Open Banking and Machine Learning Help Against False Rejections

Machine Learning

Elizabeth Grahamproduct manager enter sectsays that false declines carry significant risks for banks and merchants where consumer loyalty matters.

As a result, merchants and issuers are losing money through false declines. More money is lost than potential fraud, Graham said.

Graham told PYMNTS:

As she explained to PYMNTS, a false rejection (a type of false positive) occurs when a legitimate customer attempts a transaction and is ultimately denied the transaction. She said the problem could be on the issuer or the merchant side of the commerce equation. A false positive when a legitimate customer is flagged for fraud will falsely label the transaction as risky.

According to Graham, regardless of the nomenclature, they are all pretty much the same in terms of customer experience, resulting in unpleasant digital journeys and considerable reputational damage.

“It’s incredibly frustrating and not only do consumers find it annoying, but it also feels like there was a personal insult,” she said.

She cited her own recent online experience as an illustration. Graham said he recently completed a major home renovation and was shopping online for big-ticket items like sofas. At checkout, the transaction failed — for no discernible reason, Graham said whether the reason was merchant-related, bank-related, or credit-related with his card problem. He says he didn’t know at all if he was there.

There are also negative spillover effects on the other side of the equation, leading to increased operational costs for merchants and issuers. The increasing number of customer complaints coming through customer service calls, emails and chats consumes staff time dealing with these complaints.

A Complex and Growing Digital Ecosystem

Friction may not be so surprising. Graham said e-commerce is enabled by a complex ecosystem involving merchants with fraud and analytics tools and issuing banks with their own risk management systems and parameters. In an environment where digital commerce has grown exponentially due to the pandemic, multiple lines of defense are required.

“An entire industry of companies exists to provide tools and services that detect or attempt to detect card-not-present fraud before a transaction is authorized. We use these fraud management software and services to approve or deny credit card transactions at the payment stage,” she said.

Cautiousness of banks and merchants, as noted above, has caused an increase in false rejections that exceeds the amount of fraud believed to have occurred.

Abandoned shopping carts lead to irrecoverable lost sales. Merchants who see consumers leaving their sites are more likely to see dissatisfied customers flee to competitors. Consumers tend to put rejected cards in the “back” of their wallets and choose to use non-rejected cards, so they lose money.

Technology, of course, offers a variety of instruments and options that stakeholders can employ to combat fraud more effectively and enable “good” customers to transact. No strategy is 100% effective, but there are some promising developments underway.

“Open Banking payments integrate with existing options such as credit cards, debit cards, and digital transfer services typically offered by e-commerce retailers at checkout, thus alleviating some of the problems experienced with false rejections. I can,” she said. With Open Banking, anyone with a bank account can initiate quick and secure payments, she added.

She said open banking has been around for a while but is still not an industry standard.

But as she pointed out, “All payments go through strong authentication, through banking apps and biometrics.” He added that there is an opportunity to challenge customers and strengthen authentication protocols. Additionally, no card details are shared with merchants through open banking.

Merchants don’t hold that data, Graham says, and don’t need to implement a strict set of risk rules for ongoing transactions that they may ultimately be held responsible for.

Machine learning can help reduce friction

No conversation about advanced technology is complete without a discussion of artificial intelligence (AI). Machine learning is particularly valuable in online commerce as it helps to abandon the rigidities of traditional rule-based systems that typically flag a series of transactions. She said machine learning models could help coordinate the seasonality of shopping seasons and cross-border commerce by looking at buying behavior across large consortia of consumers to identify and understand patterns. said there is

It’s important for merchants to constantly review their fraud prevention solutions and work with publishers to share insights and streamline online commerce, she said.

“As e-commerce booms, increasingly sophisticated online consumers will spend their time transacting with retailers who offer a seamless digital experience, of course with a high level of security.” she told PYMNTS.

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *