Vesta, an end-to-end transaction guarantee platform for online purchases, released a new report analyzing millions of digital transactions it reviewed between Q1 2020 and Q1 2021, the company said in a press release.
Highlights include:
- The percentage of orders identified as potentially fraudulent ranged from 10 to 13%
- Fraudulent transactions’ average value ranged from $126 to $155
- The rate of fraudulent orders varied between 0.8% to over 30% for individual merchants
- 13% of all Q1 2020 transactions were likely fraudulent
In the press release, Ron Hynes, CEO of Vesta discussed the balance merchants must strike between protecting themselves against fraudulent orders, and losing too much revenue by declining legitimate orders:
“On the other hand, if you decline too many legitimate transactions in an effort to fight fraud, you end up with significant losses. For example, if that same merchant doing 5 million transactions per year has an average order value of $125 and blocked 30% of all transactions when only 13% were fraudulent, they’re now losing more than $100 million in annual revenue”, he said. “That’s what makes CNP fraud such a challenging problem to deal with – you have to strike the perfect balance between fighting fraud while maximizing approvals of legitimate transactions.”
In addition to sharing the new data, the report also presents ways for merchants to improve their overall eCommerce fraud prevention strategy. According to the press release, machine learning that can detect sophisticated coordinated attacks by analyzing orders and drawing complex conclusions across multiple merchants is required.
“Unfortunately, we’re seeing fraudulent transactions with indirect linkage increasing across the board, and the average value of those transactions is higher than those with direct linkage, making it an even more expensive and complicated problem for merchants to navigate,” Hynes said. “The only way to effectively identify and prevent CNP fraud with indirect linkage is with machine learning, and it’s critical to use models that have been trained against millions of global transactions and can therefore draw connections that humans simply can’t see.”
Source: https://www.vesta.io/news/press-release-card-not-present-fraud-report-2021
