Can We Build Useful Fraud Intelligence from Near-Misses?

Most fraud data comes from losses: someone was harmed, and a report was filed. But for every successful scam there are many attempts that didn't land.

Near-misses are rarely captured in a structured way, yet they contain the same scripts, lures, and infrastructure as the attacks that succeed, often earlier. What would it take to collect them, and what could we learn if we did?

Short note. More to come.