Drive Growth With Smart Ecommerce Fraud Prevention: Celerant 2024 Recap

All ecommerce businesses, large retailers and small businesses alike, are one bad experience away from losing their customers to a competitor. With all the options available to shoppers, attracting and maintaining a loyal customer base must be a business priority to protect the bottom line. At the same time, the ecommerce space has become a prime target for fraud. So, how can businesses grow revenue with high approval rates and provide a superior customer experience without making it a hassle?

At the 2024 Celerant Conference, ClearSale Client Solutions Director of Enterprise Accounts, Rick Sunzeri, presented to attendees on utilizing smart ecommerce fraud prevention tactics to drive growth. The presentation featured original research conducted by ClearSale on consumer attitudes about ecommerce fraud.

These insights and suggestions from the presentation can benefit all ecommerce businesses.

Ecommerce Fraud Schemes

As ecommerce evolves, so do fraud schemes. Two types of fraud have become particularly prevalent: account takeover (ATO) fraud and policy abuse.

Account takeover (ATO) fraud — Cyberattack sophistication and data breaches allow fraudsters to gain access to stolen data and make unauthorized purchases and commit identity theft. Businesses suffer due to fraudulent transactions, chargebacks and lost revenue, not to mention damage to the business's reputation. 

Policy abuse — Tactics like return fraud and coupon abuse have become more common. This type of fraud is more difficult to stop because it doesn’t involve stolen data or compromised passwords. Return abuse schemes are even powered by fraudsters sharing information online on businesses to target for unclear or lax return policies. 

The Ecommerce Dilemma

Ecommerce businesses have always sought the right balance between strong fraud protection and a seamless customer experience (CX) with fast transaction approvals. Unfortunately, too much emphasis either way can lead to bad outcomes.

Caution first, customer experience second

As customers lean into truly omnichannel shopping behavior — browsing on one or more devices and purchasing on another — ecommerce businesses must implement more sophisticated fraud detection and prevention strategies than grand, sweeping fraud filters and automated systems. 

Many ecommerce businesses lean heavily on filters and fraud detection algorithms. These models may be effective at identifying fraudulent transactions, but they tend to be overly cautious and trigger false declines. This is especially the case when those algorithms prioritize minimizing fraud losses over maximizing legitimate transactions in pursuit of zero-fraud, a lofty goal but one to be wary of. Here’s why.  

Limitations of fraud filters alone

Relying on solutions that employ a collection of fraud filters can present their own issues. 

Too many fraud filters can create a layering effect where they end up cancelling each other out. Another issue is that fraud filters don’t account for reasonable human behavior and can mistake a perfectly good order for fraud. For example:

• Mismatch filters won’t approve an order that’s being sent to another person as a gift.
• Velocity filters might decline a customer sending the same product to their whole family.
• Purchase threshold filters will consider large holiday-related purchases suspicious.
• Without context, fraud filters can increase false declines, which are more costly than fraud.

The cost of false declines

Monitoring the false decline rate can give a business early warning that fraud prevention has swung too far in the direction of caution. False declines have become such a significant issue that respondents to our original research reported it to be a bigger problem than actual fraud. 

• 18% of consumers experienced a false decline last year.
• 70% reported it to be the same or higher incidence than the year before.

What happens when consumers experience false declines? Nothing good.

• 41% of consumers will never shop on your site again
• 32% of consumers will complain on social media
• 11% of consumers won’t provide clarifying information
• 59% of consumers will consider reaching out to customer service
• Up to 65% of false declines are valid transactions

Responses to false declines vary by generation, with complaints on social media being a common response. Unfortunately, those complaints carry more weight than positive reviews, which is something called the “negativity effect.”

Reducing false declines

Just because false declines are common doesn’t mean customers won’t hold it against a business. Their reactions can range from frustration, inconvenience and embarrassment all the way to taking their business elsewhere. 

Rather than focus on getting to zero fraud, regardless of the number of false declines, businesses should focus on finding that balance between minimizing fraud, maximizing good order approval and costing the least overall. This method can help: 

Resolving the ecommerce dilemma 

The conventional wisdom has been that confirming suspicious orders creates an inconvenience for customers that negatively impacts their experience and the reputation of the ecommerce business. 

The fact that payment fraud is becoming more widespread may be changing that calculation. ClearSale’s research showed that 16% of respondents experienced fraud in the previous year and more Gen Z consumers were victims of fraud, at 21%.

Ecommerce businesses looking for the sweet spot between fraud prevention and CX may be surprised to find an unexpected ally in that effort: the customer. 

Customers Want Fraud Protection 

ClearSale’s survey data revealed an interesting shift in attitude: 90% of consumers prioritize fraud protection over easy checkout, signaling they are willing to sacrifice some of their CX preferences to be protected. And that’s just the beginning. 

Another emerging trend from our original research is that consumers are beginning to consider security more important than price. While this trend still varies by generation, overall, 33% of consumers would rather have a safe ecommerce experience than get a lower price.

Communication Boosts CX

Most customers understand fraud prevention is done with their best interests in mind. They appreciate it when businesses prioritize security and take proactive steps to protect them, even if it means temporarily delaying the fulfillment of an order or requesting additional verification.

Not only that, a whopping 84% of consumers said that an online store that doesn’t protect them from fraud will lose their business. 

