Сhargebacks negatively affect businesses through costly fees, lost shipping, transaction processing costs, and the time wasted on disputes. More to it, excessive chargebacks can lead to reputational damages or merchant account termination.
What are chargebacks
Disputed transactions or chargebacks are credit card charges that customers request to reverse on a certain basis. Such disputes range from a customer honestly returning something they didn’t really mean to purchase to a fraud trying to trick the system.
But chargebacks are not only about companies losing time and money. Seeing an overwhelming chargeback ratio, payment processors can limit the merchant’s monthly volume of payments, terminate the account entirely or put it in a Terminated Merchant File (TMF). In the worst-case scenario, the only alternatives will be to go costly, choose a “high-risk” payment processor, or create a new legal entity and re-start your business.
So, how can businesses fight chargebacks?
Approach 1: Manual review (payment method verification)
One of the ways to prevent chargebacks is manual payment method verification. When a user is to make their first payment, they are asked to take a selfie with their bank card.
The task is to check that the account owner is not a fraud, that they actually own the card/wallet and do not make multiple transactions using dozens of cards stolen from other people.
However, eyes can’t always distinguish the fakes from real payment proof or expose imposters who make accounts under a different name. Moreover, such checks do slow down payment processing as they require for the transaction manager to look through every suspicious payment—impossible for big corporations with thousands of transactions per day. Manual reviews are also costly to implement as they require more workforce to execute the checks.
Approach 2: Automated review
For an automated chargeback review, a user is required to pass KYC. When making their first payment, an individual has to upload a photo of their bank card (sometimes liveness check is also required). The data will be cross-compared to the personal data (such as the user’s name) gathered from the previous steps of registration. Then, the system allows or rejects the transaction depending on the result.
Approach 3: Big data analysis of devious patterns
The approach uses various tools to ingest data from checkout flows, rich payments data, card network and bank data to detect anomalies and predict fraud risk. The databases can be hosted locally, eliminating any network delays, and be seamlessly integrated into the existing flows.
- IP Geolocation and Device Location (MaxMind, Locaid)
Exposing the geolocation of an IP address or device from where the purchase was made and spot unusual patterns. The feature helps improve profitability and compliance by excluding IP addresses identified as proxies, concealed users and other anonymizers.
- Device Fingerprinting (Kount, ThreatMetrix)
Combining certain attributes of a device to identify it (operating system, the type and version of web browser, the browser’s language setting and the device’s IP address). The uniqueness of the fingerprint makes fraudulent requests to a website stand out as it shows the requests are made from the same device.
- Negative Databases (Accertify, Stripe, Sift)
Screening the cardholder for previous negative history related to fraud. Using data like TC40s, SAFE reports or early dispute notifications, you access a global list of suspicious IP, mailing and email addresses. Every time fraud is detected anywhere within our network, the entire database automatically updates to ensure that your business benefits from real-time protection.
- Social network data analysis (Spokeo, BeenVerified)
Investigating social structures—individuals and their ties, relationships, interactions and links that connect them to fraud. Establishing suspicious patterns of behavior through graph database technology, exposing individuals, groups, relationships, unusual changes over time/geography, and anomalous networks.
- Fraud scoring services (Experian, Mastercard, Sage Pay)
Providing all transactions with a score and rating that points out the level of risk the transaction poses for the business. Additionally, fraud scores are adjustable to a company’s risk level and fit the established flow.
Although there are no 100% guarantees that frauds won’t find a way to get to you or your clients, enabling all of these features is the way to minimize the possibility.
Machine learning algorithms defend against fraudulent logins, payments and look for anomalies in user patterns, given past activity. However, it is not perfect. There are some blind spots that the system can’t cover such as false positives. Orders can assimilate fraud because of an unusual location, delivery address or some other factor.
You can solve this issue by combining machine learning with 24-hour support, spotting fraudulent payments more accurately before they are approved, potentially cutting costs by decreasing chargebacks, reducing manual reviews, and improving user-experience.