Now that you have categorized your chargeback codes into high-level chargeback reasons, the next important thing to do is understand the key driving factors behind each chargeback. Once you know the causes, you can start reducing your chargebacks.
Chargeback causation can be broken out into three major categories:
- Service and/or Product Issue
- Friendly Fraud
- Malicious Fraud
Here I’ll explain the key characteristics of each and how to isolate them in the sea of data.
Service and/or Product Issue Chargebacks
Service or product related issues can often cause chargebacks when the customer is unsatisfied with the experience. Some examples include:
- Problem with delivery
- Bug caused order error
- Refund problems
- Order or service quality did not match expectations
- Bad customer support experience
- Unclear policies around cancellations, returns, or exchanges
The key in isolating service/product related chargebacks is to follow the history of the user and the order. When responding and analyzing chargebacks, there should be a basic flow where the fundamental pieces of a transaction are reviewed. This flow should start with the purchase and follow the user actions up to present day. In this review, you should check to see if the customer ran into problems and/or communicated with your team.
For your business, you will need to walk through various order scenarios and isolate all of the potential problem areas. Create a list of these areas and decide how you would know that a problem occurred. Document this process and record your findings. With these findings, you can isolate what parts of your business can be optimized to reduce chargebacks. Should you have a better order tracking system? Is your customer support team causing unnecessary chargebacks? Do you have a product issue that you were unaware of? Analyzing chargebacks can be an excellent way to survey the health and functionality of your product or service.
Friendly fraud is when a user with perfectly normal user behavior files chargebacks. This differs from a service and/or product related chargeback because there is no discernable problem with the order. Everything looks normal.
The purchaser is the authorized card holder, but also the person who filed the chargeback. Depending on your business, the magnitude of friendly fraud can be quite high but these chargebacks are essentially impossible to detect or prevent before they occur.
The major reasons that a friendly fraud chargeback occurs:
- A customer does not recognize the charge
- A customer no longer want to pay for the charge (think virtual shoplifting)
- A family or household member made the purchase
To recognize friendly fraud chargebacks, you need to analyze the user behavior. Friendly fraud is specifically looking at users who file chargebacks as unrecognized or unauthorized. While friendly fraud has commonalities with Service and/or Product Issue chargebacks, those chargebacks can be specifically assigned to a problem with your service. Friendly fraud is more ambiguous – everything on the order looks normal, but it was disputed.
To isolate friendly fraud, focus on platform level normal behavior as well as normal user behavior for this particular user. If user behavior is consistent across the user history then it is likely friendly fraud as opposed to malicious fraud. Some key traits to observe normal user behavior over time are:
- The device used and time zone have not changed
- IP stays in the same area
- Shipping and billing are consistent
- Multiple credit cards with same billing information used historically
- Direct customer communication is present
- Purchasing patterns (types of products and price range) are consistent
Malicious fraud is when a non-authorized account holder makes a purchase, then the real account holder files a chargeback. This is when stolen identities, stolen credit cards, or account takeovers come into play.
Malicious fraud is generally found on chargebacks with codes representing fraudulent or unauthorized charge (no surprise here). Since malicious fraudsters are using stolen credit cards and act quickly, you might be able to isolate some standard characteristics:
- Minimal to no activity on your platform
- Abnormal user behavior (opposite of what’s listed under Friendly Fraud)
- High order value
- Either no communication or extremely quickly types communication
- Communication with minimal to no punctuation or capitalization
- If there are long emails exchanged back and forth, it is more likely to be friendly fraud
- Closed account or no activity post purchase
- Connected to other new accounts via the same device, IP address, shipping/billing address
- Device time zone, ID, and/or IP information looks abnormal / changed from past order history
This is a nonexhaustive listing out of a large pool of tricks that can be used to identify fraudsters. In my later articles on malicious fraud, I will dig deeper into the characteristics. Another point is that if a fraudulent or unauthorized chargeback does not look like friendly fraud or a product/service issue, then you can categorize it as malicious for now.
Isolating why chargebacks are filed is fundamental before you can aim to reduce them. That being said, it is not an exact science and there will be mistakes. Pick a margin of error for whomever you assign to review and allow flexibility. At the end of the day, you don’t need 100% accuracy to reduce chargebacks. It is likely that friendly or malicious fraud will be the bulk of your chargebacks. However, product and service related chargebacks can the easiest to reduce and offer great insight on where your business can improve.