Allow and deny lists
Manual overrides on checkout — trust a buyer past every risk action, or refuse one outright, keyed on email, domain, phone, IP, or device.
Risk scoring is automatic; lists are your manual word on top of it. An allow list waves trusted buyers straight through; a deny list refuses known-bad ones before an order even forms. Both live at Protection → Lists.
What you can match on
Each entry matches one identity facet of a public checkout:
| Match on | Example |
|---|---|
[email protected] | |
| Email domain | bulk-resale.com |
| Phone | +420777123456 |
| IP address | a specific address |
| Device | a device fingerprint |
Add an optional note — why this entry exists — which only you ever see.
How the two lists behave
- Allow list — "Trusted buyers who bypass every risk action." Their orders are still scored (your metrics stay honest), but never challenged, held, or blocked. Use it for your own team testing checkout, a corporate buyer placing a big legitimate order, a VIP whose VPN keeps tripping the IP signal.
- Deny list — "Checkouts matching any of these are refused outright — before an order is created." No order, no seat hold, nothing. For buyers you've already caught reselling.
Deny wins ties. If a buyer matches both lists — allow-listed email, deny-listed IP — they are refused. The safe default: a single deny signal outranks any amount of trust.
When to reach for lists vs rules
- Use rules for the general population — they scale automatically with the risk score.
- Use lists for the specific exceptions rules can't know: this person is fine, that domain is a reseller. Lists are surgical; rules are policy.
Entries take effect on the next checkout — removing one stops it applying immediately — and every add/remove is audited.