Fighting fraud in a diverse payment landscape
The threat of eCommerce fraud in Europe has risen in line with the growth of eCommerce itself. In fact, Europol’s Internet Organized Crime Threat Assessment (IOCTA) 2016 revealed that eCommerce fraud has now been classified as a “key threat”. Is it any wonder when 66% of the EU’s total card fraud is coming from card-not-present (CNP) transactions?
This trend is something that should concern consumers and merchants, but also banks and law enforcement agencies – as more of us conduct business online, we make the online world even more attractive to fraudsters. Each component in the chain needs to be more vigilant and aware of the threats facing them, particularly in Europe where the online payment landscape is more fractured than that in the US.
The EU might be a single market but consumers across the continent have very different preferences when it comes to how they pay for things online. The Netherlands prefers direct bank transfers, Portugal favours the e-cheque, and Scandinavia is embracing alternative payments. In the UK, one in five UK consumers regularly use PayPal to buy goods and services online, whereas in Germany, paying on open invoice is well established as the most popular method of online payment, used by 70% of German online consumers.
Especially payment via invoice gives consumers a chance to make sure they are happy with their purchase prior to payment, yet it demands a great deal of trust from merchants that customers will pay their invoices in full and on time, once the product has been delivered. In 2015, Zalando, the German online fashion retailer, was reportedly defrauded of €18.5m by fraudsters using false names and addresses to order goods and not paying them afterwards.
Credit and debit card purchases, popular in places like the UK and the US, also carry their own risks however. As two of the top three largest ecommerce markets in the world (behind only China), both countries offer lucrative targets for fraudsters. In recent years, cyber criminals have turned their attention from online banking fraud to eCommerce payment fraud which soared by 18% from 2015-2016 in the UK, and 16% in the US.
The rise in fraud has been attributed to an increase in data breaches, along with increasingly sophisticated phishing scams successfully tricking users into giving up personal details, which then enable fraudsters to commit account takeovers, illegitimately ordering items from a genuine customer account.
Despite a diverse payment landscape, there are many similarities between countries in terms of consumer preferences. Overall, debit and credit card payments remain popular, yet the continued rise of alternative payment methods is set to change the landscape dramatically over the coming years.
Meanwhile, the European Union is undertaking its most significant changes to data privacy regulation in 20 years, with the GDPR (EU General Data Protection Regulation), which replaces the Data Protection Directive 95/46/EC.
The intention of the GDPR is to bring harmony to data privacy laws across Europe, protecting and empowering all EU citizens when it comes to the privacy of their personal data. While many organisations sell the idea that greater access to personal data enhances security, this is not necessarily the case.
As a leader in fraud prevention through Europe and beyond, we certainly know the power of data sets and link analysis in identifying and stopping fraud. Personalised information can be kept separate from anonymised data, such as device identification data, while still enabling anti-fraud teams to fight fraud efficiently and effectively.
Fraudsters move quickly and adapt to different payment methods around the globe, looking for weaknesses to exploit. Fighting these fraudsters requires the same speed and ingenuity; increasingly it also demands this at scale, with around 300m online shoppers in Europe spending €510bn+ each year and rising. To do so requires a mixture of technology and human expertise, which is where machine learning analytics can help to let the good transactions through and keep the bad ones out. Artificial intelligence itself is not a golden bullet for fighting fraud, but works at its peak when it is paired with human knowledge and experience.
Fighting fraud requires careful analysis of historic customer behavior in order to predict purchasing patterns. Evaluating existing data sets, including customer names and addresses, shopping basket content, device related information and other purchase data within a machine learning fraud prevention software can help merchants to decide on whether orders are of a higher risk level than they are prepared to tolerate.
But the best defense against fraud is not a man and it is not a machine alone – it is a combination of the two. Machine learning and artificial intelligence support knowledgeable fraud managers in scalable, accurate identification of fraud to ensure they can stop it before it damages their business.
Download the original interview (PDF)