Why ecommerce brands need to get loud about order and returns fraud — and how to prevent it

When it comes to fighting order and returns fraud, Theodore Roosevelt’s favorite adage “speak softly and carry a big stick” doesn’t cut it. 

If brands want to ward off bad actors, they need to get loud and carry the biggest stick they can find. 

“Once people figure out you're not a soft target and that you’re watching and checking for this, you’ll see a significant reduction in fraud," says Kyle Bertin, CEO and co-founder of return processing software Two Boxes.

With fraudsters becoming increasingly sophisticated (and working more collaboratively), the time to get vigilant is now. 

The SOP Community recently brought together a panel of ecommerce experts to talk about the trends they’re seeing and how to be systematic about implementing changes that will protect your business. The panel included: 

• Kyle Bertin, Co-founder and CEO of Two Boxes
• Adam Caton, Head of Customer Operations and Logistics at Rapha
• Carson Gilvey, Director of Operations and Supply Chain at CREO
• Sean Scanlon, Managing Partner at Karlon Group

The conversation was rich with actionable insights that you can implement today. Here’s what they had to say.

Watch the panel recording

Understanding popular order and returns fraud strategies

There are three main types of fraud that Bertin sees with the brands he works with at Two Boxes: wardrobing, empty box fraud, and fake tracking ID (FTID) fraud, which he benchmarks at about 10%, 3%, and 3-5% of all returns, respectively. 

Wardrobing is the classic buy-for-an-event-then-return-used strategy. 

“Unfortunately, we see this a lot with athletic apparel,” Bertin says, “where stuff comes back and someone clearly wore it to an exercise class and it’s gross and smelly.” 

Nice. 

Empty box fraud jumps to the next level of deceit. This is when someone returns a box that is empty — or filled with rocks to try and simulate the weight of the product. 

And lastly, we have FTID fraud, which Bertin says is growing at an alarming rate: “When we published our benchmarking report, it was about 1.5 to 2% of the returns that we saw, but we've actually seen it climb more to 3 to 5%. It's growing really fast, and I think it's the most nefarious.” 

When engaging in FTID fraud, bad actors will:

1. Buy from your brand
2. Initiate a return
3. Elect to use a parcel shipping option (even if you offer a drop-off consolidated returns offering)
4. Take the return their label and augment it so that it appears as if it was delivered somewhere close — but not actually to — your brand’s warehouse 

From the customer service team’s point of view, it looks like it was dropped off by the customer because of the limitations in carrier tracking capabilities. From there, the fraudster will reach out to your customer service team and demand a refund. 

Unfortunately, this is something that Caton has been noticing at Rapha: “We’ve experienced a lot of this, particularly in our UK market, where people go the route of laying the blame with the carrier. They say they sent it back, but it never gets back to the warehouse. Regardless, they can say, ‘Where’s my refund?’”

Caton notes the growing sophistication of this kind of behavior: “These people are smart. They don’t place really high orders, they’ll do multiple orders of a relatively low value — products that they know they can probably sell. We see it on reselling platforms all of the time. We can actually see the batch picking order number on the product and trace it all the way back to the customer, but there’s nothing we can do at that point.” 

The group agreed: Once products have been dispatched, there’s nothing that can be done.

Preventing order and returns fraud 

The panel had a number of ideas regarding what to try — and not try — when it comes to taking action against fraud.

The first step is making sure you have the data and dashboards in place to inform your strategies and make fraud detection easier. Some things you’ll want to keep an eye on are: names (bad actors often won’t input names); fraudulent addresses; average order value and any orders that fall significantly outside of that range; how many times customers have returned and if you can identify a pattern; and what your product costs.

“Know your numbers, know what your AOV is, know what your product costs are as a leader in your company,” Scanlon elaborated on that last data point. “Let’s say one customer service team member is having to spend one hour on one order that looks fraudulent, that very well could be a loss to you depending on your product cost, so it’s important to assess the ROI on the time your team is spending on these reviews.” 

