A tale of two transactions

How the power of data and AI can reduce false declines and improve the customer experience.

Meet Emily.

She just earned a big win at her ad company and a big bonus to go along with it. She wants to celebrate her success with a large retail purchase — a new electric car she’s been coveting. But the splurge is unusual for her typical purchase patterns. Her bank, still operating on traditional guardrails around risk management, declines the transaction. Her excitement wanes, and frustration sets in — she uses another payment method to complete the purchase.

When a customer experiences more than three declines on a card, they are 2.5 times less likely to use that card again.1 False declines, also known as false positives, are legitimate transactions that are rejected by a bank or payment processor but they can be reduced with smarter risk technology.

Let’s try again.

Meet Maya.

Her bank uses data and insights from artificial intelligence (AI) and machine learning data models to help reduce the possibility of false declines and, in turn, customer frustration. She uses her Visa card, powered by advanced risk technology and analytics, to purchase airline tickets online for her upcoming honeymoon. Her card’s smarter data analytics account for her recent spending behavior and approve the transaction, not only avoiding a false decline, but preserving her and her partner’s joy in making a once-in-a-lifetime trip.

Modern fraud detection systems that use AI and machine learning can analyze vast amounts of data in real time, which makes them really good at detecting whether a transaction is legitimate or fraudulent without causing unnecessary false declines. With an intelligent security approach, merchants can improve the customer experience through a more seamless, secure checkout; customers maintain trust in their credit card’s ability to never fail in moments of need; and financial institutions help reduce fraud through advanced decisioning and rules management in real time. 



At Visa, we have been developing our AI in payments for more than 30 years, investing more than $3 billion in AI and data infrastructure over the last 10 years to enable the safer, smarter movement of money and proactively identify and prevent fraud. We help banks confirm that you actually are the person making the purchase, whether you’re buying a new car, tickets to a concert or football game or simply checking out at the grocery store. With our AI-driven risk solutions, we help reduce fraud ($27B in 2022),2 prevent breaches and protect the integrity of the entire payment ecosystem so you can go on to lip sync with your favorite artist, cheer on the touchdown, experience a new culture or try a new recipe at home without ever skipping a beat.

Learn more at Visa Risk Solutions.

VisaNet data in India from Jun ’21 to Nov ’22. Decline Impact Analysis: Pre-Period: Rolling 3-month period between Jun ’21 to May ’22: Analysis Period: Rolling 1-month after Pre-period: Post Period: Rolling 3-months after Analysis period. Eligible Cards: Consumer Credit Cards having one approved transaction in Pre-period and greater than six months on book are considered eligible cards

VisaNet data from Jan ‘22 – Dec ‘22


Tag: Artificial Intelligence Tag: Payment Security Tag: Digital commerce

The Visa Blog Newsletter

Sign up for the latest news and views from Visa