Being able to effectively identify and proactively prevent fraudulent

transactions is vital if financial institutions are to look after their

customers and keep their own losses to a minimum.

As technology has advanced in recent years, banks now have far more

tools in their armory to assist with this. For instance, the era of big data

analytics means it's far more feasible now to study huge amounts of

information, such as buying patterns, in order to spot key red flags.

Despite this, losses are on the rise. In 2014, figures from the Nilson

Report revealed global losses as a result of payment card fraud reached

$16.31 billion, an increase of 19 percent from the previous year. 

This shows there's still much more to be done to ensure banks are able to accurately spot fraud. After all, it can be easy to implement an overly

cautious approach that results in a high rate of false positives (where a genuine transaction is incorrectly declined) - something that can be extremely irritating to customers. But by the same token, a more passive approach may let genuine fraud slip through the cracks. 

A holistic view of your customers

To counter this, it's important that banks take a complete enterprise view of their customers, rather than focusing

more narrowly on a few key factors, such as a user's history on a single debit or credit card.

In order to determine if a transaction is genuine, financial institutions need to look at everything they know about

the customer, from their geographic and demographic profile, to how their activities relate to wider trends and

historic behavior.

For instance, mobile transactions are particular vulnerable to fraud, with figures from LexisNexis suggesting m-

commerce accounted for more than a fifth of fraudulent transactions in the US in 2014, despite only making up 14

percent of transaction volumes.

Therefore, it will be understandable if banks pay particularly close attention to this channel - but they must also

consider a customer's profile when doing so. So for instance, if most of an individual’s normal interaction with

their bank comes from the mobile channel, does this mean an m-commerce transaction is more likely to be

genuine?

Using all the available data

In order to stay one step ahead of the fraudsters, it will be vital that banks use every source of information

available to them, and are prepared to share data when relevant.

For example, eBay Enterprise notes that its researchers have spotted a spike in fraudulent online transactions at

around 8am on Wednesdays - timed so that goods will arrive just before the following weekend.

This type of understanding into patterns of fraud can help banks develop a better overall picture of what a

fraudulent transaction may look like. These patterns are often difficult to determine in isolation, but when data

from multiple sources is analysed and applied to a customer's profile, it becomes easier to determine what

constitutes normal behavior for that individual.

Understanding how a person normally interacts with their bank - from what channels they prefer to use to when

they do their shopping or banking - can help financial institutions make sure any alerts correspond to their known

behavior, therefore making fraud detection much more accurate.