The e-commerce sector is growing rapidly with the wide spreading of online marketplaces. Unfortunately, this makes the industry the target of fraudsters. In this article, we discuss e-commerce fraud methods and how to detect them with Datactive.
As of the first six months of 2021, e-commerce volume in Turkey increased by 75.6% compared to the same period of the previous year and reached 161 billion TL*. Such growth has made the industry the target of fraudsters. As a result, e-commerce sites must give more importance to detecting and preventing fraud, against the risks of chargebacks, customer or seller dissatisfaction, and damage to brand images that may occur because of fraud.
Detection of e-commerce fraud with existing systems is generally provided by tracking unusual activities during shopping. However, in some cases, rule-based systems may be insufficient to detect fraud strategies that are disguised as legal procedures by organized groups. Such organized fraud rings use legal ways to bypass rule-based systems, and hide or divide their transactions. Therefore, a comprehensive analysis platform is required for the detection of organized crime rings and their strategies by discovering a crime pattern. To explain this requirement in more detail, we will describe common organized e-commerce fraud methods and how to detect them with Datactive.
Before reviewing organized e-commerce fraud methods and their detection ways, we recommend you to visit the product page to learn more about Datactive.
Orders Given with Stolen Credit Cards
Orders from different accounts with the same card increase the suspicion of the card may be stolen. It can be either a single person or an organized group who shops by using the stolen card.
Through Datactive, inquiries can be made by creating scenarios by using the entities of person, credit card, delivery information, personal information, etc. to discover all the previous activities of the person, previous activities with the same credit card, delivery information of all or specific orders and his or her connections. For example, we can create a scenario by describing that two (or more) people use the same card, but they do not have any relationship with each other (address, partnership, family members, etc.). The results may indicate the hints of highly organized credit card fraud.
Orders from Different Credit Cards, but from The Same IP Address
Orders by using different cards through the same IP address in a short time may also indicate stolen cards or an organized fraud network. In this case, the suspect or suspects can be detected, and the bank can be informed through the queries that are stated in the previous heading.
Unusual Sales Figures
High seasonal differences in sales figures of any company may indicate illegal activities and should be investigated in detail. For example, a company that does not make any sales for three months and sells 500.000 ₺ in the next month may probably involve illegal activity. With Datactive, such fluctuations of companies can be tracked and detected through scenarios that are created to define this suspicious pattern.
Sell a Purchased Product to Someone Else at a Cheaper Price
This is a common fraud scenario and is hard to detect with rule-based anti-fraud systems. In this method, the fraudulent account first orders the product then sells it to someone else at a cheaper price.
The fraudster provides the product delivery to the arranged person by requesting some changes during the order phase. After the product is delivered, the fraudster requests a refund by claiming that he or she has not received the product. This method damage both the e-commerce site and the seller adversely.
With Datactive, it is possible to track these fake accounts almost instantly. For example, if an account has requested a refund too often, especially for high-value products, this defines a suspicious pattern for this fraud method. Moreover, alarms can be set for this scenario to follow them up.
Fake Purchases Made to Earn Coupon or Gift
Many e-commerce sites give coupons or gifts for a certain amount of shopping. Such promotions of the company can also be the target of fraudsters. Fake accounts deceive the system by first ordering and then requesting a refund and earning gift coupons for the purchases they did not make. They can also sell these coupons online.
With Datactive, it is possible to detect such accounts by querying previous refund requests of the account. At this point, situations such as accounts with a large number of refund requests, and orders requested to be refunded in a very short time (according to the date to be given as a filter) define suspicious patterns. It is possible to discover the results of a scenario that will be created through Datactive by querying details such as previous purchases, all purchases above or below a certain amount, all refunds, through a person, an order, or card assets.
Although these e-commerce fraud methods are the most common organized methods, fraudsters may have more complex strategies. While rule-based systems cannot detect this type of organized fraud, Datactive enables to detection of organized crimes by querying crime patterns with many different filtering options and transferring the results to the entity-link analysis view.