Preventing Fraud in the New World of E-Commerce – A Technology Strategy to Maximize Protection and Minimize Loss

The pandemic fueled a dramatic shift to e-commerce that no one saw coming. By the end of 2020, online shopping in the U.S. had increased by a whopping 44%, with online sales now accounting for 21.3% of total retail sales.

Preventing Fraud in the New World of E-Commerce - A Technology Strategy to Maximize Protection and Minimize Loss

In this new world of e-commerce, there’s no dispute that manual identity reviews alone can’t keep up with the growing volume of online transactions—nor can they keep up with the increasing sophistication of fraudsters attempting to fool retailers. The potential losses due to fraud can be significant, not only in dollars but in terms of customer loyalty and brand reputation. Read this article to learn why you need automated identity verification to prevent fraud and how to create a strategy that maximizes protection and minimizes loss, including:

  • How the wrong fraud detection strategy can impact your customer’s experience
  • How to tell your identity verification system is working
  • How to choose a third-party identity verification partner

Content Summary

Manual fraud detection processes can’t keep up
Customer experience and risk
Automating identity verification
Chargebacks
How to tell if your identity verification solution is working?
Ekata identity verification based on APIs and machine learning
Developing an effective e-commerce identity verification strategy
Top to-dos before selecting your thirdparty identity verification partner
Conclusion

IN 2019 THE WORLD CHANGED. No one could predict the global pandemic and even fewer were prepared for how consumers would respond. By the end of 2020, online shopping in the U.S. increased by a whopping 44%, to $861.02 billion, the highest annual e-Commerce growth in two decades. Online sales now account for 21.3% of total retail sales. According to PwC analysis, this trend is irreversible. Some forecasts predict that by the end of 2022, digital commerce transactions will have grown by another 20%, to $5.8 trillion worldwide.

Some forecasts predict that by the end of 2022, digital commerce transactions will have grown by another 20%, to $5.8 trillion world-wide.

Manual fraud detection processes can’t keep up

Against the backdrop of ever increasing online transactions, bad actors have found new opportunities to steal from any retailer whose fraud detection systems and processes can’t detect them. In 2020 for example, The US Federal Trade Commission Received roughly 1.4 million complaints of identity theft, compared to 650,000 in 2019, and 444, in 2018.

At the same time as companies were struggling to detect bad actors, identity verification systems and processes were declining legitimate transactions, mostly impacting potential new customers. In an e-commerce world where retailers compete based on customer experience, these errors come with a high price tag. In a 2017 survey of global fraud, The Merchant Risk Council concluded that merchants decline 2.6% of all e-commerce orders due to a suspicion of fraud.

For those customers that were falsely declined, 33% will never again shop with that same retailer.

Customer experience and risk

While rejecting good transactions will have economic consequences, a bad customer review maybe even more damaging. “Many customers will only rate something when they’re dissatisfied,” says Roberta Gamble, Partner and VP at Frost & Sullivan, a research and consulting company, “and if the company can’t complete its service and a customer gets rejected unnecessarily, that’s a big challenge.” “If customers are getting their identity compromised, there’s also a big risk that they will caution others not to deal with a particular retailer,” she adds.

Minimizing risk is a huge part of why retailers need a fraud detection strategy, not only to protect themselves from economic and reputational harm but to also protect the value they can provide to their customers, explains Noelle Wiggins, Product Manager at Ekata, an identity verification solutions provider. “When online retailers think about reducing risk, they typically think about reducing the risk of bad actors participating on their platforms. However, the other way they need to think about risk is about the value of the customer experience and the value of their trust. “One bad PR event where a customer identity was compromised can have devastating effects and will impact business credibility,” she adds. Maintaining trust and online safety for consumers is critical to maintaining credibility and customer loyalty.

Automating identity verification

The need to verify customer transactions correctly at an ever-increasing speed and volume, particularly at peak sales times, has many retailers turning to automated decision platforms for verification. In the early days of e-commerce, the manual review was an effective way to catch low-hanging fruit because the bad actors weren’t very sophisticated,” Noelle Wiggins explains. “When you’re dealing with high volumes and highly sophisticated BOT types of fraud, humans can’t detect it as well as an automated system might,” she concludes. And when consumer behavior becomes unpredictable, it compounds this problem. For example, the first months of COVID created a Black Friday situation well into the new year, and, as retailers scaled back their fraud detection staff after the holidays, they were faced with a serious problem. Very sophisticated retailers only had a small percentage of their transactions going to manual review, says Wiggins, but mid-sized companies were relying on manual review. “Those companies optimized their basic rule sets for normal volumes and had exactly the right number of review agents. However, with the pandemic, all of a sudden they had Black Friday level volumes coming in and there was no way they could dig themselves out of that fast enough to handle the necessary reviews.”

“When you’re dealing with high volumes and highly sophisticated BOT types of fraud, humans can’t detect it as well as an automated system might.” – Noelle Wiggins, Product Manager at Ekata

Late in 2020, as the pandemic roared on, the sheer volume of holiday sales demonstrated that relying heavily on manual reviews had a significant downside. Companies learned that simply adding new manpower to peak shopping times wasn’t going to work in the new e-commerce environment. “During peak holiday seasons the rush to shorten the fulfillment cycle puts added pressure on fraud detection, making manual processes all the more inefficient,” says Roberta Gamble. The urgency to deliver in record time shortens the window that companies have to identify the validity of their transactions, she adds, and traditional fraud prevention systems are not able to handle the larger volumes. “Solving that problem comes down to automation and agility, she concludes, and companies felt that in a big way in the last two months of 2020.”