Consumers are concerned enough about fraud that they welcome communication from ecommerce stores to confirm a purchase — 81% of consumers appreciate a purchase confirmation call from an online business.

“Obviously, businesses and consumers prefer instant decisions about purchases. However, consumers are seeing the value of an extra set of eyes on suspicious orders. They would rather respond to an SMS or wait for a purchase confirmation over being automatically declined or, worse, being the victim of fraud.” — Rafael Lourenco, ClearSale Vice President

When businesses proactively reach out to confirm a suspicious order, it demonstrates that they are vigilant about protecting their customers' security. This helps build stronger relationships and avoids turning away good customers.

In light of these changing consumer attitudes, ecommerce businesses may want to rethink fraud prevention as a customer experience asset, not just a revenue protection strategy. Being loud and upfront about fraud prevention measures can reassure customers and gain their acceptance of minor delays in order processing. 

Rethinking the Role of Fraud Prevention

When we consider the main challenges for ecommerce, our research shows that four of the five are influenced by the quality of fraud prevention and customer experience. Even product selection affects customer experience and may require you to adapt your fraud prevention strategies to protect highly targeted products.

These are measures that can positively impact both fraud prevention and customer experience.  

The value of machine learning

Unlike filters that make a simple yes/no judgment, machine learning can identify different factors in a given order and assign values to them to generate a more complex fraud risk score. For example, machine learning systems can look at whether the order’s billing and delivery addresses match. If they do, the score doesn’t increase. If they don’t, it increases quite a bit.

Another factor machine learning can consider is product category, as some types of products are more sought after by criminals. Cell phones, for example, are small, expensive and easy to resell, so they get a high score. Mattresses, on the other hand, are big, bulky and difficult to deal with. They actually decrease an order’s risk.

A benefit of machine learning is that each business can set its own risk thresholds. Once an order reaches a certain score, it may be rejected or sent for expert review. This gives the business more flexibility to maintain the right balance. 

Fraud runs on stolen data

When criminals steal card data, it’s sometimes incomplete. Maybe they don’t have the ZIP code or the CVV. So, they test on sites where they can keep entering numbers until they find the right combination, and then they make a small purchase to verify. 

Limiting entry attempts and requiring keyed-in card numbers can prevent bots from using your site for card testing. Keeping the reasons for rejection to yourself prevents criminals from learning how they can do better.

Fraudsters sometimes call customer service after getting an order approved with the victim’s shipping address and ask to change the delivery location. The solution here is to refer those requests to the fraud analyst team or cancel the order and have the customer do it again with the new address so it can be part of the screening data. 

CVV data should always be required. Device fingerprinting, as we discussed earlier, can show if multiple orders from different “customers” are actually coming from a single device. Strengthen your company’s data protection, and you’re helping to fight fraud.

Brand impersonation is increasing 

There are many ways for criminals to impersonate a brand. The one with the lowest barrier to entry is social media—it’s fast and free to create a profile, page or chat, and then they can start engaging with your customer base. 

In many cases, social media is the top of a brand impersonation funnel that starts with engagement and ends with fraud, theft or counterfeit goods — often all three. 

Facebook alone takes down 1 billion or more fake profiles every quarter and has done so for many quarters. But, of course, there are many social networks and messaging platforms where impersonators operate. So, the scope of brand impersonation on social media is huge.

For businesses, though, the impact of brand impersonators can add up fast. 

What businesses can do to protect themselves

In summary, as businesses rethink the role of fraud prevention, these strategies can help achieve the balance that keeps both the customer happy and the bottom line growing:

• Identify your false decline rate to set a benchmark and realistic targets for reducing those declines.
• Continue to monitor your false decline rate using post-processing audits.
• Communicate with customers to confirm suspicious orders. Our survey reveals they’ll appreciate it. 
• Use some form of machine learning to analyze incoming orders to help identify good orders and fraud.
• Monitor your brand on the web, marketplaces, social media and ad networks.

How ClearSale Protects Companies from Fraud and Brand Impersonation

ClearSale's fraud prevention combines AI analysis with a second review for a multilayered approach. The advanced algorithm quickly approves most orders, while experienced analysts review suspicious cases and refine the system for even better accuracy over time, increasing your approval rates and revenue.

• Step 1: Order & Customer Data Collection 
  • The system tracks the customer journey and integrates with major platforms.
    
    
• Step 2: AI Analysis With Scoring 
  • Machine learning identifies fraud patterns and assigns risk scores.
  • Low-risk orders are auto-approved, while others are flagged for review.
    
    
• Step 3: Second Review 
  • Trained analysts consider order details, customer history and external data.
  • Most flagged orders get approved after manual review.
    
    
• Step 4: Double-Check and Contact (if Needed) 
  • A second analyst verifies the findings and may contact the customer.
  • Only after double-checking can an order be declined.

Brand protection

Brand Protection by ClearSale handles continuous monitoring, documentation and takedowns of imposter accounts all across the internet to protect business revenue and reputation. With this handled, businesses can focus on their core business goals, like growth and customer experience, instead of constantly reacting to brand impostors. 

Contact our team with any questions. You’ll learn why ClearSale is consistently ranked as a G2 leader for Fraud Protection.