Mobile commerce a whole new ballgame when it comes to fraud detection

ecommerceMobile commerce (mCommerce) is already big, and quickly growing. Just consider these two data points: it’s projected that in 2017, more than 2 billion smartphone and tablet users will make as least one mobile transaction. And by 2019, it’s estimated that more than 60% of the total global retail sales will be made on mobile.

The software and hardware ecosystems behind mobile platforms continue to develop, innovate and integrate. As mobile matures, and tablets and smartphones (particularly the latter) continue to become the all-important convergence devices of our connected world, mobile will transform our experience of retail, just as it has profoundly changed communication and multimedia content consumption.

For many consumers in the developing world, their first and primary portal to the Internet is their smartphone, and this is especially true now that lower-cost models are available on the market. Due to the fact that they’ve effectively leapfrogged the traditional desktop or laptop computer, these consumers really are “mobile first”. To reach them, companies need mobile-first website design, mobile-first payment processing – and mobile-first fraud detection.

The threat of fraud on mobile

Mobile commerce is profoundly different from the rest of e-commerce in a key way: it is really a conglomerate of several retail channels. Each of these channels – shopping on a merchant’s standard website from a mobile device, in person payment at a physical location using NFC for payment, using a laptop that uses a mobile device as hotspot…and the list goes on – is a unique way to use a mobile device for shopping, and each one presents an opportunity to fraudsters.

In addition to the diversity of ways mobile commerce can transpire, screening out fraud via the usual indicators of IP address won’t work as simply as it does for traditional ecommerce because IP addresses on cellular devices change much more frequently than fiber, cable and DSL broadband connections do. Also, latency on cellular connections is often significantly higher than it is on those broadband connections.

Fraud prevention often depends on pinning a customer’s identity (especially that of a return customer) to a device. This can work great for customers who shop from home using their desktop: in this situation the IP will very rarely change. On mobile it’s far more difficult because that exact same device can connect to your online storefront over both Wi-Fi and cellular networks, and in both cases that IP address can constantly change. Not only do IP addresses change more frequently on mobile than on desktop or laptop, but a customer’s behavior on mobile is also different, introducing another variable to be accounted for.

No point in rejecting good customers

All these features and variables in mCommerce require a sophisticated, intelligent and mobile-first fraud detection approach to prevent unsustainable losses due to chargebacks and their fees, and to allow legit transactions to transpire. Riskified, a provider of an ecommerce fraud detection solution, uses advanced machine learning algorithms which are adaptable and self-tuning, which enables the system to keep up with fraudsters’ ever-evolving tactics. While there is a real potential to get stung by online criminals, their solution blocks fraudulent orders, yet is also able to identify legitimate customers, reducing the amount of lost revenue due to false declines.

Mobile presents a huge opportunity to grow your revenue and customer base. If your fraud detection system isn’t prepared for the challenges unique to mobile, you could be rejecting lots of legitimate business or letting fraudsters haul away your profits by outsmarting your outdated procedures and software. Consider updating or replacing your current solution and stay one step ahead of the criminals instead of being caught immobile on mobile.