Most DTC operators track one number on returns: the rate. If it's sitting near the category average, they move on. That's the mistake. Return rate tells you how much stock is coming back. It tells you nothing about why, and it hides the part that's actually costing you money: a small, repeatable slice of your customer base that is not returning products, it's exploiting your policy.
This is the breakdown of what return fraud and policy abuse actually look like inside a DTC brand, why the "no questions asked" policy that helped you convert is now costing you margin, and the exact framework to close the gap without adding friction for the customers who are buying from you honestly.
Policy Abuse Is Now the Number One Fraud Threat
$849.9 billion in merchandise will be returned across US retail in 2025, equal to 15.8% of all sales. Online is where the damage concentrates: 19.3% of online orders come back, against roughly 8.7% for physical retail. DTC brands, selling almost entirely online, carry the heavier side of that split by default.
Not all of that is honest. The National Retail Federation put return fraud losses at $101 billion in 2023 and projects that figure will clear $115 billion by 2026. Appriss Retail and Deloitte, using a broader definition that folds in policy abuse alongside outright fraud, found 15.14% of all US returns in 2024 were fraudulent or abusive. That's roughly one in every seven returns landing back on your shelf (or in the bin) as a loss, not a legitimate return.
Here's the part most operators miss. Refund and policy abuse has now overtaken payment fraud as the number one ranked fraud threat across ecommerce. Almost two-thirds of merchants report rising first-party misuse, and more than one in four have seen it grow by 25% or more year on year. This isn't a niche problem in high-return categories like apparel anymore. It's a margin leak showing up across every vertical that sells direct to consumer.
Your P&L doesn't have a line item called "return fraud." It's buried inside COGS, shipping, and refund processing, which is exactly why it survives so long without anyone catching it.
The 4 Fraud Patterns Hitting DTC Brands
Almost every fraudulent or abusive return you'll see falls into one of four patterns. Knowing which one you're looking at determines how you stop it.
Wardrobing
Buy, use once, return for a full refund
A customer buys an item, wears it to the event or uses it for the occasion it was bought for, then returns it, tags carefully reattached, for a full refund. It accounts for roughly 60% of return fraud cases and affects an estimated 30% of fashion retailers specifically, but the same pattern shows up in beauty tools, home goods, and anything with a single-use occasion attached to it.
37% of online shoppers admit to doing this at least once. It's not a fringe behaviour. It's common enough that if your policy has no time limit and no condition check, you should assume it's already happening at scale in your account.
Empty Box / Item Not Received
Claim the box arrived empty, light, or damaged
The tell is almost always weight. A return that should weigh two kilos shows up as two hundred grams on the inbound scan. Combined with an "item not received" or "box was empty" claim, this pattern is designed to extract a refund while the customer keeps the product.
Retailers tracking fraud tactics report overstated return quantity (71%) and empty box or "box of rocks" returns (65%) as the two most common tactics they see, ahead of counterfeit or decoy item swaps (64%).
Friendly Fraud (Chargebacks)
Keep the product, dispute the charge with the bank
The customer receives the product, then files a chargeback claiming they never received it or never authorised the purchase. Unlike a return, this bypasses your refund process entirely and goes straight to your payment processor, often with a chargeback fee on top of the lost product and revenue.
This is the pattern most likely to be missed by a returns team, because it never shows up as a return in your system. It shows up as a payment dispute weeks later, disconnected from the original order in most people's mental model of the problem.
Serial Returners & Bracketing
5-10% of customers drive 30-40% of all returns
Bracketing is buying multiple sizes or colours of the same item with the plan to keep one and return the rest. 63% of consumers admit to this practice. It's not always malicious, but it consistently loads fulfilment, shipping, and restocking costs onto orders that were never going to convert into full-price revenue.
Layer on top of that the smaller group of true serial returners, typically 5% to 10% of a brand's customer base, who account for 30% to 40% of all returns. These are the accounts a well-built risk score will catch fastest, because the pattern (frequency, value, and speed of return) is consistent and identifiable order after order.
Why your "no questions asked" policy is the problem
Most frictionless return policies were built by marketing to reduce checkout hesitation, not by anyone thinking about what happens six weeks later. That trade-off made sense when return volume was low. It stops making sense once wardrobing, bracketing, and serial returning have had a year or two to compound inside your customer base. A blanket "no questions asked" policy doesn't just tolerate abuse, it actively signals to the small percentage of customers who exploit policies that yours is worth exploiting.
The Framework: Stop Fraud Without Punishing Good Customers
The goal isn't to make returns hard. Frictionless returns are still a genuine conversion lever for the vast majority of honest buyers. The goal is to stop treating every return the same, because a first-time customer with a genuine sizing issue and a five-time serial returner should not be going through the identical process.
Build a risk score from return history
Pull 12 months of return data and segment by customer: return frequency, return value as a percentage of order value, time between delivery and return request, and whether items come back worn or used. This alone will surface the 5-10% of accounts responsible for a disproportionate share of your returns.
