{"url":"https://www.clear.sale/blog/fraud-screening-optimization-for-more-ecommerce-transaction-approvals-fewer-cancellations","title":"Transaction Screening Optimization: The Perpetual Balancing Act of Fraud Risk, Customer Behavior and Consumer Expectations","tldr":"Slow approvals make 70% of shoppers walk away. Learn how to optimize fraud screening for more ecommerce approvals and fewer order cancellations.","markdown":"TL;DR\n\nFraud prevention usually balances automation and expert review to limit fraud and false positives, but a third variable often gets overlooked: customers who cancel orders when approvals are slow. With 70% of consumers unwilling to buy from companies with long wait times, manual review must be both accurate and fast. Merchants can optimize by monitoring auto-approval thresholds, feeding every order's outcome back into machine learning, staffing enough analysts, and tracking cancellation KPIs alongside fraud and false decline rates.\n\nEcommerce fraud prevention typically focuses on finding the right balance of automation and expert review to minimize both fraud and false positives. However, there’s another variable that’s sometimes overlooked as merchants and fraud-prevention providers set up their anti-fraud programs: the risk that impatient or confused customers will cancel their orders before they’re approved.\n\nWith the ecommerce sector more crowded with options for shoppers than before the pandemic—and with customer expectations for excellent, immediate service higher than ever—merchants can benefit from optimizing their fraud control processes to minimize order cancellations as well as fraud and false positives.\n\nFraud, false positives and customer cancellation considerations Of the three issues we’re discussing, fraud is the one that merchants focus on the most, and with good reason. Fraud losses increase every year, and in 2021 each dollar of fraud costs North American retail and ecommerce merchants $3.60, compared to $3.13 prepandemic, according to LexisNexis data.\n\nMerchants who understand the short- and long-term risks of false positives work hard to minimize them. That’s because when a good order is rejected, the profit on that order is lost, and the customer relationship is often lost as well. [ClearSale’s State of Consumer Attitudes](/blog/online-fraud-and-consumers), Fraud & CX 2021 Survey of online shoppers in the U.S., Canada, Mexico, Australia and the U.K. found that after an order is declined, 40% say they won’t shop again with that merchant and 34% will post negative social media comments about the merchant. False positives can cause lost customer lifetime value and brand damage that can increase the cost to acquire new customers.\n\nCustomer cancellations can happen for just about any reason, including finding the same item at a lower cost or simply changing one’s mind. However, slow order approvals can also prompt customers to cancel the order and buy it elsewhere, instead of waiting to see if their order will ultimately go through with the first merchant. This is a bad customer experience, which creates the risk that the customer will never return. It also means the merchant loses their profit on the order as well as the cost of fraud screening for it.\n\nBalancing automated order approval and manual review Automatic order approvals eliminate the risk of customer cancellations caused by slow approvals. With the right rules and resources in place, automatic approvals can function without unacceptably increasing the merchant’s risk of fraud. They’re also inexpensive, at pennies per transaction.\n\nIt may seem logical, then, that automated order rejections would help merchants streamline their order process and save on fraud control, but automatic rejections raise the risk of false declines. In our customer attitudes survey, 25% of online shoppers said they experienced at least one decline, with 49% reporting more declines in 2020 than in 2019.\n\nThe solution here is to send suspicious orders to a manual review team for investigation and approval or rejection. This costs a few dollars per order, but that cost is small compared to the potential customer value losses and other costs of a false decline. The risk in terms of CX here is the time it takes to manually review the order. Seventy percent of consumers say they won’t buy from companies with long wait times, per a global Salesforce study, so manual review must be both accurate and fast.\n\nOptimizing fraud control for maximum revenue and minimal loss. A few key actions can help you ensure that your fraud control processes are delivering the best possible outcomes in terms of fraud reduction, false decline prevention and cancellation prevention.\n\nReview and monitor your automated approvals to ensure that your threshold is right for current conditions. For example, some merchants adjust their automatic approval cutoff point during sales peaks based on revenue versus loss calculations for sales during those periods.\n\nIncorporate machine learning (ML) into your entire fraud control process. By screening every order and feeding the results back into your anti-fraud algorithm, you can improve your ML’s ability to identify good orders as well as possible fraud. Over time, this can reduce the volume of orders that require manual review to be safely approved.\n\nMake sure you have enough fraud analysts available, in-house or through a provider, to quickly review flagged orders with minimal delays. Analyst availability is especially important during sales peaks, when fraud control can become a bottleneck in the order approval process and when customers are especially sensitive to delays in completing their purchases.\n\nTrack your store’s order cancellation KPIs as well as fraud and false decline KPIs. As you adjust elements of your fraud control program, such as adding more analysts for manual review or moving your automatic approval cutoff point, take note of the impact on order cancellations and fine-tune those adjustments as needed.\n\nManaging all of these variables can be a challenge, especially as fraud risks, customer behavior and consumer expectations keep changing. Implementing a plan to monitor and update your fraud controls to prevent chargebacks, false declines and order cancellations can reduce fraud losses, customer churn and revenue and resources lost to cancellations—all while giving customers the ecommerce experience that they expect now.\n\nOriginal article at: [https://www.paymentsjournal.com/transaction-screening-optimization-the-perpetual-balancing-act-of-fraud-risk-customer-behavior-and-consumer-expectations/](https://www.paymentsjournal.com/transaction-screening-optimization-the-perpetual-balancing-act-of-fraud-risk-customer-behavior-and-consumer-expectations/)\n\n## Frequently Asked Questions\n\n### What third factor is often overlooked in fraud screening?\n\nBeyond fraud and false positives, merchants often overlook order cancellations caused by impatient or confused customers who abandon a purchase before it is approved. As the e-commerce sector grows more crowded and service expectations rise, optimizing to minimize cancellations matters alongside fraud and false positives.\n\n### How much did e-commerce fraud cost per dollar in 2021?\n\nAccording to LexisNexis data, each dollar of fraud cost North American retail and e-commerce merchants $3.60 in 2021, up from $3.13 before the pandemic. Rising fraud losses are why merchants focus heavily on fraud, though false positives and cancellations also carry real costs.\n\n### Why are automatic order rejections risky?\n\nAutomatic rejections quickly deny suspected fraud but raise the risk of false declines, because many good customers behave in ways that resemble fraud. In ClearSale's customer attitudes survey, 25% of shoppers reported at least one decline, with 49% reporting more declines in 2020 than in 2019.\n\n### How fast does manual review need to be?\n\nManual review must be accurate and fast because 70% of consumers say they will not buy from companies with long wait times, per a global Salesforce study. Slow approvals push customers to cancel and buy elsewhere, costing the merchant the sale and the screening expense.\n\n### How can machine learning improve fraud control?\n\nBy screening every order and feeding the results back into the anti-fraud algorithm, machine learning gets better at identifying both good orders and possible fraud. Over time this can reduce the share of orders that need manual review to be safely approved.\n\n### Which KPIs should merchants track to optimize screening?\n\nMerchants should track order cancellation KPIs alongside fraud and false decline rates. As they adjust controls, such as adding analysts or moving the auto-approval cutoff, they can watch the impact on cancellations and fine-tune the settings."}