Once these baselines are set, and you know what is worth pursuing from an ROI perspective, here are some things you can do to be more proactive in your fight against fraud.

Collaborate with your warehouse or 3PL to develop an SOP for fraud detection

A great force for protecting your brand from fraud exists in your warehouse — if you commit to putting a strong SOP in place and fostering healthy collaboration. 

Gilvey has seen this firsthand. “The folks who are packing the orders are such a good line of defense. I have so many examples of times where we caught fraudulent orders because someone was just paying really close attention.”

Ultimately the folks packing and processing products are going to be familiar with what normal and abnormal looks like, especially if it’s your brand’s warehouse. Don’t hesitate to lean on them to sniff out suspicious orders and train them to spot what those might look like. If everyone knows the bottom-line impact of fraud on the business, it’s easier to mobilize the warehouse, the customer service team, and others to help 

But if you’re working with a 3PL, a more hard-line approach could be in order: “3PLs don’t necessarily care about the return side of their business,” says Scanlon, “so you have to spend more time being very specific in the SOPs.”

What does specific look like? 

According to Bertin, that could mean detailing exact specifications for different garments around what to look for when they’re returned. Does the zipper still zip? Are there signs of wear? Can the item be rehabilitated if it’s high enough value? 

Or also knowing when it makes more sense to cut your losses. 

Another best practice is to order from and return your own products every now and then to check that your processes are being run the way you’ve detailed. If you’ve implemented an SOP, it’s worth QAing that it’s being done to your standards and per your agreement.

Issue refunds after processing returns — and build a buffer between purchase and fulfillment 

Return refunds commonly take place when a return is initiated, but Bertin recommends rethinking this if you’re seeing a lot of instances of fraudulent returns. 

Instead of refunding upon the initiation of the return, wait until the return has been processed. This way you can catch empty boxes, goods that have been worn, or other violations of your return policy.  

“Just try it,” Bertin recommends and see how it goes. More often than not, he finds that it doesn’t affect the customer experience and moves the needle significantly when it comes to catching fraudulent returns. Just be sure to let your customer service team know so they can tag any complaints or feedback that suggests it might not be the right fit for your brand.

Additionally, Scanlon recommends adding a buffer between when a customer purchases something and when it gets fulfilled. This doesn’t need to be a significant amount of time, just enough for order fraud detection processes to take place. 

Take advantage of native Shopify fraud alerts 

If you use Shopify, Scanlon recommends taking advantage of their native fraud detection capabilities. 

The caveat? 

Make sure it’s not turning away real customers. 

“Sometimes the 10-step process that they go through can be too onerous and prevent legitimate customers from making purchases. Because of that, I would recommend having your customer service teams set up ticket tags for this specific instance. So if you see a trend of legitimate customers reaching out and getting blocked you can go back and revisit the process and make sure it’s not too strict,” advises Scanlon. 

A/B test order insurance 

Lastly, you can always experiment with implementing order insurance at checkout. This is something Gilvey tried in a previous role. 

She was working with really high AOV orders and was finding that the costs associated with replacing lost orders were really high. Her team investigated a few insurance providers, made a decision, and implemented it. 

While they found that the actual customer experience of using the insurance was good, it was leading to a drop in conversion. They did some math to determine whether the money they were saving from the insurance claims was comparable to the money lost from the conversion drop. Turns out it wasn’t.

“I think the decision for any ecommerce brand is very product specific,” Gilvey says. “It’s taking a look at your cart value, taking a look at your conversion rate, and running a test. I would advise folks to look at it holistically and not just look at the potential savings in this one narrow area.” 

Determine your baseline metrics, and start experimenting

At the end of the day, fraudulent activity is becoming increasingly organized and sophisticated, and it’s a major concern that brands should be combatting in every way they can. 

If you’re feeling overwhelmed, start with the data mentioned earlier in this article. Once you have your baseline metrics, you can start to experiment with these tactics and see how they work for your brand. Everybody’s customer, product, and processes are different, so everyone will require a unique solution. 

And remember: If you want to discourage fraudulent behavior, you’ve got to get loud.