Millennials — the first generation of digitally native consumers, or those born between 1980 and 2000 — are now the most populous group of consumers in the world. Their purchasing power is estimated to exceed $13 trillion worldwide by 2020, according to the Brookings Institute. US millennials are expected to spend the most on holiday shopping this year, spending as much as four times more than baby boomers.

As retailers overhaul their omnichannel offerings to appeal to these consumers’ needs, we have noticed a critical blind spot in their approach: imprecise and outdated fraud prevention strategies that fail to accommodate millennials’ unique lifestyle and shopping behavior, creating friction during checkout.

There has never been more at stake for retailers courting a consumer base that grew up during the digital revolution and has high expectations for seamless technology and service. This is especially true for the final online checkout process in the millennial customer journey. An August 2018 survey showed 83% of online shoppers in the US will abandon their carts during the checkout process if it is too long or too complicated, and 55% went on to say they will never visit the site again.

In this post, we will breakdown the two most common eCommerce fraud prevention measures that turn millennial consumers away, and how retailers can fix them.

Missing Context on Millennials’ Dynamic, Itinerant Lifestyles

Millennials are more diverse and global than any preceding generation. Technological advances such as the internet and cloud computing allow millennials to have a location-independent lifestyle as “digital nomads.” The World Economic Forum’s annual Global Shapers Survey 2017 reported that for a large majority of young people, identity is not about region, geography, or ethnicity. About 40% said being “human” is what defines them most, and 18.6% said being “global” or a citizen of the world. Plus, the cost of international travel has never been cheaper or easier. That means when and where millennials are shopping online is dynamically changing, and so can their shipping addresses. An American freelance programmer working overseas may use her US credit card to place an order originating from a Bangkok IP address with a US retailer, and ship the goods to her next destination: an Airbnb in Berlin. That adds up to a lot of geographical data mismatches for the merchant’s fraud detection system to reconcile.

Many retailers decline legitimate orders placed by itinerant millennials because they give heavy weighting to AVS (address verification system) mismatches, and rely on rules-based engines that decline foreign credit cards or international billing and/or shipping addresses. By doing this, they fail to see the broader context that explains these seemingly disjointed and suspicious order details.

For instance, a Riskified fashion merchant previously declined legitimate orders for winter coats, which were worth 15% of their online revenue that quarter in North America, because of AVS mismatches. By emphasizing the billing and shipping address mismatch, the retailer failed to see that the orders were placed by Chinese students who had just moved to Boston for university. They had not been in the US long enough to build credit, and so used their Chinese credit cards to buy the coats to be delivered to their dorms. Judging purely by the Guangzhou billing address and the shipping address in downtown Boston, it’s easy to judge this order as fraudulent. However, factoring in other pieces of information such as a .edu email address, and geolocation and provider of the IP address will offer context: that these are some of the 350,000 Chinese millennials coming to the US for higher education, not fraudsters trying to defraud merchants.

Dangers of Over-Reliance on Geographical Data

The mismatching addresses don’t have to be as extreme as US versus China. In another instance, an online furniture seller repeatedly declined orders by millennials who were purchasing pieces of furniture for their new apartments — exactly the type of customer a furniture merchant would want to acquire and retain. The customer’s banking details had not yet been updated with the new home address, and so the retailer was not able to link this new location to the customer.

According to Pew Research Center, American millennials are more likely to rent than own homes due to having come of age in the shadow of the global financial crisis and having faced macroeconomic pressures that limited access to affordable housing and linear careers. That means, even if they aren’t digital nomads or international students, millennials are quite literally on the move, and on the regular. Again, exactly the type of customer a US furniture retailer would want to acquire and retain.

In this instance, context is again critical. Fraud managers can look at the devices from which these orders are being placed to see if they are the same device the customer has used to place past orders at older addresses. Social media can also be useful, to see if the customer has posted updates about their recent move. But individually checking these data points can be cumbersome and time-consuming. At Riskified, we automate review of all available information, including data across our ecosystem, the customer’s cyber footprints, and third-party databases to reveal the story behind the order in fractions of a second.

