by Carlos Häuser
February 10, 2017

As online fraud becomes more pervasive, merchants need to strengthen their defences against potential cyber thieves – by extending classical risk tools

Online commerce has been steadily growing over the past years. Although ecommerce provides merchants with significant gains and consumers with the added comfort of making purchases in a real-time fashion, fraudsters are becoming increasingly sophisticated and more efficient in identifying and exploiting vulnerabilities.

A successful approach to fighting online fraud is one that detects malicious activities while providing customers with a seamless experience

A successful approach to fighting online fraud is one that can detect malicious and fraudulent activities, while providing customers with a seamless experience. For example, if the fraud protection means taking extra authentication steps like the 3-D secure (an additional security layer for online credit and debit card transactions) it could be that the shopper quits the transaction – which may affect the conversion rate in a negative way. So what sort of possibilities are there that take both sides into consideration?

The benefits of the classical rule set solution

The two current major risk approaches on the market are rule set solution and the new big data method. The main advantages of the rule set solution are the variety of rules that are able to check and correlate different parts of a transaction. Since it is a standard solution that is ready to go and requires no data, merchants can start their business immediately. Furthermore, the rules make the concrete decline reasons recognizable and thus easy to monitor. One rule, for instance, is the IP-BIN check. This common method checks fraud patterns for credit card transactions. The IP address is used to find the customer’s current location or internet access point, while the bank identification number (BIN) identifies the card issuing bank and its country of origin. This allows specific combinations of two countries thus identified to be accepted or rejected. That being said, this strong rule-set may lead to strict declines which can cut further revenues or sales.

The merchant specific big data approach

The big data approach goes a step further and helps merchants who want to stop fraud early. By developing predictive models based on both historical and real-time data as well as experiences the big data approach allows retailers to get a more detailed picture of what typical fraud signals might look like and thus make a more granular decision. This modern fraud detection method provides retailers with connectivity and deep correlation between merchant specific data. A further advantage is the self-learning ability of the system: new rules can be developed around changing circumstances. To use the data approach properly, an information pool is needed as well as current IT security standards.

The best of both worlds: the fully automated weighed fraud prevention

So why not combine both parts, to build up this pool? This could mean less black-and-white rules and more shades of grey. In this case, this would mean a detailed implementation based on a much bigger linking of the regulations between the rules with each rule / decline receiving a score. The rules which produce the most ‘false positive’ results, are assigned to a score which does not produce a decline on its own but a decline in combination with other relevant rules. This process follows a grading of the rules which are rated by scores (determined via an analysis of the past transactions) and if a certain score is reached the transaction is blocked.

In the fight against fraud, it’s clear that big data will play a crucial role – together with the classic rules system

In the fight against fraud, it’s clear that big data will play a crucial role – together with the classic rules system. Combing the two analytical tools will serve retailers and customers: Avoiding false declines will benefit the conversion rate and provide a smooth online shopping experience.