As peak season fast approaches, more people of all ages will be powering up their phones to browse and buy. How can merchants dial up their anti-fraud efforts and keep mobile checkouts ringing safely?
Whether through a mobile shopping app or nudged from an email link to the webstore, using a mobile wallet or a social “buy button,” there’s more opportunities than ever for mCommerce.
ACI’s own customer data shows that on average almost a third (30%) of customer ecommerce transactions are now from mobile devices. In the U.S. it is as high as two in five (42%).
Mobile is a key part of the brick-and-mortar experience too. Consumers are using devices in store for “phygital” (physical digital) services like click-and-collect and buy-online-pay-instore (BOPIS), to make mobile contactless payments and to access “buy now, pay later” (BNPL) options at the point of sale.
Unsurprisingly, there’s been a 55 percent rise in high-risk traffic from mobile devices and fraudsters are piling in to take advantage.
According to the latest ACI eCommerce Fraud Index, mobile fraud attempts increased by 1.22 percent in the first half of 2020, and a further 1.32 percent in the first half of 2021. BOPIS channels had a 7 percent fraud rate compared to other channels at 4.6 percent. Meanwhile, mobile-oriented sectors like gaming and telco experienced the highest levels of attack, and we saw the migration of bot attacks to mobile environments.
mCommerce is expected to be at the heart of many retailers’ holiday sales strategies in 2021 with mobile apps being singled out as a key success factor
Protecting this vulnerable card-not-present channel against payments abuse, including account takeover, synthetic fraud, spam, phishing, promo scam and repetitive bot attacks, calls for better anti-fraud strategies.
Many merchants are hoping that smoother mobile authentication, using the new 3DS2 standard, will help. However, early feedback from some merchants indicates that false declines are unexpectedly high, dependent on the connector. Certainly not the result they want heading into Black Friday and Cyber Monday.
The answer? End-to-end, multi-layered fraud prevention strategies that go beyond compliance
Positive profiling uses artificial intelligence (AI) and machine learning to analyze existing customer behaviors – based on existing data from transactions, GPS and geo-location tracking – to separate genuine mobile transactions from fraudulent ones and drive false declines and chargebacks down.
However, experience shows that traditional machine modeling solutions need to be continually retrained as fraud patterns change. Often trends are only revealed a few days – or even weeks – after they take place.
It’s clear that fraud detection needs to get smarter and faster to stay one click ahead. One way to achieve this is using incremental learning, a new iterative and real-time approach to modeling. This helps by:
- Enabling models to “think and adjust” for themselves, ensuring they’re always hyper-relevant
- Automating changes to remove pressure on anti-fraud and tech teams
- Outperforming traditional methods over longer time periods
- Working better for mobile payments where context and location continually changes.
Seasonality, economic fluctuation, changes in customer behavior and new mobile fraud attack strategies constantly emerge. Mobile criminals adapt fast – so should your fraud protection.