Bottom Line: E-Commerce and online retail transactions & revenue continue to set new year-over-year records in 2020, creating one of the fastest-growing opportunities for bad actors and cybercriminals to defraud digital businesses of millions of dollars a year.
The swiftness that stay-at-home orders created new online buying habits is the crucible e-commerce and online retailers’ platforms needed to earn greater trust. ACI Worldwide Research announced this week that global e-Commerce retail sales had achieved 209% year-over-year revenue growth. Kount’s latest eCommerce Trends, Spring 2020 Report, further quantifies the high growth e-commerce and online retailing is experiencing today. Kount found that same day or next day shipping was up 237% in mid- March, and peaked again in late April, up 305% compared to the same week of 2019. Kount also found that transaction volumes in key product categories are driving a large percentage of the growth. Wellness and vitamin online transaction volumes increased 91% year over year in mid-March. By the third week of April, Sporting Goods online card-not-present transaction volume increased 599% year over year. Crafts and Wine transaction volume soared nearly 900% compared to last year.
In total, there’s been a 110% year-over-year increase pure e-commerce orders in the U.S. alone, according to the COVID-19 Commerce Insight dashboard. The following graphic from the COVID-19 Commerce Insight dashboard reflects e-commerce’s explosive growth since the beginning of the year:
COVID-19 Commerce Insight dashboard, from Emarsys in cooperation with GoodData
Fast-Growing E-Commerce Attracts Many Types Of Fraud
Just as e-commerce and online retailers are responding to the rapid growth in orders by finding new ways to scale shipping, delivery, and service, there needs to be just as strong of a focus on fraud detection and protection. The following are a few of the many approaches bad actors take to defraud e-commerce sites, online retailers, and consumers buying online:
- The most common type of e-commerce fraud is Chargebacks, one of the most expensive types of fraud an online retailer will experience. It’s a technique that often leads to additional fees, loss of inventory, services, and can even lead to them not being able to accept a specific type of credit card. Chargebacks can be the result of either criminal or friendly fraud. In either case, chargebacks are costly and take an excessive amount of time and cost to resolve.
- Friendly fraud escalates the more online revenue an e-commerce site or online retailer generates and can account for between 40% to 80% of all fraud losses. Also known as First Party Fraud, this type of fraud occurs when customers request refunds from their issuing banks, claiming that transactions on credit card statements are fraudulent. Rather than contacting the business where a purchase was made, a consumer simply calls their bank, which, lacking verification information, refunds the consumer and issues a chargeback to the company. Friendly fraud can be either accidental or intentional.
- Gaming and wireless fraud are one of the fastest-growing areas today, given the 126% year-over-year growth of online sales in this area as of April 2020. Credential stuffing, Account Takeover, and Streaming Potluck schemes are all used to defraud e-commerce sites and online retailers today. Credential stuffing is the use of mass login attempts to verify the validity of stolen username/password pairs. This information is used to take over accounts. Streaming Potluck is a technique where each member of a group brings their account for a streaming service and shares it. Each member of the group has access to all services. Depending on the service provider’s business model, unauthorized account sharing can cause lost revenue.
- Account Takeover (ATO) is rampant in gaming and spreading across every area of e-commerce today. The online gaming industry relies on real-time approvals of small transaction amounts to keep players engaged. Transactions need to be approved nearly instantaneously and at scale for a game to be financially successful. Fortunately, there are fraud prevention solutions that can analyze a transaction in less than a second, taking into account the many data points from other current and past transactions to see if it is real or fraudulent. From 2016 to 2017, losses from account takeover (ATO) rose 122% and reached $9B in 2019.
Thwarting Fraud Using A Real-Time Identity Trust Network
E-commerce sites and online retailers are facing an overwhelming workload of transactions, unusual activity, and fraud signals their fraud analysts are doing their best to keep up with today. Many are relying on systems that generate false positives, rejecting loyal customers who are legitimate buyers. That’s one of the most common ways an online retailer will lose a customer.
False positives and slow response times often happen when the transaction is first assumed to be fraudulent. Instead of forever trying to fine-tune a risk score, quantifying trust with long-standing customers is showing the potential for improving fraud detection online. Kount is taking this approach by fine-tuning trust scores based on their Real-Time Identity Trust Network. The Trust Network is based on supervised and unsupervised machine learning algorithms that deliver personalized responses to transaction requests in milliseconds. Using supervised machine learning algorithms and techniques, Kount can learn from every customer interaction, gaining new insights into how to fine-tune identity trust for every customer’s transaction. The following graphic explains their approach for bringing identity trust into the center of their platform:
By prioritizing trust over risk scores, e-commerce and online retailers can reduce false positives, friction, and lost sales. Apart from strengthening fraud prediction and prevention, improving user experiences is by far the most important area of any e-commerce strategy today. Having a trust score that continually learns about each customer to reduce transaction delays further and improve accuracy using machine learning is proving effective in thwarting fraud. Trust is the best revenue accelerator there is, and now is the time to enhance e-commerce and online retailing with a more precise approach to completing sales.
Quantum Metric COVID-19 Online Sales Impact Data, March 11, 2020