Solutions for plugging the $103 billion profitability leak while still delighting customers

According to research from Opinium and Cebr, commissioned by Riskified, returns, refunds, and exchanges cost retailers $394 billion in key ecommerce markets. And more than a quarter of those returns are abusive or fraudulent, creating significant costs and customer experience dilemmas for merchants. Omnichannel returns (in which goods are bought online and then returned online or in-store) can be especially problematic, with fraud costing upward of $19 billion each year.

One complicating factor is that multi-channel data is often fragmented and incomplete, making it difficult to identify the behavioral patterns that signal fraud.

At this year’s National Retail Federation conference, we spoke with Pedro Ramos, CRO of Appriss Retail, and Jeff Otto, CMO of Riskified, about the unique challenges of omnichannel returns fraud.

Q. What exactly do we mean by omnichannel returns?

We generally categorize returns into three buckets:

  1. ‘BISRIS’ (buy in-store, return in-store)
  2. ‘BORIS’ (buy online, return in-store)
  3. ‘BORO’ (buy online, return online)

Across these return types and channels, there are also multiple customer touchpoints, from checkout to claims support. Each generates isolated data for the merchant. Together, those channels and touchpoints make up the omnichannel returns ecosystem.

Q. Why are these returns so hard to police?

Most merchants don’t have a complete picture of customers across all these channels and touchpoints, so they face a dilemma: Do they maintain customer-friendly, unguarded return experiences for good customers or combat the bad actors who inevitably take advantage? Traditional approaches make it an either/or problem because fragmented data makes it hard to tell the good from the bad, and getting it wrong is costly.

Surveys have found that if you falsely accuse good customers of return abuse, one in three will never shop with your brand again, leaving merchants vulnerable in these omnichannel scenarios.

Q. What are the most common MOs in omnichannel fraud?

Scammers take advantage of omnichannel returns in multiple ways. For example, fraudsters who buy online might submit false item not received (INR) claims both in-store and online, potentially collecting a refund or replacement item twice. 

They may purchase coveted sneakers, a luxury bag, or other easily resellable items and then return a fake item online. Or they may simply steal—in-store or by falsely claiming a chargeback online—and return the stolen goods for a refund.

As you can imagine, sorting through all these claims and returns to identify fraud creates significant costs and burdens for merchants.

Q. How can merchants close the data gaps across channels? 

The solution is to ensure you have a 360-degree view of customer behavior across all your channels and touchpoints. This is 100% achievable: AI-powered technology can essentially guard your entire buyer journey across all channels in real time.

Using global merchant data and advanced models, merchants can gain visibility into the hidden interchannel connections that reveal fraud or abuse patterns.

Machine learning can uncover the true identity behind a shopper account by analyzing diverse data points across all transactions in the network. With an uninterrupted view of behavior, it’s easier for merchants to follow the tracks fraudsters leave.

Q. Will that negatively impact good customers?

Quite the opposite! With this approach, merchants can actually reduce friction for good customers. When you train hundreds of machine learning features on this specific problem, you get surgical precision for each individual fraud and abuse decision.

The result is that you can delight your great customers with experiences like VIP customer support, no-questions-asked refunds, friendly returns, and loyalty discounts while applying friction only to abusers and fraudsters.

Q. Beyond the CX benefits, how does this approach improve profitability? 

The benefits of this approach have a ripple effect. When merchants can ensure consistent and accurate decisions, they enhance both efficiency and customer trust. They see higher approval rates, fewer fraud losses, and stronger margins that boost revenue while reducing risks.

Interested in the finer points? Join us for a deeper dive

In a webinar on March 4th, Jeff and Pedro will share more about delivering superb customer experiences and reducing fraud costs across the omnichannel customer journey.