By Omri Dubovi1
1VP of Product Management, , SafeCharge, a Nuvei company
The role of payment and its impact on organizations has evolved considerably over the last few years, from being purely a commodity to becoming a strategic business parameter.
Businesses have realized that focusing on making their payments process life cycle efficient can be extremely profitable, with the global volume of online payments set to increase by 11% a year between 2015 and 2020 (source: Cap Gemini & BNP Paribas report).
While this boom in online payments brings a wealth of opportunity for businesses, it also raises some serious challenges. Merchants now need to worry about increased fraudulent activities as well as increased competition.
In a world where technology is evolving rapidly, it is necessary for businesses to keep up with technological progress in order to stay relevant. Artificial intelligence (AI) and machine learning (ML) are creating new opportunities for businesses in healthcare, IT and even manufacturing. A conscious use of ML and AI in the payments space can be a solution to a wide range of business complexities and a smart tool to fuel further growth.
Many consumers might not realize, but every transaction made contains a wealth of information for the merchant they are buying from. During an online payment process customers are sharing what they bought, the platform they used, device model, operating system, time zone, general location, delivery address, device ID, mobile fingerprints and more. If businesses input this data into a payment system that uses AI or ML, they can collect this meaningful information, analyse it and produce tangible benefits, such as increased security, better customer experience and more sales.
While there are many benefits from integrating a payments environment with AI and ML, this chapter will discuss four of the most interesting uses with the purpose of showcasing how businesses can leverage the collected data to detect and prevent fraud, design successful payment processes with smart routing, improve customers’ user experience and provide deeper intelligence for merchants.
The increase in online payments brings the unavoidable risk of digital-payment fraud. A traditional rule-based – and largely manual – fraud detection system tries to spot criminal activities by monitoring a range of variables, such as location, the type of merchant and the amount being spent. For example, if a customer suddenly appears to be spending more than usual on a given item with an unfamiliar merchant in a previously unvisited location, this activity will most likely be flagged as a possible fraudulent transaction. However, the problem with the standard fraud detection system is that the rules are too rigid, and in today’s highly digitally dependent world, many purchases do not fit into a rigid rule-based model of fraud detection. Using an AI-based system, businesses can perform fraud checks in the blink of an eye. A financial institution that integrates AI and ML modules into its payment infrastructure can not only detect fraudulent activities in real time, but also recognize anomalies and successfully distinguish and predict fraudulent transactions.
The flip side of this, however, is that AI is only as good as the data it receives. When poor data is fed into a system, all that an organization can expect is poor output. Therefore, a high level of due diligence is required to make sure that the data which is being used has been properly governed, otherwise it could incorrectly categorize customers’ risk scores, causing a frustrating customer experience, and high chances of fraud.
Most critically, systems can learn from each transaction, constantly improving and becoming more effective. The use of AI can allow payments companies to look at transaction data in new and more effective ways, growing the amount of successful legitimate transactions while shrinking the number of illegitimate ones.
Payment technology can do more than manage payments. Intelligently designed systems have the ability to not only increase conversions but also provide various ways to optimize the payments journey, resulting in security and business growth. Each transaction on a business platform needs to be approved by financial entities, such as acquiring and issuing banks. A payments system that is integrated with ML and AI technology can use smart routing technology that intelligently calculates various parameters and routes transactions to specific acquirers to maximize approvals. To ensure no transaction loss, soft declined transactions (tech failure, time-out, etc.) are automatically rerouted. For businesses with more than one acquirer, transactions can be routed based on “pricing” or “acceptance rates”. The routing rules are designed to meet individual business requirements. ML can maximize card payments acceptance rates by ensuring that transactions will go through the optimal route based on a merchant’s payment preferences; efficiency is increasingly improved and processing time for payments can be tangibly reduced, in turn decreasing any human error.
Through ML algorithms, payment companies can analyse data in new and innovative ways to better understand their customers. With AI and ML, payments companies can search rapidly and efficiently through their payments data beyond the standard set of factors like time, velocity and amount. ML can predict customers’ behaviour and convert this knowledge into better customer segmentation. Cross-selling and upselling can be optimized by providing a single view of the cardholder data across multiple channels. Payments data – which include point-of-sale data, data collected from Google ads, and merchant sale site data – can all be used to predict trends. Merchants can then better understand what they should sell and how they should set it. For example, Harley Davidson, one of the most popular motorcycle companies in the world, used the AI tool to analyse existing customer data from the company’s CRM. On analysing parameters like which customers had completed a purchase in the past, how much time they spend on the website, high-value customers, and so on, the AI tool was able to predict campaigns that would work, and those that wouldn’t. By using AI, Harley Davidson was able to increase sales leads by 2930% in a span of three months.
Payment providers have been collecting large amounts of data for some time, and businesses are realizing that they can use this data to benefit merchants. ML algorithms can analyse transaction data to find patterns – seasonal dips in revenue, for example – and help business owners plan and compensate. They can also provide targeted marketing capabilities like reward programmes and analytical dashboards to help business owners manage their inventory, capture new sales and optimize their businesses for each consumer. By integrating ML and AI into the payments process, merchants have the possibility of seeing the entire picture, from the shop floor to stock rooms. Merchants can have a better understanding of what the clients want and, therefore, remain more profitable as a business.
Taken as a whole, AI holds many promises for payments technology companies by providing a more powerful payments product, driving consumers and merchants toward more digital commerce opportunities, and by creating a safer and more secure ecosystem. The potential of these technologies and how they transform the payments process and customers’ experience will only grow in the coming years. Payments have evolved from being a cost centre to be a defining factor for business success, and the use of ML and AI can put the power back in merchants’ hands for the future.