Ecommerce sales in the United States continue to gain a greater share of overall retail sales each year. The Department of Commerce reports that in 2019, total ecommerce sales increased almost 15 percent over 2018 and represented 11 percent of total retail sales. There is no question that with the current COVID-19 environment, our daily habits have undergone tremendous change. As part of that change, I expect that ecommerce sales will increase at a greater rate in 2020 than in 2019.
Following social isolation guidelines, consumers and businesses are turning more and more to conducting their commerce transactions online. Prepaid carry-out, drive-through, and delivery orders now dominate the dining industry as inside dining options have been largely shuttered. Large retailers have been promoting online ordering and ship-to-home delivery options as their stores are closed. TransUnion reports that in the week from March 11 to 17, when the World Health Organization classified COVID-19 as a global pandemic, ecommerce transaction volume increased 23 percent over the previous week.
This spike in ecommerce traffic will likely bring with it a parallel spike in criminal activity, possibly adding to the increasing fraud levels in ecommerce. This shouldn't come as any surprise. It will be important for the good guys not only to be expecting this but also to be prepared for it by making swift adjustments that match the challenge.
One of the key adjustments to consider and apply quickly is properly tuning algorithms for detecting ecommerce fraud. In normal times, anomalous-pattern detection schemes are relied on to expose fraudsters. Elements such as the type of stores commonly used, frequency of usage, average or range of transaction values, and more go into making up an overall usage pattern for a given customer. While these transaction risk models have become very sophisticated over the years, they are challenged by abrupt changes in usage patterns, especially at an individual account level. They need to be smartly and quickly adjusted. Issuers and merchants need to balance the decision of denying transactions—which brings with it the risk of disgruntled legitimate customers and lost revenues—against approving fraudulent transactions and taking financial losses. No easy task, but doable and necessary to undertake, with constant attention.
Working collaboratively with merchants, consumers can help to surprise the criminals as fraud fighting evolves. The good guys win if we exercise patience with one another and remain mindful of the balance between purchase friction and fraud avoidance as fraud-fighting tools and methods adjust. Both sides being considerate of the needs on both sides of the transaction—working together, again, with patience and willingness to engage, perhaps differently than we've been willing to in the past, could yield results that everyone (except the crooks) is happier with, in both the short run and long run.
We know fraud management teams will be busy managing their fraud-detection tools and processes and expect they will rise to the challenge. We also expect consumers are ready and willing to assist in ways that are helpful as well. The constant chess match with the criminal element will continue, and we look forward to seeing a chess piece on the good guys ' side of the board with some new moves to help aid in the fight against the bad guys.