Auto-approve low-risk, low-value returns
The biggest single lever for protecting customer experience is instant approval for returns from clean-history accounts under a set value threshold. This keeps the process frictionless for the vast majority of your customers while freeing your team to focus attention on the returns that actually warrant a look.
Add condition checks and time limits on higher-risk returns
For flagged accounts or higher-value items, require photo evidence of the item's condition, verify inbound package weight against expected weight, and shorten the return window. Fraud attempts cluster near the return deadline, so a tighter window on its own removes a meaningful share of wardrobing attempts.
Trigger verification through Klaviyo before refunding
For any return that trips your risk score, route the customer into a flow that requests photo confirmation, original packaging, or ID verification before the refund is issued. This is a policy decision expressed as an automated flow, not a manual argument your support team has to have order by order.
Report return fraud alongside contribution margin, not separately
Return fraud is a margin problem, not a customer service problem, so it should live on the same dashboard as your contribution margin per order. If a specific SKU, channel, or cohort is showing an abnormal return rate against its peers, that's the signal to investigate before it quietly erodes another quarter of profit.
The brands that handle this well aren't the ones with the strictest policy. They're the ones whose policy treats a clean-history customer and a serial abuser as two different problems, because they are.
What This Looks Like in Practice
A wellness brand I worked with had a return rate that looked entirely unremarkable against category benchmarks, sitting a point or two under the industry average. When we segmented returns by customer instead of by order, a different picture appeared: 7% of their customer base was responsible for 34% of all returns, and that same group had a return-to-purchase-value ratio nearly four times the account average.
We didn't change the headline return policy. We built a risk score, auto-approved returns for the 93% of customers with clean history, and added a photo verification step for the flagged group. Within a quarter, return-driven margin loss on that segment dropped by more than half, and customer complaints about the returns process didn't move, because the vast majority of buyers never noticed anything changed.
The fix wasn't a stricter policy. It was a smarter one.
Inside the system
How we build this for brands
The return fraud framework above runs on the same infrastructure we use for margin protection generally: profit and cash-flow dashboards built from live Shopify and order data, with a reporting agent that flags leakage weekly rather than at month end. For returns specifically, that means surfacing accounts and SKUs whose return-to-value ratio breaks from the norm before it shows up as a dent in contribution margin three months later.
On the customer side, verification flows for flagged returns get built and deployed directly in Klaviyo, so the policy is enforced automatically rather than argued over by a support team on a case-by-case basis. Part of this runs live for portfolio brands today; the full system is what we deploy when we take a brand on.
Margin Audit
Find Out How Much Return Fraud Is Costing You
I'll review your return data, segment your customer base by risk, and show you exactly where policy abuse is eating your contribution margin. No pitch deck. No fluff. Just the numbers and what to do about them.
Book Your AuditFrequently asked questions
What is return fraud and how common is it for DTC brands?
Return fraud covers wardrobing, empty box returns, friendly fraud chargebacks, and serial returning. Appriss Retail and Deloitte found that 15.14% of all US returns in 2024 were fraudulent or abusive, costing retailers around $103 billion. The National Retail Federation projects that figure will exceed $115 billion by 2026. Policy abuse has now overtaken payment fraud as the number one fraud threat ranked by merchants.
What is wardrobing and how much does it cost retailers?
Wardrobing is when a customer buys an item, wears or uses it once, then returns it for a full refund, often with the tags removed. It accounts for roughly 60% of return fraud cases and affects an estimated 30% of fashion retailers. Around 37% of online shoppers admit to doing it at least once.
How do I stop return fraud without hurting loyal customers?
Segment your return policy by customer history instead of applying one blanket rule. Give repeat buyers with a clean return history fast, low-friction auto-approval. Apply stricter checks such as photo evidence, weight verification, or ID confirmation to first-time buyers and accounts with a pattern of high-value or frequent returns.
Should I charge a restocking fee to prevent return abuse?
A restocking fee on worn, used, or non-defective returns is one of the simplest deterrents against wardrobing and bracketing, and most consumers accept it as fair when clearly stated at checkout. It works best combined with a shortened return window. It should not apply to genuine defects or first returns from otherwise clean accounts.
What is bracketing and why does it hurt DTC margin?
Bracketing is when a customer orders multiple sizes or colours of the same item intending to keep one and return the rest. Around 63% of consumers admit to this behaviour. It inflates apparent conversion and average order value while loading fulfilment, shipping, and restocking costs onto orders that were never going to convert into full-price revenue.
How can Klaviyo help detect and prevent return fraud?
For flagged or high-risk return requests, you can trigger a Klaviyo flow that asks for photo evidence, proof of original packaging, or ID confirmation before the refund is approved. The same data can feed a risk score that routes low-risk returns to instant approval and high-risk returns to manual review, so your team only spends time on cases that need it.
About the author
Caner Veli is a DTC operator who has helped 350+ brands fix broken growth engines. He built Liquiproof from zero to 3,000+ global retailers in under 6 years. He now runs the same playbook, supported by AI systems he built himself, for DTC and CPG brands.