Many of us have experienced a rejected credit card transaction when shopping online. Questions flood the mind such as “Did I pay my bill?” “Have I exceeded my credit limit?” “Was my card stolen?”
Too often, fraudulent activity alerts triggered for card-not-present transactions (those conducted online or over the phone) result not from customer actions, but instead from “false positives,” such as a misinterpreted spending pattern or an unfamiliar purchase from a merchant in another country. These events can have a ripple effect for all parties involved in the transaction—from frustrated customers abandoning their purchase to merchants losing revenue, to a strain on fraud investigations at the credit card processor.
Microsoft has been working together with American Express, an industry-leader in fraud prevention, to improve the accuracy of fraud protection for online retailers. The key is enabling merchants and card issuers to share information about the risk exposure of the transaction and incorporating this information in their own evaluation to make a more informed assessment. It’s a model that promises to decrease fraud cost, increase acceptance rates, and improve the customer experience.
Why siloed fraud protection tools increase false positives
Each year, online merchants and card issuing banks suffer hundreds of millions of dollars in losses caused by fraudulent card-not-present transactions. Liability is often on the merchants in these cases, which can be compounded by chargeback fees and a potential increase in bank declines, driving down revenue. Banks suffer losses as well, as more risk-averse merchants may decline their customers’ legitimate transactions.
While regulators are trying to help with initiatives such as the Payment Services Directive (PSD) 2—a European regulation for electronic payment services—merchants and banks are proactively adopting more measures and making increased investments to combat fraud.
One of the most promising technologies to combat fraud is AI and machine learning. Today, merchants and card issuing banks alike are heavily focused on predicting and preventing fraud by relying on their own business intelligence.
Merchants use what they know about their customers, such as insights gleaned from customer account and past purchase information, and any other information they may collect, to combat fraud. Meanwhile, card issuing banks use data about their customers such as the customer transaction and repayment history, as well as merchant data like merchant fraud history, to predict the likelihood of a fraudulent purchase.
While both merchants and banks are working towards the same objective of predicting and preventing fraud, they are using discrete data sets. Typical to any prediction system, these fraud prevention solutions are bound to have false positives, resulting in legitimate transactions mistakenly declined.
The solution? A system that enables merchants to share a bit of insight about their own risk assessment with card issuers—a bridge between discrete fraud protection networks.
Bridging merchant and bank fraud protection networks
Microsoft fraud detection technology safeguards more than 1 billion transactions across Microsoft services every year, which has reduced fraud-related costs by $76 million and boosted revenue by hundreds of millions more over a two-year period. In 2019, Microsoft introduced a commercial version of its proven fraud protection technology, Microsoft Dynamics 365 Fraud Protection, which is equipped with unique, advanced AI and other digital capabilities to help online merchants build trust, streamline transactions, and amplify their fraud protection activities.
That’s where we are teaming up with American Express.
We worked with American Express to evaluate how merchants can seamlessly share risk assessment attributes and transaction data through American Express’ industry-leading Enhanced Authorization solution. Enhanced Authorization is a free solution from American Express which enables merchants to share additional data elements about a transaction with American Express to improve fraud detection. Tested on the Microsoft ecommerce platform, we defined a small payload of data with Microsoft’s AI-based assessment of the transaction, what we call “transaction trust knowledge,” that would be provided through the Enhanced Authorization API to American Express giving them deeper insight about the risks of a single transaction. The test demonstrated that, when deployed widely, merchants can expect an increase in transaction revenue and reduced cardholder friction, resulting in a better customer experience.
Thrilled by the results, Microsoft incorporated American Express’ Enhanced Authorization solution into its Dynamics 365 Fraud Protection technology, so that any merchant can opt-in to send transaction trust knowledge to American Express using the Transaction Acceptance Booster service feature.
At Microsoft, we believe in our mission to empower every person and every organization on the planet to achieve more. And with the Transaction Acceptance Booster feature, Fraud Protection can help merchants reduce fraud, achieve higher bank approval rates, and deliver a better shopping experience to their customers.