Riskified: ‘There’s a sliding scale of policy abuse’
Policy abuse expert at Riskified, Eyal Elazar, talks about the ways in which retailers’ policies can be abused, how to tackle it during peaks like Black Friday, and why it’s a growing concern today
Can you tell us a little bit about your role at Riskified?
I’ve been working with large enterprises for the last 15 years, focusing on the adoption of growth solutions. I joined Riskified two years ago during an interesting time when policy abuse had really started to scale and become a larger problem than traditional fraud. I’ve had the opportunity to really immerse myself in the challenge and evolution of policy abuse from the ground up.
What are some hidden costs of policy abuse and rigid terms and conditions during retail peaks like Black Friday?
During the holiday shopping season, merchants put a huge amount of budget to fund initiatives like free returns. In fact, we see that roughly 20% of global ecommerce revenue is spent on these policies, which can backfire and instead result in fraudulent returns, misuse of promotion codes and empty boxes being returned instead of the purchased item. Behaviours like this can drive substantial financial losses, tarnish brand reputations, and disrupt the seamless shopping experiences of honest consumers.
Another costly problem is the manual processes involved when dealing with issues like refunds and returns claims; our research shows that a majority of merchants have no automated system in place to address policy abuse. This approach is costly, time-consuming, prone to human error, and opens the door to creating poor customer experience when a team doesn’t have the scale and resources.
How can retailers tip the scales to manage fraudulent activities without hurting good customers?
Retailers can protect their bottom lines through a number of strategies: by listening to the data, automating fraud detection and collaborating across the business.
By scrutinising large data sets that include account details, cross-network claims and order history, customer behaviours, and previous fraud incidents, businesses can identify trends and patterns that predict future incidents of policy abuse.
Meanwhile, an automated system identifies, flags, and prevents fraudulent activities in real-time, nipping fraud in the bud before it causes significant financial damage.
Collaborating across the business also brings down silos to ensure diverse departments communicate and collaborate on fraud prevention during peak periods of policy abuse.
What are some trends in the online threat landscape that are really worrying for fashion retailers, and what are they doing to battle them?
As retailers introduce return fees, we’ve seen a rise in serial returns abusers who resell items at a higher cost rather than losing money from a refund. To single out the efforts of one-off fraudsters, more merchants are introducing loyalty programmes to offer free returns to existing customers with legitimate concerns, like an item being the wrong size.
Due to generative AI and the proliferation of ‘fraudsters-for-hire’ on the Dark Web, ‘Account Takeover (ATO)’ attacks are becoming more sophisticated, giving bad actors endless information and guides on how to buy goods without paying (or even making a profit). Retailers need to focus on data transparency and connectivity to surface abuse trends and keep pace with methods as they evolve.
There’s a sliding scale of policy abuse, and the boundaries between illegal and legal continue to blur. The financial pressures of a squeezed economy accelerate behaviours like chargeback fraud and promo code abuse, where seemingly good customers exploit loopholes in merchants’ policies. To tackle policy abuse, retailers should use data and machine learning models that are focused on identities rather than emails or accounts.
How are retailers benefiting from AI and merchant data networks to better understand the identities behind transactions?
The key to understanding the identity behind a transaction, is of course data. But beyond data collection, automation is the key to data transparency. Once your data is visible and connected, your team can start to explore fraud partnerships that offer merchant networks with expanded data to draw upon; uncover who actually makes an order and/or claim, and what their patterns of behaviour are across other sites that signal that they are a risk for abuse; apply and automate data-based decisions; and focus on fighting other nuances of policy abuse, such as wardrobing and coupon abuse.
By analysing historical data and patterns, predictive analytics can enable merchants to forecast possible instances of policy abuse, helping them to be proactive in their approach. Ultimately, it will take automation and AI to realise the full value of data visibility to expose threats and opportunities.
In your opinion, do you think another cost-of-living Christmas is upon us, and why?
UK retailers are bracing for a turbulent winter: GfK’s Consumer Confidence index showed the willingness of shoppers to spend has plunged since the summer months. While the cost of everyday essentials outpaces the increase in average UK incomes, saving money will continue to be a priority. With consumers looking to save money wherever possible, merchants need to get to grips with policy abuse to save every pound of profit this holiday season.