The Hidden Operational Costs of Fraud

Evan Schuman is a guest contributor to the Sift Science blog.

It’s no secret that fraud is costly for online businesses. But are you tracking exactly where your money is going? A new report from Javelin research found that fraud costs merchants more than 7.5% of their annual revenue –  a figure computed by looking at a combination of fraud management costs, false positives, and chargeback losses.

The biggest surprise? “The majority of these costs came from fraud management expenditures, accounting for around 75% of costs,” according to the report. In fact, the study discovered that fraud management eats up a larger and larger portion of merchants’ overall operational costs. “Digital goods merchants are still at the top of the list (23%), followed by hybrid merchants (16.6%) and physical goods merchants (14.9%), increasing from 20%, 13%, and 14%, respectively in the previous year.”

We talked to Al Pascual, a Javelin research director and head of their fraud and security coverage, about how to approach the hidden costs of fraud.

Fraud’s hidden costs are operational

We already know that online fraud prevention is all about degrees. There’s no perfect setting that will stop all criminal activity and permit all legitimate transactions, so fraud teams find themselves balancing a bias toward stopping more fraud (and, in so doing, blocking some legitimate transactions) with letting more legit actions happen (which naturally lets more fraud through).

Managing the tug-of-war on fraud settings can be costly in and of itself. Set those settings too strictly generated some obvious losses (a customer whose transaction gets rejected takes her money and gives it to a competitor) and some less-obvious ones, such as a lot of customer service time being gobbled up by angry customers wanting to have their transaction go through. And then, if you have a top-notch customer service department, you’re also spending money on handing out deep discount coupons as apologies for the customer inconvenience.

And this doesn’t even get into the huge cost of hiring people to manually review suspicious transactions to determine whether they should be accepted or rejected. Earlier this year, Sift Science found that 59% of online businesses facing fraud were concerned about spending too much money on manual review.

“A good part of our fraud spend is on the people side, on operations,” Pascual said. “It’s a constant tradeoff between legitimate transaction volume and the acceptable amount of fraud.”

Reallocating fraud spend

Pascual has a concrete suggestion for how to manage these challenges: “take money from customer service” to pay for more relaxed fraud settings. That idea has a lot of good thought behind it. From a spreadsheet perspective, relaxing those settings will free up customer service to help customers, while also reducing the number of customers your systems have made angry. Also, no need to dole out those discount coupons.

But there’s another point to keep in mind when navigating the balance between stopping fraud and enabling sales. Not every company needs to be as worried about the knock-on effects of false positives (namely, angry customers going to a competitor), Pascual argues. He advises clients to take a serious and meaningful look at competitive situations, zeroing in on the specific products they’re offering when evaluating risk. What are the chances of a customer who is falsely rejected due to overly stringent fraud settings taking their business elsewhere? “In a space where there are not a lot of competitors, you can be a little more cavalier with false positives,” he said.