Custom Rules Guide
Introducing Custom Rules
Welcome to Custom Rules. MaxMind is pleased to provide this functionality to minFraud Score, minFraud Insights, and minFraud Factors customers. Custom Rules are not supported for the minFraud Standard or minFraud Premium services. We’re confident that you’ll find Custom Rules of benefit, both in terms of automating disposition of transactions as well as conserving your manual review resources.
How they work
Use Custom Rules to embody what you know about fraud patterns and practices associated with your business.
Each transaction you send for review by the minFraud service includes important data, data the minFraud service analyzes in order to provide the information you need to accept or reject a transaction. Custom Rules enable you to leverage this data and automatically identify good and bad transactions, as well as transactions that require assessment through manual review.
To get started, sit down with a list of transaction data you send to the minFraud service (your input parameters), as well as the data the minFraud service returns on transactions (output parameters). Then, consider how you use this data to identify fraud. Perhaps you review any transactions associated with high risk shipping addresses, or always accept orders with a total order amount less than $20.
Once you’ve determined the criteria you use to accept, reject, or manually review a transaction, create the associated rules using the Custom Rules interface, accessed from the MaxMind Account Portal. You can create the rules you need to automate transaction dispositioning and no developer assistance is required.
Benefiting from the results
Once the rules are in place, your minFraud query results include a disposition derived from applying your rule set to each transaction. The disposition output indicates if the transaction was accepted, rejected, or requires manual review. Transactions are accepted by default if they do not match to any rule. This enables rapid order fulfillment of good transactions, quick action against bad transactions, and easy identification of transactions requiring manual review.
Accessing Custom Rules
Before you begin, please check the version of the client API you have installed to ensure that it includes disposition as part of the minFraud service response. Then, to access Custom Rules, log in to your account and select Custom Rules from the left-hand menu.
Creating a Custom Rule
Note: Once you create a rule, you can place it in an inactive state, but you cannot delete it. We therefore recommend that you plan out your rules, prior to creating them.
To create a custom rule, follow these steps:
- Enter a parameter.
Choose to build a rule based on an input parameter, that is, a data point sent as part of a minFraud query, or an output parameter, a data point provided in response to a minFraud query. You can also create rules using custom inputs.
- Set an operation.
Use an operator to find transaction parameters matches, for example, a riskScore greater than 20 or a country match to Nigeria.
- Define a value.
Define the value to find transaction parameters matches, for example, a riskScore greater than 20 or a country match to Nigeria.
- Set an action.
Add an additional condition to your rule or select an action. Refer to more detailed instructions to learn how to create one.
Selecting an action determines the disposition assigned to a transaction matching the rule. The dispositions are reject, accept, manual review, or accepted by default.
Transactions are accepted by default if they do not match to any rule.
To create a custom rule, first select an input or output parameter to serve as a basis for the rule.
In this example, we’ll use the input parameter Order amount to place any orders with a total amount more than $497 in the queue for manual review.
- From the Custom Rules page, choose a Parameter type of minFraud Inputs, and then select Order Amount from the Parameter selection panel:
- The Rules Builder displays. Next, set an operation. We are interested in manual review of order amounts greater than $497, so that is the operator we select.
If you are working with a parameter to build a rule that requires matching (for example, you are using a postal code as an input parameter), you will be presented with the option to choose operators such as MATCHES and STARTS WITH.
- Next, define the value as 497:
- Then, complete the rule Manually review orders with a total amount greater than $497 by setting the disposition to Manual Review.
- Confirm custom rule content and order the rule.
Creating a Compound Rule Example
Compound rules enable you to combine two conditions using the conjunctions AND or OR.
Use a compound rule to more precisely identify the type of transactions you wish to accept, reject, or send to manual review.
In this example, we’ll create a compound rule that uses the input parameters Is gift and Has gift message to accept transactions that are of lower risk for triangulation fraud.
- From the Custom Rules page, choose a Parameter type of minFraud Inputs, then select Is gift from the Parameter selection panel:
- From the Rules Builder, set an operation and value in association with Is gift. We are interested in transactions marked as gifts, so we select the operation IS (the only option) and a value of TRUE:
- To add the second condition, click on the green add another condition button. Because we want to look at transactions including the input is gift AND the input has gift message, we choose the operator AND between the two conditions:
- Next, define the second condition. In the Rules Builder, click on Choose a Parameter Type and select minFraud Inputs, then select has gift message from the Parameter selection panel.
We are interested in transactions that have a gift message in addition to being marked as gifts, so we select the operation IS (the only option) and a value of TRUE:
- Then, complete the compound rule Accept orders with the inputs is gift AND has gift message by setting the disposition to Accept.
When you create a new rule, you will be prompted to position it in the list of existing rules. It’s very important to consider carefully each rule’s position.
If a transaction matches more than one rule, it will be dispositioned according to the initial match.
Consider the following example. A company does not want to accept orders from Iraq, due to state department restrictions on business with that country. They also want to accept all orders with a riskScore less than 5.
The rule to reject orders from Iraq must precede the rule to accept orders with a riskScore less than 5. Otherwise, Iraqi orders with a riskScore less than 5 would be accepted.