5 Strategies to Fight Chargebacks
According to comScore, U.S. online sales – both desktop and mobile eCommerce – grew by double digits in 2016. Unfortunately, card-not-present fraud grew just as fast. In fact, 2016 saw the highest overall amount of eCommerce fraud ever. And 2017 is expected to see even higher eCommerce fraud as EMV and other forces shift fraud to the online realm.
More bad news? Losses from fraud negatively impact profitability in a dramatic way. For example, a large online retailer with a not-unusual 0.08% chargeback rate would suffer $2.2 million in annual product losses and chargebacks (e.g., processing 15,000 orders per day at an average of $40 per order).
Conversely, investments in enhancing fraud prevention can positively impact profitability, again, in a dramatic way. By reducing the chargeback rate to 0.04%, that same large online retailer could add more than $1 million to its bottom line.
But what investments in improving fraud prevention deliver the best return-on-investment? The first thing to understand is that fighting chargebacks isn’t a monthly event that takes place at a single point in time. There are multiple prevention and interception “catch points” where you can take action to reduce chargebacks.
Here are five prevention and interception strategies that can help you win the fight against chargebacks at every stage of the chargeback cycle:
- Deploy Superior Fraud Screening and Analytics. Use multiple technologies that screen multiple dimensions of every transaction to reduce fraud, reduce the number of manual reviews and reduce “false positives” (good transactions that only look suspicious). At a minimum, you should be integrating Device ID, Geolocation, Transaction Velocity, and Order Linking to maximize accuracy and precision. However, the more screening data you can collect, the more you can take advantage of the Big Data network effect (see 3. Apply Advanced AI Machine Learning). Kount employs dozens of proprietary and patented fraud screening technologies—integrated at the code level—to generate hundreds of data elements per transaction across billions of transactions from over 180 countries...all analyzed in milliseconds.
- Orchestrate Real-time Data. When transaction data by itself is insufficient to render a clear decision, outside data sources can fill in gaps to provide crucial context. The more data you can bring to the fight, the better your chances to precisely differentiate fraud orders from merely suspicious (but valid) transactions. In addition, your systems should orchestrate data from these third parties in real-time so decisions are based on your risk criteria and give proper weight to factors based on your specific tolerance for risk. Kount orchestrates and integrates multiple best-in-class third-party data sources – in real time – to deliver optimal context for detecting and preventing fraud...but only on an “as-needed” basis to minimize data costs.
- Apply Advanced AI Machine Learning. AI machine learning has the massive processing and memory capacity to spot patterns in Big Data that are undetectable to humans. It can extract highly predictive features in real time to enhance fraud detection in low-information scenarios (e.g., first-time fraud).
This “online” machine learning should be enhanced with offline learning that provides non-stop evaluation and refinement of AI results by human fraud experts to ensure machine learning algorithms are continuously optimized. Kount’s advanced machine learning uses patented graph theory algorithms (PERSONA™) to detect and respond in real-time to emerging threats...but according to rules you set to maximize revenue while holding down fraud.
- Exploit Expert Human Intelligence. Fraud prevention systems that deliver optimal ROI are informed by best practices developed by fraud prevention experts. This expertise is enhanced by the ability to deploy rules set up by thinking humans that enable a strategic and customized response to rapidly-changing fraud tactics. Kount fraud experts provide guidance and insight to each merchant, helping them get better at fraud mitigation. Kount’s Rules Engine then makes it easy for online retailers to fine tune rules and thresholds so that detection and prevention responses are customized to each customer’s specific appetite for risk.
- Capitalize on Chargeback Alerts. Intercept “bad” transactions even after they’ve been approved, so you can avoid product losses and chargeback fees. How? It typically takes 45 days from the time a chargeback is actually incurred until a merchant is notified. Yet card issuers often know within hours about a chargeback. It’s the lengthy reporting process that introduces the 45-day “Chargeback Lag.” Innovative, cooperative networks of merchants and card issuers—like the one founded by Ethoca—avoid this “Chargeback Lag.” With Ethoca Alerts, card issuers send electronic alerts in as little as 1 hour so your online eCommerce operations can intercept bad orders before they become chargebacks. This helps you avoid chargeback fees and merchandise losses, identify other fraudulent transactions linked to the original bad order, and approve more borderline transactions for higher sales while holding the line on chargeback losses. Kount integrates Ethoca alerts and reporting into its fraud prevention process to automate greater efficiency. Admin time is typically reduced as much as 75% and automation of the process for linking other fraudulent transactions to the bad order(s) in question improves fraud prevention.
If you’ve been losing more and more time and money to chargebacks, follow these five winning strategies and start winning the fight against product losses and chargebacks. Want to know more? Attend the “How to Win at Every Stage of the Chargeback Cycle” webinar at 10:00 am PST / 1:00 pm EST on Wednesday, March 22.