Multi-layer Security Fraud Mitigation Lowers the Cost of Detecting Fraud
November 01, 2016
Accuracy is the primary means to lower the cost of fraud detection. When an automated solution is less accurate, the frequency of manual reviews rises along with associated costs of reacting to false positives and other bad triggers. When the automated solution brings more accurate results, the need for manual reviews drops. The odds of accuracy grow with a balanced, multi-layered approach to automated controls. Related benefits include lower costs by less intervention of fraud professionals and less friction in the buying process that might irritate a good customer. Some data points in this year’s LexisNexis True Cost of Fraud study show the importance of this principle and how to achieve its benefits.
1. Fraud is Growing So Merchants Need Help
As card-present fraud declines with the rollout of EMV, card-not-present fraud is growing for e-commerce and m-commerce merchants. These merchants are discovering that detection of remote channel fraud is a mandatory cost of reducing losses. The study revealed merchants’ average volume and value of fraudulent transactions rose sharply during the past year –fraud grew from 1.32% to 1.47% as a percentage of revenues (see Fig. 3 in the study). In conjunction, efforts at verifying customer identity are not going well. The average percent of flagged transactions sent for manual review by online and mobile merchants was 51% and 58% respectively, compared to the 42% average of all merchants (Fig. 11). These data suggest that remote channel merchants do not thoroughly trust their automated fraud detection tool or tools to make transaction decisions without human backup. Larger remote channel merchants also complain the cost of controlling fraud is too high (Fig. 12), which is clearly linked to the higher rate of costly manual reviews. Resolving this dichotomy entails boosting merchants’ confidence in the accuracy of their automated tools. Earning this trust requires demonstrable evidence of accuracy such as cutting the rate of false positives and blocking more attempts at fraud.
2. Achieving More Accuracy with a Multi-layer Approach
Automation of fraud detection entails the use of one or more solutions. The use of multiple tools or solutions is called a multi-layered approach to controlling fraud. The layers begin with the standard core solutions, such as card verification or PIN/signature; advanced identity solutions, such as geolocation and device ID fingerprinting; and risk transaction assessment solutions, such as rules-based filters and machine learning. Some merchants try to get by with a single so-called “silver bullet” solution. According to Fig 22, their results aren’t so good. As a percentage of declined transactions due to false positives, 33% to 34% were experienced by merchants with a few simple, core solutions. Merchants using 3-4 solutions as limited layering had slightly better results at 29%. Merchants with five or more multilayer solutions experienced just 21% – a reduction of about 38% from core solutions alone. Using a multilayered approach for detecting fraud decreases false positives; the accuracy gives you confidence to rely more on automated solutions.
3. Multi-layer Controls Bring Better Results
Reducing false positives is a vital first step toward “better results” – the indicators executives care about such as blocking fraud attempts and lowering cost of operations. The latter occurs because the automated solution automatically steers a smaller quantity of higher risk transactions into manual review. Less manual intervention requires a smaller staff of fraud analysts, which lowers the cost of operations. And the end result, shown in Fig. 23, is that the percent of successful fraud attempts significantly drops when large remote merchants use a multilayered approach. The e-commerce and m-commerce merchants respectively experienced a prevention rate of 40% and 33% per month without multilayer; they prevented 68% of fraud attempts with a multilayer fraud solution. An added benefit is that more good transactions are approved in a swift, timely manner – resulting in a smooth customer purchasing experience.
The key takeaway from these numbers is that a multi-layer fraud solution ultimately costs less and performs better than a handful of simple controls. Relying on a single “silver bullet” solution is also risky. We agree with the study’s recommendation to use a multilayered approach for fraud prevention. This approach will help your organization to improve accuracy and effectiveness of fraud control. Multi-layer is the foundation of blocking fraud before it occurs, and doing it in a more cost effective manner.
About the Author
Jeff Wixted is the VP of just about anything at Accertify. These days he oversees our Product Management, Decision Sciences, Implementation and Support teams at Accertify. He also represents Accertify as a member of the Merchant Risk Council's global advisory board.