Product Rule Filters Refactor
For Fraud.net clients that use E-commerce and Marketplace APIs, we have refactored how the product filters work within rules.
Previously
Product rule filters evaluated the entire product object across all items in a transaction, leading to potential false positives. For example:
Rule: Product Price > $125 AND Product Brand is Nike THEN send to queue
Products Purchased: $135 Converse shoes + $75 Nike shoes
Result: Transaction sent to queue, as one product met the price, and another met the brand condition, even though no single product matched both.
Currently
Filters evaluate each product individually, reducing false positives and aligning rule results with the intent.
Rule: Product Price > $125 AND Product Brand is Nike THEN send to queue (same as above)
Products Purchased: $135 Converse shoes + $75 Nike shoes (same as above)
Result: Rule does not trigger.
This update ensures rules now focus on relevant products, helping investigators target the intended transactions. Active rules with multiple product filters will now behave in the way described in the Currently example.