The winning tools to conquer today's fraud challenges

By Amanda Mickleburgh

In today’s fast-moving, digital commerce landscape, competition is rife, loyalty is increasingly hard to build and consumer demands are a moving target. A seamless, fast, convenient purchasing experience is now expected and conversion rates remain a key measure of success. With all this in mind, online merchants can’t afford any friction in the payments process. They also can’t afford the losses that come with the growing threat of fraud.

The fight against fraud

While all businesses are vulnerable to fraud, ecommerce merchants are particularly susceptible to attack – and that fraud can come in many forms and from many sources. The introduction of stronger security measures like Chip and PIN have increased the protections around physical payments, making it harder to use stolen payment cards, or produce counterfeit ones. Fraudsters will always move to exploit the weakest link in the payments chain and this move has gone towards digital payments. In 2017, fraud losses on UK-issued payment cards totalled £566 million, 72% of which was through remote, or card-not-present (CNP) purchases.

While customer transactions are becoming more remote, fraudsters’ dealings are also increasingly well-hidden, as they take advantage of data and cutting-edge technology being more accessible than ever. Sensitive information such as card details and personal data is freely available to buy on the dark web in alarming quantities. Gained from identity theft and data hacks, this information allows criminals to purchase goods with stolen or fake card details with little chance of being detected.

Fraudsters are also often able to gain access to a consumer’s payment accounts, posing as the genuine customer to conduct their activities and purchases through ‘account takeover fraud’. They then gain access to the purchased goods by changing the shipping address or intercepting the delivery.

The increasingly sophisticated methods employed by fraudsters, along with the remote nature of online purchases, make their activities difficult to spot – making it harder to tell customers and fraudsters apart.


Find the value in visibility

Data and sophisticated analytics are a core part of the answer to this conundrum, since they can help merchants build a clearer picture of customer and fraudster behaviour.

Machine learning techniques can help merchants to spot trends and trigger relevant reactions. By applying pattern recognition techniques to transaction data, (from both fraudulent and genuine transactions) machine learning models build algorithms that can predict the probability of a transaction being fraudulent.

When machine learning models are properly trained (using mass amounts of relevant transaction data) tailored by sector and configured by experts, they can help increase fraud detection accuracy by as much as 40-50%*. The increased effectiveness also results in reduced false positives and ensures genuine customers do not get unnecessarily declined or delayed by manual review processes.

Critical to building rich intelligence is a good understanding of fraud trends within and across market segments and geographies. This can only be achieved with a continuous integration of up-to-date information from both internal and external sources. This is where analysing global fraud intelligence gathered from across a variety of merchant businesses can deliver significant value. By offering the opportunity to see details of newly detected fraud, consortium data allows merchants to track fraudsters more easily and block them to quickly contain losses.

Fraud analytics are not just about increasing fraud detection accuracy. By utilising behavioural analytics and screening the customer, rather than just the transaction, analysing fraud and payments data can in fact help to build positive customer profiles and support sales. In practice, positive profiling has been shown to reduce decline rates by 59%**.


Operate in multiple dimensions

A single-layered fraud solution often doesn’t present a difficult challenge for today’s tech-savvy, resourceful fraudsters. Relying on a single tool can also leave merchants open to additional losses.

For instance, while 3DS 2.0 is expected to become widely used (over time) as a standard answer to the authentication requirements of PSD2, there are still ways for fraudsters to circumvent these measures. Also, in cases where an acquirer chooses to apply an authentication exemption, the liability for any resulting fraud could still come back to the merchant. In our 2018 risk review, our research showed that, for select types of merchants, authenticated fraud accounted for more than 80% of all fraud.

Multi-layered decisioning takes advantage of information that is available at or before authentication and during authorisation. The use of multiple tools such as device information, geo or IP location, behavioural biometrics, and scoring using Artificial Intelligence provide a wealth of opportunities to determine the risk associated with a transaction.

