Effectively combating fraud is not a one-size-fits-all effort. It’s critical for trust and safety solutions to be both flexible and powerful so that the teams using them can stay accurate and agile.
Sift’s latest product release puts transparency and control front and center, giving fraud teams additional levers to pull and more insights to indicate when and how to pull them when dialing in fraud-fighting operations.
Validate changes prior to publishing with Workflow Backtesting
Improving (or simply updating) existing workflows is now a whole lot simpler. Workflow Backtesting pulls in detailed metrics that can be compared against the existing workflow prior to going live, reducing errors and allowing analysts to preview the impact of any new changes. Fraud teams are free to iterate away and continuously discover new ways to stop more fraud, all while creating better customer experiences, and without worrying about publishing an errant update.
Workflow Backtesting is available for all active workflows.
Maintain specific, stable rates of friction with Percentile Scoring
Many fraud teams closely monitor customer friction rates (e.g., blocks, manual reviews, 2FAs), but keeping those rates stable can be tricky using a raw risk score alone.
Percentile Scoring allows teams to set a fixed friction rate using a percent (i.e. the top 2% of all scores) rather than using a raw score in isolation (i.e. all scores above 85), making it that much easier to keep customer insult rates down and maintain steady workloads for review teams.
Percentile Scoring can be used with Payment Protection, Account Defense, and Content Integrity for all events available via Events API and Score API. It can be used in Workflows to automate block and approve decisions, to apply adaptive friction (authentications, notifications), and to populate manual review queues.
