Facts and Figures About Fraud in Mobile Commerce

By Stephen Fidgeon

We can’t know for sure, but it seems safe to guess that not long after the first human traded — thus beginning commerce — a fraudster came along and tried to take advantage. Where commerce leads, fraud follows. Today, that means that m-commerce is the new frontier, and fraudsters have begun exploring.

While there's no doubt that there's a level of risk in m-commerce, the good news is that these transaction are — on the whole — safer than traditional desktop e-commerce transactions. We can’t say that it will stay that way for long, but merchants should embrace this technology and increase their revenue, confident that they're not being fleeced. For now, merchants largely have the upper hand in m-commerce. If they make smart decisions and adapt their platforms to best approach mobile, it will be a very valuable channel for a long time to come. Riskified's full report is available here, and I suggest you take a look, but I’ll distill a few key points.

I’ll reiterate the most important point from our report: m-commerce is safer than you think. While the safe approval rate for desktop e-commerce transactions is around 90 percent (with substantial variation across industries), mobile transactions have half the rate of fraud. In general, merchants should be approving roughly 95 percent of orders made through mobile devices. Being overly cautious may seem like a smart business decision, but, in reality, it’s leaving money on the table.

Knowing that the vast majority of m-commerce purchases are safe, what approach should merchants take to increase their rate of approval without additional chargebacks? As with all card-not-present (CNP) fraud prevention, retailers should collect as much information as possible. One of the most shocking findings from our report is how many merchants don’t even track fraud rates based on the channel through which an order was placed. Although more than 80 percent of our merchants accept orders through mobile, only 52 percent track fraud rates by channel. That’s a huge data point that they’re ignoring, and it’s extremely likely to skew how they handle those orders. I’ll give you two examples.

I’ll start with what shoppers purchase. A single-item purchase in traditional e-commerce is risky, with a safe approval rate of only about 88 percent. Safety increases substantially for baskets of two to five products — over 95 percent and then settles around there for even larger baskets. In contrast, the safe approval rate of purchases made through mobile devices hovers at around 95 percent for baskets from one to five items before dropping to about 92 percent for larger baskets. That might seem incongruous, but if you think about how people shop, it makes sense. The mind-set of traditional e-commerce shoppers is often to research, consider the purchase and make a deliberate decision. They’re also likely to be looking at prices and making the most of shipping costs by ordering a number of products at one time.

Now think about the last purchase you made from a mobile device. You may have had a conversation with a friend about a product over dinner and decided that you just had to have it. Or maybe you were struck by something on TV while sitting on the couch and ordered an item through your tablet. Many m-commerce sites are designed with this in mind, making it easy to purchase with a single click or swipe to buy. It’s a different mind-set, with the tools to match. That helps to explain the discrepancy.

The second example is to consider the shopping experience — i.e., how consumers interact with the e-commerce site. The very act of shopping on a computer vs. on a mobile device is quite different. Desktop shoppers can see the full site, easily navigate through fields and input their information with a full, traditional keyboard. That isn’t the case with a mobile device. Mobile shoppers may make mistakes in entering their data, but it doesn’t necessarily indicate fraud. The numbers bear this out. An AVS mismatch is a serious red flag with traditional e-commerce, as the safe approval rate goes from about 93 percent on a full match to under 78 percent with only a partial match. For mobile purchases, it’s much less of a concern. Safe approval for a full AVS match is 94.6 percent, and it barely budges for a partial match, at 94.2 percent. That difference is pretty easily explained by big fingers on a small screen.

AVS match, use of a proxy, size of basket, channel used, all of these are important data points that add up to a full picture of the customer. Ignoring any of them will give merchants incomplete information and hamper their ability to make an accurate approval decision. A retailer that considers the whole story when evaluating orders will have fewer chargebacks, happier customers and higher revenue.

Stephen Fidgeon is the director of communications at Riskified, a SaaS fraud and chargeback prevention technology.