Also called: Fraud screening
What a fraud filter is
A fraud filter applies screening rules to transaction information before or during checkout. The rules look for indicators such as a mismatched billing address, a card security code that does not match, unusual device or location details, or repeated attempts using the same card.
Depending on the system, a flagged order passes, goes on hold for review, or is blocked. A filter focuses on applying specific rules to transactions. Fraud detection is the broader process of identifying suspicious activity, and filters are one tool within it.
A filter that blocks more fraud but also rejects good customers can cost you sales you never see.
How fraud filters work
You or your payment provider set the rules and their actions. An address-verification check compares billing address details with issuer records. A velocity check limits repeated attempts within a set period. Other rules can use the customer's IP address, shipping details, product codes, or purchase patterns.
Some systems let you run rules in a passive or test mode. The filter returns results without acting on them, so you can see which orders would be held or blocked before applying the rules to live sales. In an active mode, a matching rule triggers its configured action.
How filters affect sales and costs
A blocked fraudulent order can prevent a loss and a later chargeback. A false decline blocks a legitimate customer instead, costing you the sale and possibly a future purchase. Strict rules that reject every order with a billing and shipping address mismatch can block valid gift orders or customers who recently moved.
Fraud filters do not change interchange or card-network assessments. Their direct price depends on the product: pricing models include monthly fees, volume tiers, or charges tied to matched transactions. Manual review also takes staff time, so compare filter costs and review workload with the fraud losses and legitimate orders you prevent or reject.
What to check before changing rules
Review both fraud caught and legitimate orders blocked. For each rule, compare its results with completed sales, later disputes, and manual-review outcomes. If a country block or address rule produces many legitimate declines, narrow the rule instead of rejecting every order that matches it.
Test proposed changes in passive mode when your system supports it. A saved setting can affect live authorizations and the response your checkout software receives, so confirm that your application handles pass, hold, and block results correctly.
What you can actually change
Your filter settings control which orders receive extra scrutiny and which are stopped. Tune each rule against actual fraud and false-decline results, rather than treating every warning as proof of fraud.
- Start with address, card security code, device, and velocity checks that match your sales channel.
- Use a hold or manual review for uncertain orders instead of an automatic block.
- Track legitimate orders blocked alongside fraud caught, disputes, and review workload.
- Test rule changes in passive mode before activating them on live transactions.
- Measure false declines. Compare blocked orders with customer records and completed sales to see which rules reject legitimate buyers.
- Test before activating. Use passive results to understand how a rule affects orders before it changes live checkout decisions.
- Price the full process. Include software charges and staff time for held orders when weighing the cost of fraud screening.
Frequently asked questions
What does a fraud filter do?
A fraud filter checks transaction details against rules such as address mismatches, card security code results, device information, or repeated attempts. Based on the settings, an order passes, is held for review, or is blocked.
Is a fraud filter the same as fraud detection?
No. Fraud detection describes the broader work of identifying suspicious activity. A fraud filter is a specific tool that applies configured transaction rules and returns or takes an action, such as holding or blocking an order.
Can a fraud filter block legitimate customers?
Yes. A rule can reject a legitimate order when its details match a risk condition, such as a billing and shipping address mismatch. Review blocked orders and adjust rules that create unnecessary false declines.