What are the reasons for false declines?

False declines happen when an ecommerce business automatically declines transactions that look like fraud but aren’t fraud.

These companies might have a low chargeback rate, but they also might not know how many good orders they’re losing. The way that companies decline those orders is with strict fraud filters.

Fraud Filters

Fraud filters are the rules a company sets to prevent potentially fraudulent orders from processing in your store. Practically every modern ecommerce platform has a selection of fraud filters. When used correctly, fraud filters can warn a business that a transaction is potentially fraudulent. But if they are set up to be too strict, they can cancel an order entirely.

The most common types of fraud filters include:

Daily or hourly velocity filter

This filter controls how many sales may be submitted to your website over a certain period of time. This helps prevent fraudsters from testing stolen credit card numbers after purchasing lists on the black market. It also can highlight when a fraudster is ordering multiples of a product to capitalize on a discount or sale.

Address verification system (AVS)

This filter compares billing and shipping addresses to make sure they match. Often, fraudsters use stolen data to purchase goods and ship them to the closest possible location in an attempt to outrun the company’s manual review process.

Card verification value (CVV) filter

This filter looks for discrepancies between a card’s CVV number and the one entered during checkout. Keep in mind that fraudsters know this filter exists and can easily include it in the data they hack from datastores with weak security. It’s also fairly easy for them to collect this information when criminals commit ATO fraud and take over a victim’s accounts altogether.

Purchase amount filter

This filter looks for higher-than-usual transaction amounts. Most companies forecast based on average ticket value, so this filter allows for a threshold equal to or above that value. It’s also a good idea to use this filter for unusually low transaction amounts, as they could indicate that a fraudster is testing out stolen payment credentials and trying to stay under the radar.

Geolocation filter

This filter can be set to decline orders that originate from specific regions of the world. For example, if you know that there's been an extraordinarily high incidence of fraud within certain ZIP codes, provinces or even countries, you can prevent any of those transactions from processing.

But just because an order looks suspicious doesn’t mean it’s fraud.

For example:

An AVS fraud filter could potentially turn down an order that a grandparent made for their grandchild’s birthday.
A velocity fraud filter could mistake a customer who’s taking advantage of a sale in good faith as a fraudster trying to buy out the inventory of a product for resale.
A purchase amount fraud filter could consider every customer transaction a fraudulent purchase during holiday shopping.

The potential for mistaking good customers for fraudsters is high when fraud filters are your only tactic.