CNP EXPO 2016 Main keywords – online fraud, data and machine learning

Comments (0) Cross-Border Ecommerce, Global Card Processing, International Online Payments, Payment Industry Trends, Payments Events

As card not present (CNP) transactions grow, online fraud is growing too, and with EMV migration in the US it is expected to grow even more. No wonder fraud is so high on merchants’ list of concerns! Next to omnichannel, big data and cross-border, fraud was also one of the main keywords of the CNP EXPO 2016 edition.


Amongst the top challenges in fighting fraud is the immense volumes and velocity of data generated by the big brands. This poses new, unique problems but it also creates opportunities to transform business and create competitive advantage. So as customers give merchants more and more data, if they don’t use it to improve the buying experience and journey across their website then merchants are losing out on a huge opportunity. Together with their providers, web merchants must learn to use data as a competitive advantage.

How can we leverage so much data?

During CNP Expo 2016 panelists from leading brands discussed how data and analytics are being used to address new world challenges and redefine customer experience. However, in today’s marketplaces the tools are very dated and people are sitting on a lot of data without the right tools to make use of it effectively. Hence, the data is wasted and not used for real-time decisions.

Machine learning

Machine learning has increasingly grown in popularity as a platform to help businesses combat fraud thanks to its real-time analysis of data. This enables merchants to make necessary adjustments within minutes or seconds. “Scoring transactions within milliseconds is a challenge that wasn’t even possible a few years ago, so the future is bright”, said Sandeep Grover senior vice president of global ecommerce for Feedzai.

Machine learning has positive ROI as it continues the “learning” process from the user behavior and constantly applies the learned patterns to the future and to predict a phenomenon. It can be used for fraud prevention, content personalization, search rankings, website optimization and so on. It can help save money, maintain good customer experience, help people make the right decision because they get the right tools, and score the transaction even prior to payments.

But when we talk about machine learning models, the industry needs more education and awareness about how to use that data. Industry leaders also recommend a combination approach for fraud prevention, with big data analytics on one hand and a manual review on the other.

‘How do we control the fraud without negatively affecting the good customers?’  was another question that stirred up lots of discussions. How can we drive sales without adversely affecting these customers?

All of the panelists stressed that it is up to merchants to find a balance between impacting good customers and keeping the fraudsters from making purchases, as well as recouping potentially lost sales from identified good customers, as opposed to just focusing on identifying the fraud pattern.

The best interests of the good customers has to be the decision driver of each merchant, because we will never stop fraud 100%. Fraud is evolving and for as long as there’s human nature there will always be fraud. We can only try different prevention models, learn and emerge from them.

Last but not least, in terms of regulation and KYC, one of the main statements to keep in mind as a key takeaway from CNP EXPO 2016 was made by Amazon Payments’ Anastasia Rissis: “Regulation doesn’t kill innovation it just takes more time for the regulators to see and understand the trends. It’s a matter of working together with them, educate them and help them to keep up with the fast pace of the payments industry”.


Browse By Topic

Browse By Expert

Infographics and White Papers

Our infographics and white papers illustrate the latest online shopping trends, preferred payment methods and most up-to-date ecommerce data across key markets.