Fraud continues to be a huge problem for the retail sector. A recent report from analyst, Martec International has revealed that UK retailers experienced losses of £3.4 billion due to theft and fraud in 2012, 10% up on the previous year.

This is just half the story. In trying to protect themselves from fraud, retailers may end up suffering even greater losses. Industry estimates suggest that around 10% of good customer transactions are turned away due to stringent rules-based systems. This serves only to frustrate and alienate good customers, making them much more likely to churn. The resultant loss to business is likely to dwarf losses incurred through fraud alone.

With the downturn continuing to bite, and financial pressures weighing down on individuals across the UK, the risk of fraud is rising but it is even more important that retailers don’t risk alienating their customers and see them defect to their rivals.

In line with this, the UK Online Fraud Report 2012 from payment management company, CyberSource found that the biggest concern for 50% of physical goods retailers is inadvertently turning away good orders. The rigid rules-based approaches to tackling customer fraud that many retailers employ typically result in many false positives – where an alert is generated that a fraud has taken place, when in fact, it hasn’t.

Clearly smarter customer verification is needed to prevent loss of legitimate customer revenue and keep customers loyal and engaged. So what can retailers do to strike the right balance and ensure they are detecting fraudulent customer behaviour effectively without deterring legitimate customers from doing business with them?

Meeting a need

In addressing this challenge, retailers need to deliver a rapid response that effectively tackles the problem by helping eliminate fraud at source. Technology is now available, based on innovative behavioural change identification, which enables them to tackle this kind of issue by quickly delivering accurate, early indications of behavioural change.

By analysing data in real-time as it flows through the system retailers can establish what normal behaviour looks like and quickly pinpoint when activity deviates from that norm, enabling them to stop fraud before it even starts – to the benefit of retailer and online shopper alike.

By predicting the likely action of each customer and detecting anomalies, this approach provides significant assistance to the detection of malpractice due to the way fraudsters continually change methods to avoid detection. Real-time capability is particularly key here. To be effective, solutions applied to these challenges have to be able to automatically process thousands of pieces of data in real time, to allow retailers to identify potential fraud and make instant and objective decisions about whether or not to accept customer transactions. And the ability to work with minimal data is also beneficial in ensuring results are achieved quickly.

The self-learning capability of these solutions is another key benefit for retailers, allowing them to keep abreast of change, and refine their definition of normal behaviour for each individual customer as more data inputs are modelled. Effectively, it gives them the opportunity to get to know their customers better and understand the way they operate, benefits which a rule-based system could never deliver.

The end result is a system that offers accuracy of fraud detection, has the ability to carry on minimising false-positives and also to identify previously unseen types of change, positive as well as negative through an approach known as adaptive behavioural analytics. In turn, it allows retailers to accept more business and reduce the time they spend unnecessarily scrutinising legitimate customers, thereby saving money, growing revenue-generating opportunity and building customer loyalty.

The best systems use adaptive behavioural analytics to identify good behaviour. Customers can be encouraged in their good behaviour, leading them to spend more in the process. Critically, too, they can use the technique to identify bad behaviour, not just in terms of how predisposed they are to commit fraud but also with regard to their profitability. Adaptive behavioural analytics can help a retail business identify how likely it is that they could turn an unprofitable customer into a profitable one, for example.

Looking to the future

In the future, the compelling benefits this approach can bring retailers is likely to confine the rules-based approach to fraud to history. From the consumer’s perspective, this is likely to lead not just to a stop to the irritating block on credit/debit cards when people are travelling overseas but to a range of road-blocks they experience because of inaccurate fraud systems. We could even see an end to genuine orders being turned down due to a rules-based system rejecting orders that don’t meet specific pre-determined criteria but that are otherwise legitimate and acceptable.