Improved fraud prevention with Radar 2.0

We launched Radar in 2016 to help protect our users from fraud. We’ve blocked billions of dollars in fraud across the Stripe network for companies of all sizes—from startups like Slice and WeSwap to larger companies like Fitbit and OpenTable. Since launch, we’ve continuously invested in our suite of fraud prevention tools, and today, we’re excited to launch the result of those efforts.

The next generation of machine learning

As of today, we’ve rebuilt almost every component of our fraud detection stack to dramatically improve performance. In early testing, the upgraded machine learning models helped reduce fraud by over 25% compared to previous models, without increasing the false positive rate.

  • Hundreds of new signals for improved accuracy: We’ve added new signals to better distinguish fraudsters from legitimate customers, including certain data from buyer patterns that are highly predictive of fraud. Some signals now use new, high-throughput data infrastructure to process hundreds of billions of historical events. (Even if a card is new to your business, there’s an 89% chance it’s been seen before on the Stripe network.)
  • Nightly model training: Fraud evolves and changes rapidly. Radar can now adapt even faster by training and evaluating new machine learning models daily.
  • Algorithmic changes for better recall and precision: We’ve optimized our machine learning algorithms in hundreds of ways—from boosting the performance of our decision trees to tweaking the minutiae of how we handle class imbalance, missing values, and more.
  • Custom models for your business to maximize performance: Radar is constantly evaluating how to balance patterns from across the Stripe network with patterns that are unique to your business. Radar now trains and evaluates multiple models daily and determines which one achieves the best performance for you.

Radar cut our fraud rates by over 70% without any configuration, saving our pizzerias thousands of dollars every month and allowing us to focus on delivering the best local pizza experience possible.

Finn Borge, Product Manager at Slice

Introducing Radar for Fraud Teams

While most Stripe businesses can rely entirely on Radar’s automated fraud protection, it makes sense for some companies with more fraud risk to invest more deeply. We’ve now made that as easy and powerful as possible on Stripe. We’ve honed features to be even more useful to fraud professionals, built new features, and packaged them all together in a new bundle called Radar for Fraud Teams.

Optimized reviews to spot fraud faster Update

When reviewing payments, we now show relevant info for faster and more accurate reviews. You can see data related to the device used, compare the geolocated IP address and the credit card address, or see whether the purchase pattern is anomalous compared to typical legitimate payments for your business.

Related payments for more accurate reviews Update

We now help you evaluate payments holistically rather than in isolation by surfacing previous related payments your business has processed that match certain attributes like email address, IP address, or card number.

Custom rules with real-time feedback Update

When you create a rule, Stripe will use your historical data to show how that rule would have impacted real transactions your business has seen. We’ve added dozens of new properties you can use in rules to give your teams even more fine-grained levers.

Custom risk thresholds New

Radar for Fraud Teams surfaces a numerical risk score (0–100) for every payment. Depending on your business’s appetite for fraud, you can tweak the threshold at which to block payments to maximize revenue.

Block and allow lists New

Fraud teams now have an easy way to create and maintain lists of attributes—card numbers, emails, IP addresses, and more—that you want to consistently block or allow.

Rich analytics on fraud performance New

Get a snapshot that helps focus your fraud team. The new overview highlights dispute trends, the effectiveness of reviewing flagged payments, and the impact of rules you’ve written for your business.

Radar for Fraud Teams has already made fraud management easier and more effective for beta users at Watsi, Fitbit, Restocks, Patreon, and more.

The related payments feature helped our fraud team quickly spot a nuanced fraud ring and avoid significant potential loss. It’s been a great asset in our fraud detection arsenal.

Alison Cleggett, Head of Risk and Compliance at WeSwap

Get early access

We’re gradually rolling out the upgraded machine learning models to all users over the next few weeks. If you’d like, you can also activate the models today. Activating early requires that your Stripe integration follows a few basic best practices. Most users already follow these best practices and won’t need to make any changes to get early access—you can check your integration by logging in to the Dashboard.

Radar for Fraud Teams is also available starting today as an optional add-on. If you’re already using any of features included in Radar for Fraud Teams (like rules or reviews), there are no changes to your pricing—the machine learning updates and all new Radar for Fraud Teams features are included at no extra cost for your account.

Like this post? Join our team.

Stripe builds financial tools and economic infrastructure for the internet.

Have any feedback or questions?

We’d love to hear from you.