Every business has a different tolerance for risk. A SaaS company with low cost of goods sold might prioritize capturing as much legitimate revenue as possible, accepting that a few fraudulent transactions will slip through. Meanwhile, a retailer selling high-priced luxury goods might need stricter controls, such as collecting more customer information at checkout, even if that means adding friction.
Many businesses tell us they’ve struggled with fraud providers that were either too lax, allowing too much fraud through, or too stringent—blocking legitimate customers. What they want instead is a customizable fraud prevention solution they can fine-tune to match their specific risk appetite and business priorities.
In this guide to evaluating fraud solutions, we’ve compiled the questions to ask potential fraud providers. These are the questions our customers say they wish they’d asked before choosing their previous provider. This isn’t an exhaustive vendor comparison; instead, these considerations are designed to help you identify a fraud solution that can scale with your business.
Accuracy and effectiveness
-
Does the fraud provider use AI-powered risk evaluation, rules-based logic, or a hybrid approach?
-
What’s the size of the provider’s network that trains their models?
-
What inputs or data points are used to determine risk scores and make a fraud risk assessment?
-
How often are the AI models retrained to adapt to new fraud patterns? Are they continuous updates or periodic version releases?
-
How, if at all, does the solution adapt based on shifting fraud patterns? Can it provide proactive insights or take action without manual intervention?
-
What interventions can the solution apply to authenticate high-risk transactions, which can further help reduce fraud? Are those applications intelligent and dynamic, or are they primarily rules-based?
-
Does the tool intelligently trigger 3DS and SCA only for high-risk or mandated transactions, rather than applying it to all transactions?
-
Does the fraud protection extend to noncard payment methods such as ACH, SEPA, and buy now, pay later (BNPL)?
Rules and customization
-
Can I create custom fraud rules to manage how my business handles incoming payments and how it enforces policies to block suspicious activity?
-
Can I create lists of specific types of information (such as suspicious IP addresses or email addresses tied to fraud) and use them in fraud rules?
-
Does the fraud solution suggest rules based on my top fraud indicators?
-
Can nontechnical users write rules in plain English (e.g., “Block if the email domain is temporary”)?
-
Can I run “what-if” simulations on new rules before turning them on, either against historical data or in “shadow mode”?
-
Does the fraud prevention tool allow for manual review of suspicious transactions?
-
Can I write rules based on custom business data (e.g.,
loyalty_tier,product_category, orshipping_method)?
-
Can the system detect and prevent card testing by blocking multiple attempts from the same IP address or email within a defined time frame (e.g., five attempts per hour)?
-
Does the tool provide a granular risk score (say, 0–99) for every transaction, along with detailed insights, so I can set my own thresholds for blocking versus reviewing?
Modularity
-
Does the fraud tool provide risk scores for transactions processed through other payment processors as well?
-
Can the fraud provider offer risk scores and insights for preauthorization to inform payment routing decisions?
-
If I move processors, can I take my “allow or block” lists and fraud history with me?
-
Can the provider securely handle the raw PAN to perform a risk assessment across multiple processors?
-
Is the provider flexible enough to screen transactions at different points throughout the checkout flow?
Analytics and reporting
-
Are fraud insights, payment details, and dispute management available in a single, unified view?
-
Can I sync fraud data directly to my own data warehouse for custom analysis?
-
Is the fraud analytics dashboard updated in real time or is there a lag?
-
Does the fraud provider inform me of fraud attacks in real time? Does it offer recommendations on how to mitigate these attacks?
-
Can I use a fraud dashboard to investigate fraud patterns and emerging vectors?
-
Does the provider give me information about my standing with different monitoring programs?
-
Does the provider allow me to filter and view analytics by payment processor?
-
Can I review each transaction to understand its risk evaluation and outcome?
Disputes
-
What tools or services are offered to reduce or manage chargebacks?
-
Does the fraud provider allow me to programmatically manage disputes using an API, with functionality to upload evidence, respond to disputes, and receive dispute events via webhooks?
-
What tools or solutions does the provider offer to manage and respond to chargebacks? If available, can these tools be automated? Do they include intelligence to optimize for winning a dispute?
-
Does the provider calculate win rate probabilities for individual disputes to help me triage and prioritize responses?
-
Does it provide recommendations for evidence to submit for individual disputes?
-
Does it integrate directly with network solutions such as Verifi and Ethoca to help me manage and prevent disputes, or would access to these network solutions require a separate integration through a third-party provider?
-
Does it provide recommendations for how to mitigate and prevent disputes, down to the individual dispute level?
-
Does the provider enable offerings that notify me when an issuing bank flags fraud—before it becomes a formal chargeback?
-
Can the system automatically build and submit dispute evidence packages (logs, tracking, etc.) to the issuing bank?
Platform and marketplace specifics
-
Can I get access to risk scores for individual businesses on my platform?
-
Can I write custom rules for both transactions and accounts?
-
What actions can I take with custom rules as a platform? For example, can I pause payouts or send accounts to manual review when they meet certain criteria (such as when they reach a certain level of riskiness)?
-
Can I safeguard my business against potential losses by setting temporary reserves on user funds?
-
Can I set up real-time alerts to allow for immediate action when potential merchant fraud is detected?
-
If some businesses raise suspicion but aren’t conclusively fraudulent, can I implement enhanced verification processes, such as requiring additional authentication?
-
Can I access a list of accounts that are flagged as suspicious?
How Stripe Radar can help
With Stripe Radar, you can identify and fight fraud in real time with AI powered by Stripe’s global data. Radar is built into Stripe and seamlessly integrates with the full Stripe platform, allowing you to get started right away with zero code required. And now, Radar’s AI model returns a risk score and recommendation in real time, allowing you to apply Stripe’s risk evaluation expertise for your externally processed payments.
With Radar, you can:
- Automatically detect and block fraud: Radar reduces fraud using AI that trains on data across millions of global companies. Our scale—92% of cards have previously been seen on the Stripe network—enables us to detect and automatically block true fraud more accurately to save you money from fraud losses. While fraud globally increased 15%, fraud on Stripe has decreased by 17%.
- Reduce card testing attacks: Thanks to our Payments Foundation Model, an industry-first AI model trained on tens of billions of transactions, our detection rate for attacks on large businesses significantly increased from 59% to 97%.
- Save time and money on disputes: Unify disputes management from end to end with Radar. You can proactively refund transactions to prevent disputes before they occur with Smart Refunds, reduce your dispute rate to avoid card brand monitoring programs with dispute prevention, and use Stripe’s AI to tailor and submit evidence for eligible disputes on your behalf with Smart Disputes.
- Dynamically apply two-factor authentication: Stripe can handle European SCA requirements and dynamically apply authentication, such as 3DS, when required by the cardholder’s bank or when fraud is suspected—helping to reduce fraud by 30%.
- Manage your platform’s risk end to end: Identify fraudulent businesses on your platform, set custom account-level rules, and intervene on suspicious transactions with Radar for platforms.
- Create rules to customize fraud: You can create custom rules to manage how your business handles incoming payments, blocking any that you would consider suspicious or placing them in review.
To learn more about how Stripe can help your business fight fraud and disputes, contact us or sign up for an account.