Rethinking return fraud in retail

For years, return fraud has been treated as an unavoidable tax on growth — something retailers tolerate in exchange for speed, convenience and customer loyalty. But that mindset is outdated and a very costly approach to returns. Return fraud has become more organized, more opportunistic and increasingly embedded in the return ecosystem. 

David Johnston, NRF’s vice president of asset protection and retail operations, sat down with Happy Returns co-founder and CEO David Sobie to talk about the real state of return fraud, how retailers are rethinking collaboration and data, and why the future of mitigation may require a fundamentally different approach. 

We have seen quite the shift in return fraud over the past couple of years. How has return fraud evolved from isolated abuse into a more systemic problem — and where are retailers still underestimating the risk?

Return fraud at scale, or return fraud that’s really harmful for the retailer, isn’t about a handful of opportunistic shoppers taking advantage of a loose policy. We’re seeing the current return fraud challenge fall into two buckets: First, highly adaptive and organized return fraud schemes, and second, a cultural normalization of fraud and policy abuse. 

Highly adaptive and organized return fraud schemes can do a lot of damage over time, and it’s these schemes that we’re currently focused on defending against with our new AI-driven technologies and layers of fraud prevention. 

This focus is vital. According to Happy Returns’ internal data from 2026 so far, nearly one in five fraudulent orders come from organized networks, and the threat is only going to continue to compound.

We’ve found that these bad actors actively coordinate their tactics and recruit using social media platforms. The online conversations and the patterns we’re seeing point to a shift from occasional abuse to repeatable playbooks shared across bad actors. 

But beyond the organized schemes, we found in our annual 2025 Retail Returns Landscape report with NRF that the cultural normalization of fraud and abuse is broader than many retailers want to admit: 45% of shoppers say it is acceptable to “bend the rules” on returns, and 39% of Gen Z shoppers surveyed admit to returning a fraudulent decoy instead of the original item. 

It takes more than pulling return policy levers to keep up. Managing the return fraud risk takes a systemic, 360-degree approach. 

One of the constant challenges of tackling fraud is also balancing consumer friction. How are retailers managing friction, driving loyalty and ensuring convenience in ways that don't invite fraud or other costly behavior — and where do most return strategies get this balance wrong? 

The returns experience provides an incredible opportunity to drive loyalty, and there is a delicate balance to maintain between preventing fraud and providing a smooth, frictionless experience. 

Where many return strategies get this wrong is at one of two extremes. They either make returns so easy that there’s no real verification, which opens the door to abuse and fraud. Or they swing the other direction and add friction for everyone, which protects margin at the expense of loyalty. 

One way to find a balance between the two is through layered fraud prevention that applies friction selectively. For example, in-person drop-off is convenient, but it is also controlled with returned item barcode scanning. Instant refunds remain available, but not indiscriminately. AI scales review, but humans still verify decisions. 

The best return experience is not frictionless for everyone. It is low-friction for trustworthy shoppers and selectively higher-friction for risky returns. Layering fraud preventions such as in-person item verification, behavioral risk scoring to flag potentially risky returns and AI-powered auditing allows for this approach. 

The point is not to slow the entire system down, but to preserve immediate refunds for most shoppers while routing a small set of flagged returns into extra review before the refund is issued. 

What are retailers doing to identify bad actors, and is there anything holding retailers back from embracing more collaborative, cross‑retailer solutions? 

Identifying bad actors through layered signals is key. What looks like a single suspicious return is often part of a broader behavioral trend, and to truly detect and prevent return fraud, retailers need to understand the behavior behind it: How it starts, how it spreads and how it evolves over time. Behavioral risk scoring makes that possible, surfacing early warning signs that would otherwise go unnoticed and allowing retailers to intervene before fraud escalates. 

Returns are scored using signals such as timing, geography, frequency and historical behavior. A retailer looking only at its own returns will often see isolated incidents, while a shared-intelligence model can reveal repeat actors and emerging fraud tactics much earlier. 

Retailers want confidence that collaborative intelligence will be precise enough to justify delayed refunds, extra verification or differentiated policy treatment. That is why the layered model matters so much. Shared signals alone can indicate risk; when combined with in-person drop-off with item verification and AI-plus-human review of suspicious returns, this powerful combo can make action (like a delayed refund) defensible. 

Looking ahead, will fraud continue to outpace traditional controls and how can technology help retailers stay ahead of evolving fraud? 

Bad actors will always try every way they can to get ahead. Fraud tactics change quickly, so we all need to be prepared for that inevitability. 

Technology helps retailers get ahead in three ways. First, it catches signals earlier, before refund leakage occurs. Second, it scales scrutiny to volumes that humans cannot manage alone; AI-driven technology helps automate comparison, logging and documentation at scale. Third, it allows retailers to personalize policy enforcement. 

Ultimately, retailers do not need to choose between loyalty and control anymore; the emerging advantage goes to those that can apply AI and network intelligence precisely enough to keep returns easy for shoppers and unprofitable for bad actors. 

For those attending NRF PROTECT this June, can you share some details about Happy Returns’ featured session during the conference? 

We’re excited to be hosting a panel discussion at NRF PROTECT. We’ll be covering the current state of return fraud in the marketplace based on recent industry data and reporting. 

We hope people leave the discussion with a clearer understanding of return fraud’s impact, practical examples of how retailers are addressing it today, and a tactical framework to help loss prevention teams prioritize next steps. 

“The hidden threat in your returns: How to spot, stop and prevent return fraud” is on June 10 during NRF PROTECT at the Gaylord Texan Resort in Grapevine, Texas. To register, visit nrfprotect.nrf.com.