WP Simple Pay Blog

WordPress Tutorials, Tips, and Resources to Help Grow Your Business

How Stripe Radar Protects You From Fraud (and How to Use It)



Last updated on

Online fraud is a bigger problem than ever, but Stripe Radar is here to address it.

In 2017, there was $57.8 billion in ecommerce fraud losses, according to the October 2017 Global Fraud Index. That’s a 30% increase from the year before.  

Furthermore, nearly 12% of fraudulent transactions exceed $500. There are plenty of thieves who don’t mind stealing big.

The most common type of fraud is identity theft, accounting for 92% of all fraud. 7% of fraud occurs through direct billing, third party transfers, and gift cards.

As the ecommerce industry grows and we make more transactions online each year than the year before, the problem is only going to get worse. As a business that accepts payments online, you must take steps to protect yourself.

“Talk to any internet business and you’ll frequently hear that fraud is a huge pain in the neck for them,” says John Collison, co-founder of Stripe. “Anyone trading or doing business online has this problem.”

You could invest in a fraud management team to investigate every transaction for potential fraud, or you could use a payment processor that takes care of it for you. We strongly recommend Stripe (which our WordPress plugin, WP Simple Pay works exclusively with) because it comes with one of the best fraud detection tools: Stripe Radar.

What Is Stripe Radar?

Radar is a machine learning detection algorithm that identifies fraudulent activity on payments that take place on Stripe’s platform. It sits unseen in the background, quietly reviewing every transaction for signs of fraud and provides real-time scoring for every incoming payment based on thousands of signals.

Since 2016, Radar has blocked billions of dollars in fraud across the Stripe network for companies of all sizes.

Stripe Radar

Radar isn’t a static piece of software. It learns and evolves every day based on new models and data. It constantly evaluates patterns across the entire Stripe network and then applies new learning to its fraud detection so you get the best performance.

Considering Stripe has millions of customers and processes billions of dollars in transactions annually, they have access to a tremendous trove of behavioral data about users and usage to run through its algorithm. This is why it’s so accurate at predicting patterns of fraud.

Basically, as more payments come in through the entire Stripe network, Radar gets smarter and protects you better.

Check out how custom products vendor Teespring benefits from Stripe Radar:

When other payment processors recognize fraud, they often simply block the card, but they don’t inform the business why it happened. This is frustrating for business owners who aren’t sure why their sales get refused.

Think of it like this: A customer’s card can be declined for several reasons. Maybe the processor doesn’t accept that kind of card. Maybe the customer didn’t have enough money. And yes, there’s a chance the payment could be fraudulent.

False negatives (when the processor misses fraudulent activity) are obviously bad, but so are false positives (when the processor declines the card for fraud when there isn’t any) because it means the business misses a legitimate sale.

Stripe’s goal with Radar is to provide you with more detail about why cards are refused so you can take steps to remedy the situation and give you control over what gets refused. You may want to reach out to the customer for better payment information or take steps to protect yourself from that type of transaction in the future.

Stripe’s dashboard shows you how many high-risk payments were blocked by their established rules, how many payments were blocked by your rules, as well as your merchant dispute rate percentage (often called the average chargeback rate) well as the payments that were blocked by your custom rules.

Stripe Radar

In their effort to continually improve their fraud prevention tools, Stripe launched a major update to Stripe Radar in April 2018 that dramatically improved its performance. They managed to reduce fraud by over 25% without increasing the false positive rate.

They also optimized their machine learning algorithm and added hundreds of new signals to improve accuracy by distinguishing fraud from legitimate transactions, including the ability to track buying patterns of customers (because according to Stripe, there’s an 89% chance they’ve seen that same card on the Stripe network at some point).

Stripe Radar’s biggest benefit is that it’s already integrated with your payment processing tool. It’s simple, just like Stripe’s other services. Even though you can customize it, Radar doesn’t require a special data team or fraud specialists to monitor and operate.

Stripe co-founder John Collison says the company’s mission is to “take out the complexity from growing your business.”

Radar for Fraud Teams

Early this year, Stripe also unveiled Radar for Fraud Teams, a paid product designed for companies who are especially prone to fraudulent transactions and willing to invest more deeply in protection.

To be clear, Radar for Fraud Teams is a separate product from Radar. Everyone gets Radar, but you have to add Radar for Fraud Teams to your account and pay the additional $0.02 per transaction.

According to Stripe, “Radar for Fraud Teams helps you fine-tune how Radar operates, create custom fraud rules, and optimize manual reviews. See fraud insights, assess rule performance, and view manual review efficiency from a unified dashboard.” It comes with a few unique benefits

  • View new types of data related to your payments when you review them to determine if they’re fraud or legitimate.
  • Set custom risk thresholds based on a numerical score so you can determine your own fraud comfort.
  • Create and maintain lists of attributes (card numbers, IP addresses, emails, etc.) to consistently block and allow.
  • View better feedback when you create new rules by seeing how they affect your historical data.
  • Look at payments holistically (rather than in isolation) by comparing them to previous related payments that match similar attributes.
  • Explore and evaluate your fraud prevention efforts with rich analytics.

Essentially, Radar for Fraud Teams gives you more control over Stripe’s fraud system so you can make changes specific to your business. It’s designed for companies who have special people or teams who prevent and rectify fraudulent activity.

How to Use Stripe Radar

Stripe Radar is a free service that comes with every Stripe account. You don’t have to do anything to activate or use it. Stripe says it’s “fully integrated into your payments flow and you can start using it without any additional development time.”

To access Radar, click the menu link on the left side of your Stripe dashboard.

Stripe Radar

Under that link, three more options will appear.

  • Reviews: A prioritized list of payments you need to investigate personally (more on this in a minute).
  • Rules: Custom optimizations to Radar for your unique goals.
  • Lists: Information used to create your rules (suspicious IP addresses, email addresses of fraudulent customers, etc.). Radar comes with several default lists.

Stripe Radar uses machine learning as part of its fraud detection system, which means it gets smarter over time based on your transaction history and new data from the Stripe team. The purpose is to avoid the “one size fits all” approach to payment processing that most processors offer.

Over time, you’ll want to get the most out of Radar by setting your own rules. By adding rules, you can make some customizations to the tool to support your preferred payment workflow. You can set three types of rules:

Block rules let you block payments you strongly think are fraudulent, even if Radar doesn’t catch them. For instance, you might block payments from a particular country or a type of card.

Allow rules force Stripe to accept payments, even if Stripe Radar might flag them as fraudulent. For example, you might allow all payments from a particular IP address. Allow rules override all other rules including Stripe’s machine learning models, so use these with extreme caution.

Review rules place any qualifying payments into a review queue for you (or someone on your team) to review personally. For instance, you might insist on personally approving all payments over a certain dollar amount.

Protect Yourself

Credit card companies have become better at responding to fraud disputes, but it’s still smarter to prevent fraud in the first place. Instead of trying to claw your money back, it’s best to prevent fraudulent transactions before they ever occur. With a tool like Stripe Radar, you can build your business without worrying about the cost of fraudulent activity.

Start Accepting Payments Today

Start accepting one-time and recurring payments or donations on your WordPress website.