How to catch scammers using slightly changed names, emails and addresses
Look-alike fraud is when someone you already blocked comes back with a tiny change: jane.doe1@ instead of jane.doe@, Jon Smyth instead of John Smith, or 123 Main Stret instead of 123 Main Street. Exact-match blocklists miss these, so the same person keeps ordering.
Warning signs
- A new order whose email, name or address is one or two characters away from a known fraudster
- Plus-addressing or dots in an email (jane+1@, j.ane@) on an otherwise familiar customer
- Misspelled street names or swapped unit numbers on an address you've seen before
How to prevent it on Shopify
Things any merchant can do, with or without an app.
- Block whole email domains used only by fraudsters, not just single addresses
- When you block someone, block their name, email and address together
- Search your orders for partial matches (the street number and name) before shipping high-value orders
AI-powered fraud detection
What Interceptor does about it
- Compares every new order's email, name and address with your blocklist and the orders you marked fraud, including near misses
- Sends near misses to an AI look-alike check that decides whether it's the same person behind a small change, and shows both values side by side with the differences highlighted
- Each AI check sees only the two values being compared, never payment details or other customer data
What it doesn't do: The AI judges look-alikes of people you've already blocked or marked fraud; it doesn't guess about strangers. AI checks are capped per store each month.
Questions
Why doesn't my blocklist catch repeat scammers?
Most blocklists need an exact match, so changing one letter gets through. Interceptor also checks near misses and asks an AI whether they're the same person.
Does the AI see my customers' data?
Only the two values being compared, for example two email addresses. No order totals, payment details or anything else, and the AI provider doesn't train on it.