In September, the Pew Research Center reported that a majority of 18-29-year-olds in the US, 52%, were living with their parents. Affected largely by the spread of the coronavirus earlier this year and the economic downturn it caused, the share of young adults living at home today is higher than it was at the end of the Great Depression. 

Millennials may not be considered at-risk for COVID-19 complications, but the pandemic has impacted them deeply. One-third of millennials have been laid off due to COVID, and another 42% said their pay has been cut, Pew researchers reported. Recent months saw many young adults return to their parents’ homes as a way to save money. Others vacated their apartments as restrictions pushed them out of locked-down cities. 

Despite the economic downturn, Gen Zs and millennials are leading the shift to eCommerce. More than half of young adult shoppers, 28% of Gen Zs and 24% of millennials plan to shop online more than before. This population has tremendous buying power and the cost of a false decline could cost eCommerce merchants’ their lifetime loyalty. So how can merchants gain from this consumer group’s increased activity when their transaction details defy convention?

Don’t let AVS have the last word

We recently looked at Labor Day sales in the US to try and anticipate activity for this year’s peak shopping season. One of the emerging patterns we detected was an uptick in AVS mismatches: on Labor Day, the volume of orders with mismatched AVS went up 47% YoY, while the total order volume increased by 30%. We attributed this trend in large part to the number of American shoppers that unexpectedly or temporarily moved away and did not update their billing addresses. 

AVS matching is used primarily by US merchants to suss out suspicious transactions. Orders with an AVS mismatch are often flat out declined or routed to manual review. The current rise in AVS mismatches could easily result in an increase in false declines, at a time when many depend primarily on online shopping and overwhelm already busy manual review teams. 

One way to affirm order legitimacy–regardless of mismatches–is to use behavioral analytics. For instance, fraudsters tend to go straight to checkout, while legitimate customers shop around, compare goods, and are far more likely to check out a merchant’s returns policy. By analyzing shopper interaction with merchants’ eCommerce sites, merchants can get a more holistic view of each order. While AVS information is not altogether irrelevant, it’s not enough to make a reliable decision. Today, more than ever, young customers have an abundance of good reasons to shop from different parts of the country. As WFH, Work From Home, is very quickly becoming WFA, Work From Anywhere, merchants will need to update their fraud review systems to account for the dynamism and flexibility the future holds.

Look beyond the order details

For each order, merchants may have many data points: the buyer’s full name, email address, shipping, billing and IP address, device, phone number, etc. The key to accurately identifying a shopper even if one or more of the data points have changed is through elastic linking. Riskified uses this type of machine learning to cross-reference the order against our database. 

When our models encounter an IP mismatch, a classic risky indicator, it analyzes the order not based on whether the IP matches the most recent order, but if it’s been used in the past. A customer living in New York may purchase regularly from this location, but if their parents lived in Maryland, it’s likely that they used this IP address at one point in their transaction history. Based on this linking, our models would be able to recognize future transactions from the Maryland IP address as ‘safe.’

What happens if a shopper is new to your store, and there is no previous history to rely on? Since the pandemic began, 69% of Gen Z and millennial consumers have shopped with a new brand. Through our elastic linking technology, we don’t just compare data from one retailer, but with orders across our entire merchant ecosystem. That way, we can correctly identify and understand the context behind seemingly disparate transactions. 

Conclusion

Merchants end up falsely declining good customers when they rely on rigid and outdated fraud-management practices that don’t take advantage of the latest technologies. Young adults in particular have spending power and pose incredible lifetime value opportunities for merchants. But they’re also at the forefront of change as many have had the flexibility to pick up and move when the crisis hit. To safely approve their orders, merchants need to follow suit and rely on an adaptable system that can lead their most loyal shoppers safely to purchase.