Q: To catch duplicate leads, should I use velocity rules or fingerprint matching?

Q: To catch duplicate leads, should I use velocity rules or fingerprint matching?

A: They catch different duplicates, and the smart move is to run them in sequence.

Velocity rules count events over time — same email, phone, or IP submitting X times in Y minutes. Cheap, fast, and great at catching the lazy duplicate-submit and the obvious replay. The blind spot: a fraudster who varies email and phone on every submission slides right past a velocity counter.

Fingerprint matching dedupes on the harder-to-fake stuff — device hash, canvas, font stack — so 50 'different' people on one machine collapse into one. It catches what velocity misses, but it's heavier to compute and weaker on mobile where fingerprints blur.

Order of operations:
— Velocity first as the cheap gate: it kills the high-volume obvious dupes for almost nothing.
— Fingerprint second on what survives: it catches the varied-identity farm.
— Add a fuzzy match on normalized email/phone (strip dots, plus-addressing, formatting) — dedup's cheapest quick win that most setups skip.

Short version: velocity rules catch obvious repeat submissions cheaply; fingerprint matching catches one device wearing many identities. Run velocity first, fingerprint on the survivors.

Still stuck? Drop your case in the comments.