Manual reviews for refund abuse help differentiate customer behaviours specific to refund-related interactions. Unlike payment fraud labels (e.g., genuine or fraudster), refund abuse reviews focus on higlighting behaviors related to the misuse of refund policies.
Manual reviews provide essential labels of high-risk behaviors that can help Ravelin’s systems identify data patterns. These reviews enable continuous improvement of rule accuracy and may support model optimisation in the future, refining predictions for refund abuse detection.
There are two main reasons why doing manual reviews can be beneficial.
Manual reviews can act as a proactive defense against refund abuse, enabling the use of ravelin’s rule engine to prevent abuse as it happens.
Precise manual reviews can be used to refine the model, enhancing its ability to distinguish between genuine customers and potential refund abusers, ultimately improving predictive accuracy.
If you would like to do reviews with the goal of optimising your machine learning model, you should review customers that you are very confident are either people prevented but who are obviously genuine (false positives) or people allowed who are obviously fraudulent (false negatives).
Reviewing a customer as refund abuser or not refund abuser currently has no direct impact on future orders placed or refunds requested by that customer. To influence future actions based on these reviews, you can set up a rule in the Ravelin dashboard at the relevant checkpoint. For example, you might prevent known refund abusers from making purchases at checkout checkpoint or approve refunds for trusted customers at the refund checkpoint.
In order to use reviews to catch more fraud in real time, you could focus on reviewing customers that are newly registered. Customers creating a lot of accounts tend to be in the same location so manual review on a new account can help to stop other fake accounts that share that location from making orders.
If you do review a customer, please leave comments with your customer manual reviews. These are invaluable to Ravelin, to other people at your company and potentially to you at a later date. Explain why you reviewed this customer and any additional evidence you might have found.
Customers you are not sure about. If you are not sure how to review someone, then we suggest that you don’t review them at all.
You can tag customers, add them to watchlists or leave a comment if you’d like to keep an eye on suspicious accounts. Whilst manual reviews can be beneficial, incorrect reviews can potentially damage the accuracy of recommendations.
A customer may have different labels in payment fraud and refund abuse contexts because these fraud types are driven by distinct behaviors. For example, someone marked as “genuine” in payment fraud could still abuse policies, while a “fraudster” in payment fraud might follow refund policies.
Key Tip: Treat each label independently. Set context-specific rules and avoid assuming that payment fraud labels directly indicate refund behaviors—or vice versa.
Test your refund abuse integration
Was this page helpful?