Chargebacks

As transaction volumes increase, so do the risks of chargebacks due to identity fraud. This can have significant consequences for banking relations, Vikrant Gandhi, VP at Frost & Sullivan explains. Truly fraudulent chargebacks tend to come from two sources – stolen credit cards where the owner has realized the card details were stolen and opportunistic fraudsters using their credit cards and denying they made the transaction. “These chargebacks not only hurt the merchant and create a bad customer experience, but credit card processors and banks have their internal rating system, and frequent chargebacks could put the merchant in poor standing with them, ” he concludes.

U.S. shoppers spent $201.32 billion online in December 2020 up 45.2% from $138.65 billion in 2019.

“The dramatic growth in online sales has ushered in a new wave of fraudsters. Even the smallest merchant is now a target of global fraud.” – Noelle Wiggins, Product Manager at Ekata

How to tell if your identity verification solution is working?

The number of chargebacks is a great fraud detection signal, says Noelle Wiggins. “While not all chargebacks are fraud, a lot of them are. So if chargebacks start declining, it’s a good indication that an identity verification solution is working.” The other thing to look at in an account signup workflow or a transaction workflow is cart abandonment or signup abandonment. “If those numbers go down, this would be a good indicator that you have successfully reduced customer friction by implementing identity verification,” she adds. In terms of what Ekata does, “I like to think of us as being able to link any digital interaction to the human behind it. In that way, we empower companies to combat fraud, build customer trust, and create frictionless experiences for the customer. As a Mastercard company, we leverage our proprietary data set, identity elements, and machine learning models to correlate behaviors with identities.”

Ekata identity verification based on APIs and machine learning

Ekata offers two distinct and mutually exclusive data sources, the Ekata Identity Graph and the Ekata Identity Network that make up their Ekata Identity Engine. They use these two data assets, along with sophisticated machine learning capabilities, to power all their global API and web app solutions.

Ekata Identity Graph uses licensed data from authoritative providers to create a robust representation of a person’s digital identity by validating the five key identity elements (name, email, phone, IP and physical address) and how they are linked to each other. For businesses, Ekata Identity Graph answers a variety of questions including:

  • Does this email belong to the person?
  • Is this physical address valid?
  • Is the address residential?
  • What type of phone number is this?

Ekata Identity Network uses data aggregated from over 5 billion anonymized and hashed real-world digital interactions. It predicts fraudulent vs. legitimate users by analyzing behavioral patterns of how identity elements are being used alone and together and behaving online. For businesses, Ekata Identity Network answers a variety of questions including:

  • When was this email first and/or last used?
  • How many times has this address been used?
  • Why are there 50 IP addresses associated with a single phone number?

Ekata’s customers query their APIs or web apps by sending the data collected from customers and used in business decision logic (such as email, name, shipping address, etc.). Ekata’s APIs and web apps query the Ekata Identity Engine to return a comprehensive identity risk score and detailed information on the customer’s data elements to the business.

Developing an effective e-commerce identity verification strategy

According to Roberta Gamble, vendors need to ensure that the experience of valid customers is not marred by well-meaning, but poorly executed fraud prevention tactics. “Merchants need to ask themselves whether they’re losing customers that have become frustrated by having legitimate purchases incorrectly rejected,” she says. At the same time, she adds, “..they need to determine whether their current customer identity validation strategies can correct those errors.”

Noelle Wiggins points out that the same strategies that can minimize fraud loss if done correctly and done with precision, can also maximize value to customers. An identity verification solution can help create a more personalized experience for the customer, she says. “Once we’ve verified their identity, they don’t have to go through high friction steps anymore. We’re going to send them straight to check out and send them their product. Creating these bespoke experiences for customers is only possible if we are highly confident in who the customer is. “

Top to-dos before selecting your third-party identity verification partner

  1. Determine the legalities of sharing customer data with a cloud-based identity verification solution.
  2. Develop KPIs important to your business – ex. Chargebacks, false declines, actual fraud losses
  3. Consider the benefits of moving towards a centralized identity management function, like reduced manual review rates and customer insult rates
  4. Ask about how your data will be protected and stored?
  5. Make sure your validation partner has its security certifications up to date?
  6. Determine if you can test the solution before you buy.
  7. Ensure your partner meets your criteria for data efficacy, global coverage, privacy and security, and speed

Conclusion

There’s no dispute that manual identity reviews alone can’t keep up with the growing volume of online transactions, nor can they keep up with the increasing sophistication of fraudsters attempting to fool the retailer. The potential losses due to fraud can be significant, not only in dollars but in terms of customer loyalty and brand reputation.

High online sales volumes during holiday shopping periods have alerted retailers to the need to automate their identity verification process. The COVID-19 pandemic made many aware that quickly scaling up to meet Black Friday-like demand, was not only challenging but came with other inherent risks. Developing an identity verification strategy to meet the new normal in the online retail world should include the selection of a best-in-class verification partner that can provide the highest level of data security.