How Riskified differs from a typical fraud check
Most merchants run their own internal fraud rules and absorb chargeback losses themselves. Riskified inverts the model: the merchant ships every order to Riskified for a decision, and Riskified pays the chargeback if any approved order turns out to be fraud. This creates a strong incentive for Riskified to be aggressive on borderline cases — false approvals cost it money directly.
Practically, that means Riskified-protected merchants decline orders at the slightest pattern anomaly: a residential IP in one state with a billing address in another, a freshly-created account, a payment instrument with no Riskified-network history. Real customers get caught by this regularly, which is why "my order was declined" customer-support is so prevalent at Riskified-merchant sites.
When scrapers encounter Riskified
Same answer as Forter — pure data extraction is unaffected, anything automating checkout is in scope. Common scenarios:
- Sneaker / streetwear limited-drop automation
- Concert and sports ticket purchasing
- Limited electronics releases (consoles, GPUs)
- Luxury goods resale flipping
The failure looks the same as Forter — silent decline at the payment step with a "please try again" message.
What works against Riskified
Operationally, not technically. Aged accounts with organic activity (orders, returns, browsing patterns) score higher than fresh accounts no matter how clean the fingerprint. Billing and shipping addresses must match the IP geolocation. Payment instruments with prior clean history at any Riskified-network merchant carry over. The single biggest signal Riskified uses is the network effect — a card or device that declined at one Riskified merchant gets a worse score at every other Riskified merchant, so burning identities is permanent.