The threat of fraud and the strategies to combat it might be among the most misunderstood aspects of ecommerce today.
Fraud protection isn’t a glamor game, the way merchandising, marketing and even rapid delivery are. It is an aspect of ecommerce that doesn’t enjoy notoriety when it is working well. Only when something goes wrong is fraud protection — or the lack of it — in the spotlight.
And yet, underestimating the threat can literally destroy a small ecommerce business. And it can do plenty of economic harm to even the largest enterprise. But there might be another way to look at online fraud that lends the pursuit a little more pizzazz.
What if instead of a defensive act, managing fraud was seen as a high-ROI way to increase sales and protect shrinking margins in the retail world? That rather sexy view of fraud protection was laid out recently in a webinar sponsored by Signifyd and featuring a principal analyst and a principal consultant from Forrester.
Brendan Miller, the principal analyst, laid out the state of affairs at the outset of the webinar, “The Total Economic Impact of Guaranteed Fraud Protection in Ecommerce.”
He said the subject of risk, security and fraud was pivotal to the digital transformation of companies, but “too often this is an overlooked subject as a key component of a retailer’s digital evolution and their transformation.”
And he explained why ignoring issues around security and fraud was fraught with peril, running through several growing threats, including an increase in data breaches that put millions of digital identities at risk.
Data breaches are here to stay
“We’re seeing millions of more people go online every year across the globe,” Miller said. “That means that the massive breaches are going to continue to grow. One of our security and risk analysts just said in a headline that it’s going to be more of the same in 2018.”
That was what you would call the bad news section of the program. There was some good news. In particular, Reggie Lau, the principal consultant, walked attendees through a Forrester Consulting study commissioned by Signifyd that looked at the experience of one Signifyd customer.
The June 2017 study, “The Total Economic ImpactTM of Guaranteed Fraud Protection,”(1) looked at the total effect of Signifyd’s fraud protection system on a major retailer, who requested anonymity in exchange for sharing financial information.
Lau said looking at the savings from avoiding fraud isn’t a comprehensive enough view. Turning to Signifyd’s machine-learning-based fraud protection, backed by a 100 percent financial guarantee for any approved, fraudulent orders, has additional benefits as well.
“The business case for Signifyd is more than just comparing chargeback costs vs. Signifyd’s fees,” Lau said during the webinar. “It includes a multiplier for chargeback costs for all the related costs, like re-stocking and overhead. It also includes incremental revenue from higher order acceptance rates. Not including those items would ignore an important part of the value proposition.”
In other words, with fewer chargebacks, there are fewer related fraud costs, including the cost of having to return items to inventory. Moreover, shifting fraud liability from the merchant to Signifyd allowed the retailer to approve more orders. That obviously means more revenue. The machine learning capabilities eased the burden of manual reviews, which take time and delay orders. The move freed up time and money.
Faster fraud review increases customer satisfaction
And faster and more accurate order fulfillment had another, less obvious, positive effect: Increased customer satisfaction. There are few better ways to alienate a legitimate customer than to decline his or her order because you assumed it was a fraudulent order.
Citing the Forrester study, Lau explained that the Signifyd customer saw its order acceptance rate rise from 88.5 percent before Signifyd to 93.5 percent after. Of course that’s important because more approved orders mean more revenue. But it’s important for another reason as well.
Manual reviews are slow. When orders are slow in being approved, customers cancel them and search for alternative retailers. By increasing the number of orders OK’d automatically and quickly, the retailer was able to reduce the number of canceled orders.
“Fulfillment times are especially important during peak or holiday seasons,” Lau said, “when many consumers want their goods as soon as possible.”
The change was evident in the way that the retailer’s customers rated its customer service. The ecommerce manager told Forrester that customers who received their deliveries on time, gave the company an 80 percent customer satisfaction score. Those whose orders were late, provided a customer satisfaction score of 45 percent.
As the retailer’s ecommerce manager told Forrester: “Improving order fulfillment speed not only reduced risk to customer satisfaction, but also reduced order cancellation, especially during holiday season, when everyone wants their order before Christmas.”
The lesson: Speeding up reviews, speeds up orders, which keeps customer satisfaction high.
There were of course additional measurable results presented in the webinar. Referring to the Forrester study, Lau explained that over a three-year period, Signifyd’s fraud protection solution realized a 3.8x return on investment.
Eliminating false declines means a significant sales boost
In all, deploying Signifyd resulted in $3.2 million in revenue from orders that were shipped, but that would have been declined without Signifyd’s machine-learning insights. The retailer avoided another $2.7 million in chargebacks and fraud costs by turning to Signifyd. The merchant picked up another $732,000 in improved order fulfillment and a reduction in cancelled orders and another $479,300 by eliminating the need for staff members to conduct manual reviews.
It marks the sort of financial windfall — $7.1 million in all — that you’d think would get people’s attention. In fact, they’re the sort of results that just might move fraud prevention into the glamor category after all.
(1) Note that the Forrester TEI study was based on one real customer and wasn’t intended to represent a statistically significant sample size.
Contact Mike Cassidy at firstname.lastname@example.org; follow him on Twitter at @mikecassidy.