Voice of the Industry

How Signifyd brought 3.8x return on anti-fraud investment to a major retailer

Thursday 12 July 2018 08:00 CET | Voice of the industry

Combining machine learning with human domain expertise helps merchants protect themselves against fraud. Mike Cassidy of Signifyd shares more details on ways to achieve it.

The major multi-channel electronics and home appliance retailer was under siege.

Its fraud-management shortcomings, which had been an irritant, were becoming a major strain on margins as its business changed and grew with the changing habits and demands of consumers.

The retailer, which requested anonymity in return for sharing its financials, needed help and it turned to Signifyd’s guaranteed fraud protection model. The result: a 3.8x return on investment, as described in The Total Economic Impact Of Guaranteed Fraud Protection, a June 2017 commissioned study conducted by Forrester Consulting on behalf of Signifyd.

But the raw number hardly tells the full story.

We’ll start at the beginning, with the retailer’s key challenges. The IR 500 retailer had built its fraud protection almost as an afterthought. It relied on a rules-based system to flag suspicious orders, which were then reviewed by employees whose core responsibility was customer support. None were fraud experts or even full-time fraud reviewers.

As the retailer’s online orders grew and as it instituted a new dropshipping program that introduced new vulnerabilities, the workload became unsustainable and those reviewing orders became more conservative in approving orders, so as not to risk shipping to fraudsters.

The retailer’s order approval rate slipped below 89 percent. The declined orders — a significant number falsely declined — were costing the company millions. Customer support workers who became familiar with fraud invariably moved on to other jobs and the problems worsened.

“As business grew, workload grew in all areas for customer support, and we saw an increase in online fraud and account takeovers,” the manager of ecommerce operations for the retailer said. “Items sold through our dropship initiative also were also not eligible for review with our legacy tool. We needed more automation.”

Enter Signifyd.

Signifyd’s fraud protection model relies on big data, machine learning and human domain expertise to shift fraud liability from merchants to Signifyd. The model attacks fraud on three fronts, all of which result in a better buyer experience for Signifyd’s customers’ customers.

First, it insulates merchants from chargebacks and related fraud costs on any approved order that later turns out to be fraudulent. Second, it significantly reduces the number of false declines, meaning customers aren’t frustrated and insulted by cancelled orders. Finally, it frees up employees who were doing double-duty as order reviewers to focus on providing the kind of superior customer service that leads to a memorable customer experience.

The retailer looked over several technology-based fraud protection solutions. It chose Signifyd because the customer support team liked the idea of an automated solution; the finance team liked the financial guarantee against fraud; and the accounting team liked the idea of shifting the cost of chargebacks away from the company.

Signifyd proved to be a wise investment. According to the Forrester study cited above, the percentage of orders manually reviewed once Signifyd was deployed dropped from 8 percent to 2 percent. And those reviews were conducted by Signifyd.

The decrease in manual reviews was critical to providing an exceptional customer experience and in assuring the retailer that it was capturing all the sales it should be. For instance, the retailer noted that consumers who received their orders on time gave the enterprise an 80 percent customer satisfaction score. Those whose orders were delayed, say by a cumbersome manual review, gave the retailer a 45 percent customer satisfaction score.

“Improving order fulfilment speed not only reduced risk to customer satisfaction, but also reduced the order cancellations, especially during the holiday season when everyone wants their order before Christmas,” the ecommerce operations manager said.

Turning to Signifyd not only paid dividends when it came to buyer experience. It also paid off in terms of the top and bottom lines.

When it comes to fraud charges, deploying Signifyd saved the retailer USD 2.7 million over three years. Even better, in the area of false positives, the retailer saved USD 3.2 million in orders that previously would have been declined for fear of fraud, but were shipped with Signifyd’s guarantee.

The retailer benefited to the tune of USD 732,000 during that same three year period from faster order fulfilment and reduced cancellations from customers who’d become frustrated waiting for a fraud decision on their orders. In fact, Forrester Consulting found that Signifyd contributed to saving 3 percent of the orders that previously would have been cancelled because drawn out fraud reviews.

Finally, the retailer was able to save USD 479,300 in costs attributed to paying for in-house fraud expertise and costs that were no longer necessary after Signifyd.

Quite a turnaround — from a retailer under siege to one that found a way to make customers happy and see a USD 7.1 million positive impact on its business over a three year period.

About Mike Cassidy

Mike is lead storyteller at Signifyd. A former journalist, he covers e-commerce and the way automation is changing digital commerce. Hes a retail geek. And, as a White Sox fan, hes also resilient.

 

 

About Signifyd

Signifyd, as the world’s largest provider of guaranteed fraud protection, delivers a 100% financial guarantee against chargebacks and fraud costs on every approved order. This shifts fraud liability and allows merchants to increase sales while providing a friction-free buying experience for their customers.


Free Headlines in your E-mail

Every day we send out a free e-mail with the most important headlines of the last 24 hours.

Subscribe now

Keywords: Mike Cassidy, Signifyd, machine learning, retailer, merchant, ecommerce, fraud protection, big data, fraud prevention, case study
Categories:
Companies:
Countries: World





Industry Events