Riskified is an eCommerce revenue protection and fraud prevention solution dedicated to ‘turning shoppers into customers’. We sat down with Ana Vidal from Riskified’s Team for an in-depth talk about how merchants can protect themselves against fraud when expanding into international markets. we touched on how to prepare for an international launch, calibrating fraud prevention strategies based by location, PSD2, the Russian market, and more.
1. Let’s start at the highest level. What would you say to a merchant who wants to expand into international markets, but is currently letting a prior traumatic experience with chargebacks (in either domestic or foreign markets) hold them back?
Fear of fraud is holding merchants from capitalizing on new, growing markets, and this is not a new thing. It’s also not a coincidence that many of the fastest-growing global markets happen to have bad reputations when it comes to eCommerce fraud – so it’s no surprise that online merchants are apprehensive about expanding into this unfamiliar territory. In fact, phenomenons like geoblocking, and failure to support foreign payment methods are still all too common. The reality is, that the benefits of selling overseas are far greater than the risks. A good fraud solution with a global track record is all that merchants need to mitigate the risk and start capitalizing on new markets to help create additional revenue streams. If this fraud solution offers also a chargeback guarantee, merchants don’t have anything to fear: their income will increase with approval rates, and they will avoid the cost of false rejections.
2. What are the three most important ‘to dos’ for merchants as they prepare for an international launch?
The most important ‘to dos’ ahead of an international launch are:
1 – Know the market: understand how customers in the region shop online, and how they prefer to pay for their online goods.
2 – Localize your product and address cultural sensibilities: Some markets will do fine in English, while others will shun your site if it’s not localized. Consumers will feel understood and consequently safer if you customize your offering to them, showing that you ‘speak their language’.
3 – Protect your investment: new markets bring unknown challenges and surprises, and it is crucial to mitigate any potential risk. The best strategy is to partner with a solution whose core focus is fraud in the new region.
3. Expanding globally requires moving into unknown territory. What can merchants do to cut down on the amount of time it takes them to get educated on fraud patterns in new, international markets?
The simple answer is to partner with someone whose core focus is fraud in that region, a company with the experience, data and ability to take over the risk and help the merchant increase their benefits. Otherwise, as you say – it will take a tremendous amount of time, effort and resources to try and master these things on their own. On the other hand, a machine learning fraud solution can automatize the learning process across different geographies. For example, it may uncover that a shipping/billing mismatch is a significant sign of fraud in sneaker orders placed with Albanian credit cards, but not particularly alarming in other cases. This is the sort of insight that a solution that can account for vast amounts of data provides, and it can greatly boost order processing efficiency and accuracy.
4. You wrote about the false declines problem from a European perspective. How should merchants expanding into Europe be recalibrating their fraud prevention strategies from what they use in the United States? Are there any assumptions that remain valid across both markets?
False declines are a global problem, with many similarities, but also differences across regions. For example, Europeans are very cognizant of the risk of identity theft; they are more likely to have separate credit cards used exclusively for shopping online, and to replace them often. They also often create emails unrelated to their real names or don’t provide their full names on social media. These identity mismatches, new emails and frequent changing of credit cards, can propagate false declines. Additionally, made of up many different countries – Europeans are accustomed to buying cross-border – more so than North Americans. In that sense, the strategy taken up by US merchants is less likely to account for cross-border sales as a significant portion of order volume. But expansion into Europe mandates a command of the challenges of cross border transactions.
Fortunately, there are ways for merchants to improve accuracy when reviewing European shoppers’ orders. Firstly, behavioral analytics. By observing shopper behavior on your site, including which link they clicked to arrive at your store, and which pages they visit during this shopping journey, you can help determine the legitimacy of an order. For instance, good customers tend to visit an online store multiple times and browse around a lot before purchasing, while fraudsters go straight to checkout.
Finally, embrace machine learning. Human analysts are often simply unable to discern big-picture, country-specific fraud patterns. They’re also prone to biases–for instance regarding orders from certain countries as risky, even though the statistics say otherwise. A machine learning fraud solution can learn to treat orders from different geographies differently, which is especially important when expanding into a territory that is comprised of 44 countries.
5. Staying on the European market, PSD2 is well-intentioned but proving to be problematic in practice. What’s the biggest problem you’re seeing the directive create for merchants that they hadn’t realized would be an issue before it’s too late?
PSD2 was meant to protect shoppers, and to help merchants by shifting liability for chargebacks to the banks. But despite g the good intentions – it has created a new problem. Merchants, who have accumulated vast knowledge and expertise in managing fraud over the years, are now completely dependant on bank decisions. But the issuing banks, who have little expertise in detecting online fraud, are tasked with making decisions on what is fraud and what isn’t. They will likely play it extra safe, and will rely on customer friction – mainly requiring shoppers to undergo Strong Customer Authentication (as mandated by PSD2), which we know leads to cart abandonment and drop off. In essence, PSD2 will leave merchants with less control over decisions, and will likely result in a drop in completed sales.
6. The Eastern European and Russian markets specifically are providing explosive growth opportunities for merchants, but are traditionally thought of as risky markets. What are some things about this market specifically that merchants should be aware of in order to make better fraud prevention decisions?
Merchants need to remember that the vast majority of orders coming from the Eastern European and Russian markets are good orders. Online shoppers from that region are happy to buy cross-border and to take advantage of the great prices merchants in North America and Europe offer them.
In terms of fraud, merchants aiming to gain traction in Eastern Europe, or any new region, should make sure their solution has a proven track record in the region. Also – leveraging the power of machine learning solution is advisable – as it can learn the shopping trends, and discern the good orders from the bad faster, and with greater efficiency over time. That will keep merchants safe, while reaping the rewards from this up and coming market.
7. The new, lower chargeback thresholds rolled out by VISA and MasterCard decrease merchants’ margin for error when expanding to new overseas markets. What best practices should they be following to manage the risk during the initial rollout?
Merchants must keep their chargeback rates low in order to stay within a safe range, with enough room for error not to land in an excessive chargeback program. But this often leads to an over cautious approach that compounds false declines. Merchants need to make sure they’re always looking at chargeback rates in relation to approval rates. For merchants in risky verticals, it’s too tempting to reduce chargebacks fast by over declining. Best practice is to look at the entire picture, and find the optimal balance of chargeback and approval rates.
Merchant who can’t find that sweet spot, or are in an excessive chargeback program – are simply doing something wrong, and need to revamp the way they manage fraud. They need to invest in a fraud management solution that will keep chargeback rates at bay, without curbing growth, and to leave enough room for risk-taking – so that new products can be rolled out, and new markets penetrated.
Riskified is an end-to-end eCommerce solution that helps enterprise merchants realize more revenue and cut costs. By leveraging advanced machine learning algorithms to weed out fraudulent orders, Riskified enables merchants to fulfill a higher percentage of legitimate orders.