Your e-commerce business is growing by leaps and bounds. This is shaping up to be your busiest year yet, and the holiday shopping season is right around the corner. Inventory is starting to come in, marketing is working on the big annual campaign, maybe you’re even hiring seasonal help for packing and shipping. What could go wrong? This holiday season, with fraudsters lurking in every dark corner of the internet, anything can happen.

Even though 1 in 4 SMBs recently surveyed has been hit with fraud (each incident costing an average of $28,000 dollars), 58% of SMB executives believe their businesses are less susceptible to fraud because of their size. It’s no surprise then that growing companies are not investing in proactive fraud prevention, which could prove to be very costly. Online fraud attempts grew 13% during the 2018 holiday shopping season, and experts predict that trend to continue this year.

You don’t have to resign yourself to accepting online fraud as a cost of doing business. There are steps that you can take to fight bad actors. Most importantly, start with a proactive, holistic, and automated fraud prevention solution that will do most of the work for you. Here’s what you should know.

Rules, Models, and Sharing – How Automation Can Boost Revenue

The right fraud prevention solution will take a comprehensive approach to fight fraud through a number of techniques that help your team by automatically approving low-risk transactions and declining high-risk transactions. This is done using predetermined rules that can be fixed to a particular risk level based on data provided like IP geolocation and email address.

In addition to this rule-based approach to automation, some fraud prevention solutions also offer custom fraud models created with machine learning. Machine learning creates dynamic rules tailor-made for your business. Combining rules-based fraud prevention and custom fraud modeling with shared network intelligence allows you to take advantage of detected and prevented fraud from other retailers to strengthen your fraud prevention methods.

Let’s take a look at how each of these methods helps e-commerce retailers improve customer experience, reduce fraud loss, and boost top-line revenue.

Rules-based Fraud Prevention

Fraud prevention rules are static thresholds used by retailers to decline or approve transactions automatically. Using a fraud detection solution that incorporates transactional rules has several benefits to e-commerce retailers including reducing loss from a variety of fraud types and increasing revenue by providing a more efficient checkout process.

Fraud prevention solutions create rules based on dynamic data points that are analyzed to determine the overall riskiness of a transaction. For example, our EmailRisk Score correlates over 200 data fields such as email address, customer name, billing address, and IP address to determine the likelihood that the customer is actually a fraudster. This analysis produces an easily digestible risk score. Here’s where rules come in, working with decision scientists and your fraud prevention provider, custom rules can be created to approve low-risk transactions and decline high-risk transactions automatically.

One global retailer was able to leverage this rules-based approach of proactive fraud detection by using the EmailRisk Score to automatically approve 42% of their transactions, which also reduced manual reviews to 1% over twelve months. With that kind of accuracy for an enterprise level retailer, holistic fraud rules can help a rapidly growing SMB e-commerce merchant approve more transactions faster, ensuring their revenue continues to grow, and brand reputation improves with each excellent customer experience.

Custom Fraud Modeling

A rules-based fraud approach is a good start for growing merchants who want to stop fraud loss and improve the speed of their transactions. However, using only rules to approve or deny transactions has some drawbacks. The static nature of rules means they don’t adapt to changes in your industry or new trends in real time. This is where custom fraud modeling comes in.

A fraud prevention solution offering machine learning-based custom fraud modeling will be able to use transactional data in conjunction with other industry-specific data points to produce risk scores that are unique to your business. These models can be updated on the fly to ensure that you’re getting the best possible fraud detection for your online store. They are created and implemented by your solutions provider with no additional effort on your part, letting you focus on more important things – like preparing for those busy holiday shopping days. Retailers using custom models with the Emailage EmailRisk Score have their models updated every week, an unprecedented turnaround for fraud prevention.

While custom fraud modeling is perceived to be a privilege reserved for big brands, the reality is that any merchant can take advantage of shared network intelligence in their industry to create unique models automatically and stop more fraud than rules alone would.

If e-commerce SMBs want to compete with enterprise retailers, they must start thinking like them and proactively fight fraud to set themselves apart from their competition.

Network Intelligence

Fraud prevention is a team sport. All too often, retailers rely on their own internal blacklists to decline transactions with suspicious markers. There are a lot of pitfalls to this approach, namely that for a fraudster to end up on your internal blacklist, they must have already successfully completed a fraudulent transaction. So basically, in order to prevent that fraud, you must have already lost money, product, or both to that particular fraudster. Fraudsters are clever. If they’ve already committed fraud using a name, email address, or other pieces of identifying information with a merchant they’ll most likely discontinue use of that information with that particular store. The best way to combat this is with shared intelligence from other retailers.

Emailage gathers the information that their customers are willing to share like names, email addresses, IP addresses associated with confirmed fraud cases. This data is added to the database used to create custom models and also in transactional rules. In cases where companies have chosen to share such fraud data, fraud detection accuracy improves by 40% over hit rate when not sharing data. This improvement is no accident and translates into increased revenue and better customer experience.

For example, one unique feature of the EmailRisk Score is that if a retailer confirms that an email address was associated with CNP fraud on their website, this information updates the Emailage database. A few weeks later, the fraudster attempts to use that same email address with another retailer in the provider’s network. The transaction is immediately flagged as very high risk, having been associated with confirmed fraud in the past, and the transaction is quickly and automatically declined or passed to your fraud team for a more in-depth manual review.

Who Needs Sophisticated Fraud Prevention?

The short answer is everyone. E-commerce merchants of all shapes and sizes must work to prevent fraud proactively. Fraudsters vary their targets, spanning attacks across multiple industries and verticals in an attempt to complete the most transactions for the highest profit.

If they want to stop losses, SMBs must proactively invest in a fraud prevention solution that incorporates external data, machine learning customizations, and rules-based automation for speedy approvals.

Combining these factors for a holistic approach creates an efficient, frictionless customer experience while saving merchants money and resources. The right solution will provide streamlined processes and peace of mind that supercharges your holiday shopping season.