate – the percentage of total orders for which a chargeback was incurred, and
  • decline rate – the percentage of total orders declined by the merchant

This way, the fraud rate reflects the number of orders rejected due to fear of fraud as well as the number of approved orders that turned out to be fraudulent.

Among Riskified’s online fashion merchants, the average chargeback and decline rates is quite low, but given the fact that Riskified approves 66% of orders merchants typically decline, the average fashion e-tailer is probably rejecting more orders than necessary.

By rejecting an order, you lose not only the revenue from the declined transaction, but also the acquisition cost goes down the drain – all the budget and effort invested in bringing the customer to your site and having them place the order only to have it declined. When you take into account the lost lifetime revenue from clients whose order is wrongly rejected – you start getting an idea of the true cost of declines and realize why accurate order decisions are crucial.

Our advice to you – never decide to approve or decline an order based on a single data point. Every order should be reviewed holistically, taking into account all the available information. That being said, we hope the following tips and insights will help you distinguish friend from a fraudster.

3 Tips for Keeping your Ecommerce Fraud Rate Down

It’s common to think that items selling for the highest price are the riskiest. While it’s true that many fraudsters will target expensive clothes and accessories, the fact is that not all items are equal. Certain items are “high-risk” – meaning they are more likely to be targeted by fraudsters either for personal use or for resale value. Some classic examples of items with a high resale value are GoPro cameras, tickets to an NFL game, iPads, gift cards, and diamond rings. In the fashion vertical, designer handbags and hi-top sneakers tend to be most targeted by fraudsters.

Below is an illustration showing the risk of fraud associated with various items sold by two luxury retailers:


Fashion items and risk of fraud, Riskified (illustrative)

Tip 1. Know Your High-risk Items:
Of course, stats differ between businesses. Every store has a unique selection of items, and it may be that for your store a certain bag, or even a specific brand you carry, is targeted by fraudsters. That’s why it’s imperative to track all the chargebacks you incur, especially chargebacks related to fraud, such as unauthorized card use. Once you identify the high-risk items or brands it will be easier to know which orders should be reviewed with more care.

Shopper behavior is another key factor to consider when thinking about preventing fraud. Unlike in bricks and mortar store, there’s no way to interact with eCommerce customers in person –  no magnetic security tags, no cameras in the fitting rooms. However, a shopper’s browsing history and online behavior can provide great insight and help determine whether a transaction is legitimate.

Tip 2. Pay Attention to the Shopping Cart:
One indication of shopper behavior is the item mix in their shopping cart. Many merchants track customer behavior around the shopping cart for purposes other than fraud – such as recommending items to “complete the look”, but this same data can be used to prevent fraud.

Does the item combination within the cart “make sense” or does it seem suspicious? While it’s easy to imagine why a customer would purchase the same blouse in three different colors, a cart with several tops in a wide variety of sizes might seem “fishy”. Similarly, purchasing one designer handbag is more common than buying a dozen designer handbags at once.

Every rule has an exception, and we have seen cases of orders with bizarre shopping carts that were perfectly legitimate. For example, an order for 10 vintage coats placed by the celebrity wife of a Hollywood star, or the time a movie production company purchased clothes for the entire cast in a single order.

Tip 3. Track and Analyze Browsing History:
Another great indication of online behavior is browsing patterns. Overwhelmingly, a regular customer will take time to browse your site, look at several items, click to see the available colors and item photos from various angles. When purchasing a high value product, legitimate shoppers are likely to check your store’s returns and guarantee policies. Fraudsters, on the other hand, often have very different online behavior. For example, a fraudster might sort items from high to low price, immediately add the most expensive item to the cart, and proceed to checkout without browsing any further.

Good browsing behavior (illustrative)

(click to enlarge)

Fraudulent browsing behavior (illustrative):


 (click to enlarge)

The visuals above illustrate the difference between the online browsing behavior of good and fraudulent customers. Keep in mind that it is likely that legitimate shoppers will visit your site more than once before placing an order. In order to correctly analyze shoppers’ on-site behavior, you must consider all of their browsing sessions. We recommend collecting all of this data, as it will allow you to get a full picture of user behavior on your site.

Turning Away Good Customers is SO Last Season

Most online risk management tools flag potentially fraudulent orders based on data mismatches or statistically “high risk” indicators. As a business owner relying on such solutions, you run the risk of higher than necessary decline rates, as well as of an incredibly time-consuming process. You can either decline all flagged orders to avoid fraud, while turning away perfectly good customers in the process, and/or you can spend considerable time reviewing all these orders for fraud.

Riskified’s technology, however, was designed to approve as many orders as possible. Since the vast majority of online orders are good, if we can help it, Riskified wants to make sure legitimate customers are never wrongly turned away. Therefore, we’d like to share certain patterns and behaviors that might generally be considered to be indicative of fraud, but that are commonly found in orders placed by legitimate customers in the fashion vertical:

  1. Reshippers:
    Reshippers, also known as package forwarding services, receive a parcel on the customer’s behalf, and ship it to its final destination. It’s clear why such services are attractive to anyone committing fraud. Much like a proxy, a package forwarding service allows a fraudster to conceal his or her true location and identity.
    However, when it comes to the online fashion vertical, plenty of legitimate shoppers use reshippers. In some countries, such as Nigeria and island nations in the Caribbean, package forwarding services significantly lower the shipping price, or allow shoppers to receive items from shops that don’t ship to their country.
  2. ‘Corporate’ shopping:
    As mentioned above, high cost orders with many items or a strange item mix may seem suspicious. However, whether placed by fashion boutiques, model agencies, or movie sets, in the online fashion vertical it’s not unusual to see a single order with dozens of clothes and accessories. We often see high ticket orders placed by fashion resellers based in China and Japan. Since these orders can be worth a lot of money, it’s important to understand who the buyer is. Even a quick online search can help “explain” the order and ensure you don’t miss out on the revenue and reject a good customer.
  3. Celebrity purchases:
    Yes, some items are at a higher risk of being targeted by fraudsters, and indeed, the average customer does not purchase a dozen pairs of vintage shoes at once. Especially if you’re in luxury, vintage, or high-end fashion retail, however, it’s crucial to remember that everyone shops, including celebrities.
    The last thing you want to do is unwittingly turn away a high-profile customer who can help generate buzz around your store.
    Unsure about an order of designer clothes or ultra trendy sneakers? Try searching the customer’s name online. You might find that a celebrity from a foreign country is shopping on your site!

Online fashion retail is a booming industry, but it’s not an easy business, and fraudsters definitely don’t make it any easier. We hope this post has provided some ideas for ways to protect your business from fraud. We love talking about fraud prevention and online risk management, so if you have any questions please reach out! If you’re interested in learning more about reviewing orders for fraud, check out the Riskified resource center.

About Author

Shalhevet Zohar, Content Manager at Riskified
Enthusiastic about data analysis, writing and travelling – and about helping online merchants avoid CNP fraud.

Recommended Posts