Nearly nine in every 10 millennials live in emerging economies, according to the United Nations Population Division. More than one billion millennials currently live in Asia, and Chinese millennials – 351 million – outnumber the entire population of the United States. Retailers looking to succeed across global markets by attracting more millennial consumers worldwide, already know they need to adjust their advertising campaigns and sales offers. What many don’t realize, however, is that they must also restructure their fraud management operations to correctly identify and understand the context millennials’ lifestyle and purchasing behavior add to their orders.

Missing Context on mCommerce and BOPIS

Millennials actually love to shop in physical stores. Turns out their tech-savviness and love for convenience are not mutually exclusive with their desire for authenticity and unique experiences. Approximately 82% of millennials said they believe it’s important for a brand to have physical stores while 67% of them said they prefer to shop digitally, according to eMarketer. That means millennials want to be able to shop both on- and offline, as well as on their mobile devices inside the brick-and-mortar stores. The study also found that 63% of millennials are likely to use buy online, pick up in store (BOPIS) methods — yet another indication of their love for omnichannel shopping.

Channels such as mobile and BOPIS that weave on- and offline touch points throughout the customer journey, add unforeseen complexities for the fraud management team. For instance, fraud managers now need to be so granular in their analyses to distinguish shopping activity on mobile browsers versus merchant-specific mobile apps: according to RSA’ data science research, fraudulent orders in the past were placed more via mobile browsers such as Chrome or Safari, but now 80% of mobile fraud comes from merchant-specific mobile apps.

This doesn’t mean mCommerce and BOPIS orders aren’t worth what may seem like an extra headache. Retailers can’t afford to lose out on 63% of millennials who shop on their phones every day and 84% who use their phones for shopping assistance while in a brick-and-mortar store. In addition, 65% of millennials reported they are comfortable with making purchases on mobile devices. Another recent study showed almost 50% of millennial shoppers (in their mid 20s – early 30s) make a mobile purchase every week.

Omnichannel merchants typically track fraud by either the payment method or by the channel used, and rarely track both together, according to Merchant Risk Council’s 2017 Global Fraud Survey. By tracking fraud in silos, retailers are not able to fully understand where fraud is occurring most frequently, which payment methods are being exploited at what channel. That ultimately blindsides them from seeing the true scope of fraud throughout the checkout, payment, and order fulfillment processes.

Instead of auto-declining mobile or BOPIS orders, we advise retailers to first start collecting and splicing the relevant data points. There are mobile- or BOPIS-specific data points that should be collected and analyzed to identify order segments uniquely problematic for each merchant and the product type, channel, or geography the retailer is engaged in. At Riskified, we elastically search and link hundreds of data points spread across hundreds of millions of historic transactions. That way, we can connect a first-time mobile buyer at one merchant to purchase history at another merchant.

BOPIS is challenging because it takes away a key data point — the shipping address — that standard fraud prevention systems rely on to identify the shopper and verify the order. We have seen attempts to exploit this BOPIS vulnerability grow by as much as 250% since last year at some of our merchants, but that doesn’t stop us from encouraging our merchants to offer BOPIS. Nearly half, or 47%, of millennials say they buy online and pick up in-store more than 40% of time, according to Euclid, and that’s valuable revenue we wouldn’t want retailers to miss out on.

What we advise is to introduce a system that identifies patterns in the shopper’s behavior, such as prior purchases or shopping journey on the site or app, as soon as the customer journey begins. For example, Riskified uses proprietary Beacon technology – a small snippet of code embedded on a retailer’s site – to collect data about the type of device used, the device location, and the customer’s activity on the retailer’s website or mobile app. This data can be used for device fingerprinting, proxy detection, and behavioral analytics.

We can’t emphasize enough the dangers of rejecting or approving an order based on a single data point. Retailers need to take a dynamic and comprehensive approach to fraud review, and look at hundreds of data points and the customer interaction before making a decision. Sound labor intensive or impossible? It doesn’t have to be. Riskified has a fully automated solution using machine-learning models that is set up so retailers always reach their goals.

Request a demo to experience it, or contact us for more information.