A good fraud filter should act like a skilled bouncer. It keeps out the troublemakers, but lets the real customers through without hassle. The problem? Too many merchants end up locking out the wrong people. This guide shows you how to set up a fraud filter that actually works, without choking your revenue.
Why Fraud Filters Backfire
Fraud filters are built to stop suspicious behavior: mismatched billing details, bulk orders, high-risk geographies, and more. But when they're too aggressive or poorly tuned, they block valid customers. This leads to abandoned carts, lost lifetime value, and sometimes chargebacks anyway.
The goal isn't just to reduce fraud. It's to reduce fraud without rejecting good orders.
You need a balance.
Let's look at how to achieve it.
Start With Your Data
Before changing anything, look at your false positive rate. Pull reports from your fraud prevention tool and compare the following:
If your filter catches every fraudulent order but also turns away real buyers, it's costing more than it saves.
Common Causes of False Positives
These settings often trigger unnecessary blocks:
- IP/Billing mismatches: Many travelers or mobile users get flagged for this.
- High velocity: A returning customer placing two orders in one hour? That's not always a bot.
- VPNs and proxies: Common with privacy-conscious shoppers or international buyers.
- Unusual item count or size: Bulk orders aren't always resellers or fraud rings.
- First-time buyer with high cart value: New customers shouldn't be penalized for spending more.
Real-World Fraud Filter Settings That Work
Here's how to tune filters for Stripe Radar and Shopify without overblocking:
Stripe Radar
- Set dynamic rules. Instead of blocking if the IP doesn't match the card's country, try adding 20 to the risk score. This lets the system weigh context.
- Use Radar's custom rules for patterns you know are bad (like BINs from unsupported countries or repeat email addresses from blocked orders).
- Review orders with medium risk scores rather than blocking them. You can auto-decline high-risk only.
- Monitor decline codes from the bank. If the issuing bank says the transaction is clean, don't override it unless there's a red flag on your end too.
Shopify Fraud Protect
- Enable manual review for medium-risk orders.
- Check for device fingerprint overlaps between chargebacks and new customers.
- Whitelist your returning customers. High-value, repeat buyers often trigger filters if they shop from different IPs or cards.
Tips to Avoid Overblocking
- Always soft-block first. Flag an order for review before declining it outright.
- Review chargeback reason codes. If you're still getting disputes, your filter isn't catching the real fraud and might be blocking the wrong people.
- Watch for abandoned checkouts on filtered orders. If they spike, your filter might be too harsh.
- Run test orders from different locations, devices, and payment methods. See what gets flagged and fix it.
Fraud Filter ≠ Chargeback Shield
Remember, fraud filters catch suspicious orders before payment. They don't stop disputes that come later, especially friendly fraud or service-related chargebacks. You still need a plan for post-purchase chargeback prevention and evidence response.
Conclusion
A fraud filter should be smart, not strict. The goal is to catch bad behavior without punishing good customers. That means reviewing your settings, tuning them with real data, and avoiding one-size-fits-all rules. Use tools like Stripe Radar and Shopify's risk analysis to your advantage, but never set and forget.
FAQ: Setting Up a Fraud Filter
What is a fraud filter?
A fraud filter is a set of rules or machine learning models that automatically flag or block suspicious orders before they're processed. They're designed to prevent stolen card use, bot orders, and other payment fraud.
Can fraud filters block legitimate customers?
Yes. If a filter is too strict, it can block real buyers, especially first-time customers or people ordering from outside your typical location range. That's why tuning is so important.
How do I know if my fraud filter is too aggressive?
Look for high rates of flagged or declined orders that never result in chargebacks. You should also watch for spikes in abandoned carts or customer support complaints about failed checkouts.
What's the difference between a fraud filter and chargeback prevention?
Fraud filters operate before the transaction is processed. Chargeback prevention happens afterward, focusing on disputes and retrieval requests. You need both to protect your revenue.
Should I use automatic blocking rules?
Use them cautiously. Auto-blocking is best for patterns you know are 100% fraud. For uncertain cases, soft-flag the order or send it for review to avoid false positives.
Chargeblast Can Back You Up Where Filters Fall Short
Even the best fraud filter can't catch everything. Chargeblast helps you handle the chargebacks that slip through. We detect fraud that filters miss, build expert responses with the right evidence, and give you a second line of defense after payment. So you can keep selling confidently, without shutting the door on real buyers.
