What is Fraud Scoring?

Sydney VaccaroFraud PreventionLeave a Comment

What is Fraud Scoring
A fraud platform performs a risk analysis on an incoming transaction and gives a fraud score. The rating is based on predictive technology that can identify patterns of fraudulent activity.

Dispute management has two sides. Merchants need to prevent disputes from happening and respond to invalid disputes when they do happen. In this post, we will focus on stopping chargebacks, specifically using fraud scoring as front-end fraud protection.

Preventing True Fraud

Front-end fraud protection is put in place to prevent true fraud disputes from happening. True fraud is when a fraudster gets ahold of credit card credentials and successfully uses them at a merchant’s site or store. The actual cardholder disputes the fraudulent purchase, their card account is closed, and a new card is issued to them. The card networks have Zero Liability Guarantees, which makes merchants responsible for the fraud loss because they accepted the fraudulent purchase.

Because it is the merchant’s responsibility to prevent these fraudulent purchases, true fraud disputes are not winnable for merchants. Receiving a true fraud dispute will cost the transaction amount, a dispute fee, and the possible loss of merchandise or services. This is why it is so essential for merchants to take the necessary steps to prevent the acceptance of fraudulent purchases.

The only way for merchants to stop true fraud is by putting preventive measures in place that stop fraudulent transactions from ever happening. This front-end protection could include fraud scoring, fraud filters, and manual review.

What is a Fraud Score?

A fraud platform performs a risk analysis on an incoming transaction and gives a fraud score. The rating is based on predictive technology that can identify patterns of fraudulent activity. The higher the fraud score, the higher the likelihood of that transaction being fraudulent. To get this score, the programmed logic evaluates individual components of the transaction and compares the data to previously identified high-risk attributes. The different data elements that the fraud score gets based on include:

  • IP address
  • Email address
  • AVS results
  • Shipping and billing address match

What is a Fraud Filter?

Fraud filters are used by merchants to filter out fraudulent transactions from valid ones. These filters are customizable and should be based on your industry, customer behavior, and other factors. One of the factors that merchants can program into their fraud filter is the fraud score. If a fraud score is high, the fraud filter can automatically reject it. Similarly, the filter can send transactions to manual review or automatically approve them based on the score.

A fraud score is just one example of the possible filters that merchants can put in place. Filters can be layered on top of each other, all looking at different variables. When set up correctly, fraud filters can be extremely helpful in preventing true fraud disputes while still letting legitimate customers in. When set up incorrectly, it can cost merchants in fraud and lost sales from false positives.

Merchants Need the Full Picture

The goal of front-end fraud protection is to approve valid customers and keep out fraudsters. The only way to get your front-end fraud protection as smooth as possible is to use feedback loops. A feedback loop is when a merchant takes data from various feedback sources (dispute data, manual review team, etc.) and uses it to make the front-end fraud filters as accurate as possible.

Merchants should look at data such as:

  • What transaction turned into disputes
  • What dispute came from each manual reviewer
  • How many disputes were lost
  • What reason codes were attached to the disputes

With this information, merchants can take meaningful actions. Merchants can update the fraud filter rules, change the manual review process, or even retraining a manual reviewer.

Leave a Reply

Your email address will not be published. Required fields are marked *