Ecommerce policy abuse presents a $100 billion catch-22 for merchants. In a survey of 300+ merchants globally, the vast majority of respondents reported they can’t compete without offering generous policies and promotions — yet those same generous policies invite costly abuse. 

The key challenge in handling policy abuse is identifying the “who” behind the order or claim: Are they a scammer, a repeat offender, or a regular customer with iffy return habits? Without clear identification, businesses risk being too cautious, unfairly rejecting valid claims or charging for returns, which hurts customer experience.

In a podcast hosted by Emerj Artificial Intelligence Research, Riskified CMO Jeff Otto describes how merchants can reframe and resolve the policy abuse dilemma by looking at each purchaser as a “cohort of one” and exercising precision trust management. 

Their discussion offers a fascinating look at how policy abuse is perpetrated, by whom, and how merchants are tipping the balance in favor of good customers and higher profits. Listen in above, or catch the highlights here.

Who’s who in policy abuse

Among the challenges of combating policy abuse are the diversity of abusers involved and the fine line that sometimes exists between a desirable customer and problematic purchaser.

At one end of the spectrum are professional fraudsters “trained on the dark web” and able to plan and commit policy fraud on a large scale. They may be unauthorized resellers, for example, who create hundreds or thousands of fake accounts and manipulate freight forwarding in order to repeatedly use first-time-customer discounts on popular goods. They then flip those products elsewhere online at a profit. 

Those scammers both cut into merchant margins and deplete inventory, causing loyal customers to shop elsewhere.

On the other end of the spectrum are more “friendly” abusers who exploit generous policies in a way that creates a loss for the merchant. For example, the customer who orders six different pairs of running shoes to try on and then returns five unwanted pairs on the merchant’s dime. 

The hidden costs of this friendly abuse can be staggering. Packing and shipping, reverse logistics, and restocking costs add up, as do the environmental impacts of trashing items that can’t be restocked. But because the more innocuous abusers may not realize the economic and environmental impact of their actions (and may otherwise be profitable customers), addressing the problem requires a deft touch by the merchant.

Clearly, one size does not fit all when it comes to policy abusers. 

Using identity resolution to get to the “cohort of one”

With advanced AI-driven identity resolution, merchants are increasingly able to analyze account and transaction data to understand where each instance of abuse originates and then tailor the customer experience and policy decisions accordingly. Otto shares that the decisioning around abuse and legitimate process can be specific to the individual, the “cohort of one,” to ensure the best customer experience.  

Using a large graph of data and by examining dozens of attributes ranging from keyboard language to product type, merchants can triangulate signals, see relationships among shady transactions, and identify who is behind abusive transactions and what level of risk they pose. 

Calibrating the trust dial

With accurate identity resolution, merchants can adjust the level of friction for each customer in a way that preserves a great experience for desirable customers, discourages friendly fraud, and kicks the worst offenders out of the store for good. 

This is how merchants are resolving the dilemma of policy abuse. Instead of taking a blanket approach that simply eliminates the generous policies good customers expect, it allows for a precision approach and an overall “pro-trust” stance.

Advanced identity resolution allows merchants to adapt the purchase experience for each customer using what Otto refers to as a “trust dial” that controls the level of friction customer by customer.

For example, a verifiably loyal customer may be allowed plenty of policy wiggle room. A high-risk customer will have a firm all-sales-final policy applied. And the worst bad actors will simply be blocked from the site, even when they pop up under a new fake account.   

What’s more, the merchant can adjust where the friction takes place in the user journey, stopping a serial item-not-received abuser at checkout, for example. Other abuse may be addressed at the dispute resolution stage, either by routing risky customers to specialized customer service teams or by implementing more automated systems for disputing chargebacks.

Rethinking rules

Resolving the dilemma of policy abuse for merchants requires abandoning black-and-white, rules-based decisioning for a more nuanced approach — one built on accurate identity resolution and variable-friction journeys.

This approach not only protects the investments merchants make in creating a positive experience for profitable customers, it also stops more abuse. Otto cites one head-to-head example during a holiday surge in reseller abuse. While a blanket rules-based solution caught 27% of fraud, Riskified’s calibrated approach prevented 85%.

Listen to the full podcast to learn more about what day-to-day fraud looks like and how merchants can get out of the policy abuse trap.