Beginner’s Guide to Fraud Filters

Fraud is one of the biggest concerns for businesses that sell their products online. The impacts are widespread: Ecommerce fraud costs your business lost merchandise, lost shipping fees and potential chargeback fees. Not to mention the damage it can do to your company’s online reputation. Even if you’ve successfully avoided fraudulent orders so far, you’re still at risk.

The pandemic ushered in a new era of online shopping and with it, an industry built around fraud. Fraudsters have expanded their reach, enlisting bot technology and creating crime rings to continually find new ways to exploit a company’s vulnerabilities.
That makes fighting fraud a tricky business.
Customers have become accustomed to a particularly high level of service and an online experience that’s equal or superior to shopping in stores. You can no longer rely on a single set of rules to anticipate and thwart fraud attempts.
Companies need to adopt a comprehensive approach that is nimble and can adapt to the changing fraud landscape. One element in that approach involves fraud filters.

Nova call to action


What Are Fraud Filters?

Fraud filters are the rules a company sets to prevent potentially fraudulent orders from processing in your store. Almost every modern ecommerce platform has a selection of fraud filters. Depending on how you set up those fraud filters, they will either warn you of a potentially fraudulent transaction or 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.

What Is Triangulation Fraud?


Fraud Filter Use Varies

Companies use fraud filters for different reasons, often depending on their order volume, risk level and resources.


Small ecommerce businesses tend to lean on fraud filters as their sole fraud prevention tactic. In many cases, small businesses have few staff members; maybe the owner/founder and one or two employees – all of whom have more than enough work to do to keep the business running. Reviewing orders for fraud doesn’t make that list when their ecommerce platform can technically weed out “bad” orders.


Mid-sized and enterprise ecommerce retailers often have a team of analysts who have a secondary or manual review process for orders displaying fraud. But with the swath of transactions that come in daily – even more so during high volume seasons – these retailers turn to fraud filters to decline any order that even looks like it could be fraudulent so their analysts can focus their efforts.

The problem with both scenarios is it gives businesses a false sense of security.


The Challenges With Using Fraud Filters

Fraud filters aren’t meant to be the end-all-be-all for fraud prevention. There are too many factors and, frankly, fraudsters have become too proficient in their profession. As a result, fraud filters present a number of challenges.

Layering and canceling out rules

Companies tend to layer fraud filters in an attempt to out-maneuver fraudsters, but those layered rules can create chaos. In fact, layered fraud filters can cancel each other out, leaving the company open to rampant fraud.
Fraud filters also open companies to rampant errors.

Mistaking good orders for bad

Just because an order looks suspicious doesn’t mean it’s fraud.
For example:

  • A fraud filter to automatically decline transactions where the billing and shipping addresses don’t match, you 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.

You can see where we’re going with this. The potential for mistaking good customers for fraudsters is high when fraud filters are your only tactic.

Nova call to action

Increasing false declines

Those mistakes – when fraud filters incorrectly decline valid transactions – result in false declines. For mid-size and enterprise businesses, false declines are a real issue. They happen when good customers who expect to be approved aren’t allowed to make purchases. The embarrassment and inconvenience can result in negative consequences and even angry customers.
In our original research, “State of Consumer Attitudes on Ecommerce, Fraud & CX 2021,” 40% of customers said they would never again shop with an online retailer after being declined. And as much as 34% will take their frustrations a step further, sharing their displeasure on social media. That spells big trouble for a company’s brand reputation.
When you consider the sheer number of competitors in today’s online marketplace, losing the lifetime value of customers over a preventable mistake is costly: For every $1 in false declines, an ecommerce business loses $13.

Fraud filters definitely contribute to fraud prevention, but they must be set up and utilized efficiently.


Fraud Filters Should Be a Component of Fraud Prevention

Your fraud prevention process should be designed to carefully analyze every transaction, no matter the size, as well as monitor the big picture of incoming transactions. A single small order may not be of much concern – but hundreds of small orders coming in at the same time clearly indicate a fraud pattern and should ring an alarm.

While using fraud filters as your only fraud prevention tactic is ill-advised, they do serve a critical role in a comprehensive fraud prevention strategy. Specifically, fraud filters can help flag suspicious orders that fall into the gray area of “could be fraud but may not be.” Those transactions need to be reviewed to determine how they should be handled.

When companies use fraud filters as a first step in their fraud prevention process, they can decrease the number of false declines that threaten their reputation and bottom line and increase their approval rates.

Better yet, companies that work with a fraud prevention provider can either find an easy-to-implement solution (for small businesses) or gain a partner to help them understand the best way to use fraud filters.

Nova call to action

How ClearSale Uses Fraud Filters

You want to approve as many orders as possible, and you should be able to. The question is, do you have the solution and/or resources to make that possible?

At ClearSale, we use a hybrid fraud prevention model that incorporates several elements:

  • Fraud filters are calibrated to flag obviously questionable orders for processing.
    AI-enabled automatic approval technology leverages data insights and analytics to then approve or decline as much as 97% of orders with precision. Orders that are still suspicious are flagged for review.
  • A team of more than 2,000 fraud analysts who have identified and prevented fraud in the most high-risk regions across the globe perform a secondary, contextual review of those flagged orders (for many businesses only 2-3% of orders are in this category). In some cases, they may reach out to customers to offer “white-glove” fraud prevention solutions.
  • Once those orders are dispositioned, the data insights are fed back into the AI system to help it “learn” about the client’s industry, customers, regions and new fraud trends, which further improves auto-approval accuracy.

For small businesses, the auto-approval technology combined with their fraud filters can almost eliminate the need for their team to review orders. Midsize and enterprise benefit from a combination of services, either as an extension of their fraud teams or to help manage the increase in orders during peak seasons.

By staying alert and vigilant while working with a trusted third-party vendor such as ClearSale, you can better protect your sales, profits and customers from credit card fraud & other types of commerce abuse. To find out more, contact us today.

Ecommerce Fraud Protection for Enterprise Business

How Ecommerce Fraduud Protection Works With ClearSale