Phillip Finnegan, Managing Director – Pacific, ACI Worldwide
It’s no secret that consumers continue to embrace eCommerce with enthusiasm, with many shoppers choosing online shopping over a bricks and mortar experience even during ‘traditional’ shopping holidays like Boxing Day. Naturally, merchants are keen to adopt a digital, eCommerce or mobile-first strategy to cater to this demand.
However, it can be tricky to get the balance right between offering a frictionless payment experience and one that offers maximum fraud protection. Err too far on the side of caution and merchants risk leaving dollars on the table through incomplete transactions, while an overly lax fraud strategy will quickly erode the bottom line. Case in point, fraud costs Australian merchants over half a billion dollars annually, with card-not-present fraud accounting for over 84 percent of total fraud.
So how can merchants maximise ROI from their eCommerce activities while combatting fraud? What is clear is that a singular fraud strategy used in isolation is no longer viable. Fraud mitigation involves leveraging multiple layers of protection that work hand-in-hand to detect fraud at the source. Luckily, it is becoming easier for merchants to implement an end-to-end fraud strategy; here’s how.
Use positive profiling to know your customer
Fraud prevention is as much about who you are letting in as who you are keeping out. For example, imagine losing a potential new customer during peak shopping season, just because they have not purchased from you before, resulting in a failed transaction due to overly stringent security. That’s where ‘positive profiling’ comes in.
Imagine that same scenario, except you are able to profile the customer – not the transaction – to see if they have purchased goods from other merchants, have a good payment history and know that all their data checks out. This is the power of positive profiling, where a more accurate and complete view of the customer can be built up through ‘consortium’ data – shared across merchants. Buyer and seller benefit from a seamless payment experience, safe in the knowledge that the transaction is legitimate and secure. Consortium intelligence and analytics also work hand-in-hand to positively profile customers, sorting the good from the bad. By building comprehensive customer profiles based on detailed behavioural data from multiple data sources, merchants can better know their customers.
Implement machine learning to accelerate detection
Machine learning and artificial intelligence is transforming payments and fraud by providing unparalleled speed, processing power and accuracy. For example, machine learning and artificial intelligence can automate the underwriting process by creating risk models using an applicant’s submission and alternative credit data. Machine learning is now being used to power fraud analysis, which can identify potential red flags across applications.
Merchants should start with simple, transparent, supervised machine learning models that take into account all customer relationship data points, to get a holistic view of fraud. As the implementation becomes more mature, merchants can implement adaptive machine learning to improve decision making in line with new fraud trends and reduce false positives.
Minimise points of failure online
You, and your customers, are only as strong as your weakest link. As such there are a few simple measures you can implement to fortify common security pain points online.
Over 20 per cent of all website traffic in 2018 was from bad bots. Website security teams should have access to the most up-to-date security measures, and payment service providers can consult on additional security measures they can provide, such as consortium data profiling.
Data is also your best friend when it comes to fraud prevention. Use historical trends, such as peak buying seasons, to predict future ‘hot spots.’ 3-D Secure, CV2 responses and bank declines are the three leading factors that lead to card declines. Customers who receive a decline may try to repeatedly force the transaction, causing strain on the systems. Web teams should help to determine the cards, IP and email addresses that pose the biggest threats ahead of busy times and troubleshoot common pain points.
Tying it all together
A multilayered approach to fraud prevention and detection is essential for merchants looking to reduce fraud-associated expenses and attract new customers. Positive profiling works for merchants and customers by supporting increased purchase conversions, with reduced friction and higher acceptance levels – essential for business growth. It also reduces chargebacks through accurate customer profiling. This works hand-in-hand with machine learning, which can identify potential red flags across applications by creating sophisticated risk models in record time. Lastly, identifying weak spots in your online experience, including bad bots and fraud ‘hot spots,’ will ensure you are well-equipped to fight fraud and boost business profitability.
