There is hardly an online shop that has not fallen for at least one of the fancy tricks of a cheater’s line and suffered consequently financial loss. One way to reliably detect fraud is classic identification. This helps online shops to distinguish real people with honest buying and paying intentions from fraudsters.
But in addition to technical identification tools, it is always worth a look into the shopping cart of potential customers, since it often reveals in advance how serious the intentions of the buyer really are. There are general indications that can give a first recommendation for action:
Ordering behavior: A meaningful criterion is the number and amount of a customer's orders within a defined period, e.g. five orders within one hour. If the orders deviate significantly from the average behavior, a test is worthwhile. The establishment of maximum order values can be useful here, too. We also recommend limiting the number of orders up to a cumulative limit, which can be completed within 24 hours.
Customer journey: The "fraudster" often fills his shopping cart arbitrarily and quickly, goes straight to the cash register, and usually refrains from further information such as product descriptions. Repeated corrections to the personal data in the purchasing form are conspicuous, especially when the address data changes. Intelligent sensors allow the tracking of related information, which can already be included in the decision-making process during the shopping cart evaluation.
Untypical article number and combination: Depending on the product range, atypical number of products that are bought at the same time, raise the alarm: Orders like 20 razors are obvious exceptions that should be checked again. Similarly, there are certain combinations of articles that require caution. For example, an order for five different coffee machines is a conspicuous indicator.
Product group: There are product groups that are particularly common in shopping carts of fraud attempts. These include so-called "white goods", domestic appliances, as well as products that can be resold particularly well for a high resale value. Shopping carts containing these products should be specially watched.
Processing times: The ordering times can also be used as a sensor for possible fraud orders. For example, large pieces of furniture are usually ordered during the day and less at night. By developing standard profiles, deviant behavior patterns can be explicitly identified. In combination with other sensors, possible fraud attempts can be detected quickly and actively blocked.
To safely identify high-risk shopping carts, the various parameters should be individually combined and automatically checked. Suspicuous orders are that way already recognized prior to the identification test, additional queries used more purposefully, and unnecessary costs reduced. Furthermore, these cases can then be subjected to a more intensive (manual) check or assigned to a "secure" payment method as part of the active payment method control. All in all, the risk of default and associated costs can be significantly reduced with a well-established shopping cart rating.