A multi-layered approach provides increased protection against attacks, but it also offers enhanced accuracy and flexibility. It is vital for merchants to employ a variety of tools and processes as part of a sophisticated, tiered approach to fraud prevention – one that makes sure merchants are still in control of the fraud prevention process and their exposure to risk. A holistic solution, which combines data-driven intelligence with tailored rules, link analysis, alerts and traditional identity checks, gives merchants the best opportunity to efficiently serve customers and block fraudsters.

An essential layer to a successful fraud strategy is ensure that each facet is tailored by product, customer and payment type, channel and geography to cater for the characteristics of the customer base and the nature of fraud in each segment. Our merchants have found that implementing tailored rules as part of their fraud strategy can reduce chargebacks by up to 80%, helping to reduce the cost of fraud and minimize any impact to customer relationships. 


Run in real-time, all the time

In the always-on eCommerce world, fraudsters and customers can transact 24/7. This means that, whatever payment processing and fraud screening tools are in place, they have to operate in real-time to successfully minimise fraud and expedite genuine customer transactions.

By creating a constant loop of real-time, or near-real-time data feeds into the screening and analytics processes, merchants can ensure that transactions are checked against the latest fraud intelligence.

Taking this a step further, there is the opportunity to operate an ‘all the time’ screening process which can retrospectively rescreen previously approved transactions– powering the ability to stop fraud in its tracks and issue refunds to customer before a chargeback is even generated.

Running a real-time, all the time fraud prevention system enables continuous improvements as new payments intelligence data becomes available, enhancing customer experience and reducing costs.


Constantly adapt, because fraudsters will

It is also vital to remember that fraudsters will constantly adapt to beat the system, and genuine shoppers will continue to shift preferences; trying new ways to shop, acting in unpredictable ways and expecting more from each subsequent shopping experience. For these reasons, it is critical that merchants maintain dynamic solutions to ensure they continue to prevent fraud and increase conversion rates.

No solution, no matter how effective, can be set and forgotten, especially in such a fast-moving merchant landscape. Every merchant’s approach to fraud management must remain fluid, informed, timely and adaptive to ensure revenues and customer relationships are protected.

Where a platform itself performs the majority of the fraud screening or analysis, experienced fraud analysts can offer significant value by reviewing data and trends and taking steps to optimise platform performance. It is this constant monitoring and adjustment process that ensures both fraud detection and the customer experience remain optimal and responsive to change.


Multiple challenges require multiple measures, combined in a single approach

Fraud prevention measures have long been considered as a source of unwanted friction. It’s certainly true that overzealous fraud measures can add friction to the sales process, resulting in cart abandonment. Worse still, untempered fraud prevention strategies can significantly reduce acceptance levels and cause genuine customer orders to be declined – damaging customer relationships and revenues in the process.

However, the cost of fraud is also something that merchants cannot ignore. Beyond the obvious loss of goods and chargeback fees, fraud has a number of negative impacts on businesses. In fact, each dollar of fraud, costs online merchants around $2.94 in revenue through penalties, loss of goods, wasted shipping costs and lost time dealing with chargebacks. If fraud persists, merchants also risk being forced to pay higher transaction fees and could lose their merchant account altogether.

A key goal for merchants today is to successfully prevent fraud without hitting any of these highly damaging obstacles. Achieving this requires implementing multi-layered, flexible, real time fraud prevention strategies and solutions which are specifically designed to address the complex challenges of today’s ecommerce landscape.

*ACI Use Cases 2018

**ACI Positive Profiling Use Cases 2018

Amanda Mickleburgh, Director Merchant Fraud Product, ACI Worldwide

Join thousands of other Online Retail professionals

Get unique insights straight to your inbox for free, and improve your understanding of online retail. Subscribe to Online Retail Weekly now.

Join thousands of other Online Retail professionals

Get unique insights straight to your inbox for free, and improve your understanding of online retail. Subscribe to Online Retail Weekly now.