Return fraud drains billions from online retailers every year. If you run a small e-commerce business, you feel every dollar of that loss in your margins, your inventory, and your sanity.

The real difficulty isn't identifying that fraud exists—it's knowing what to do about it. How do you tell the difference between a loyal customer who genuinely received a damaged item and someone who's figured out exactly how to game your return policy?

Most guidance on this topic splits into two camps, neither particularly useful. Camp one says "accept all returns to keep customers happy," which is a reliable way to watch your profits evaporate. Camp two recommends aggressive verification that treats every customer like a suspect, which is equally effective at driving away the people you actually want to keep.

There's a middle path. You can enforce return policies firmly, catch serial return patterns before they wreck your margins, and still communicate like a business your customers want to buy from again. This guide provides the frameworks and scripts to make that happen.

What Return Fraud Actually Looks Like for Small E-Commerce

Return fraud isn't a single behavior—it's a spectrum. Some actions are clearly malicious, others sit in a gray zone that requires judgment, and a few are technically allowed by your policy but still unsustainable for your business.

The Patterns You'll See Most Often

Wardrobing hits apparel and accessory sellers hardest. A customer buys a dress, wears it to an event with tags carefully tucked in, then returns it claiming it "didn't fit" or they "changed their mind." The item arrives back with deodorant stains, a faint perfume smell, or seams that have clearly been stressed by actual wear. The National Retail Federation estimates the retail industry lost $101 billion to return abuse in 2023, with wardrobing representing a significant portion of that figure [1].

Serial return abuse follows recognizable patterns: the same customer returns a disproportionate percentage of their orders, sometimes across multiple accounts or slight name variations. Nothing they're doing might violate your written policy—but the economics simply don't work when someone returns 70% of what they buy.

"Item not received" fraud (sometimes called first-party fraud or friendly fraud in industry terms) involves customers claiming packages never arrived when tracking clearly shows delivery. Similarly, "not as described" claims sometimes mask buyers who changed their minds but know that reason won't qualify for free return shipping under your policy.

Bracketing has been normalized by large retailers: customers order multiple sizes or colors intending to return most of them. While not inherently fraudulent, bracketing creates real costs—shipping, handling, quality inspection, potential markdowns—that small businesses can't absorb the way Amazon does.

Why the Playbook That Works for Large Retailers Won't Work for You

Major retailers throw technology and headcount at return fraud. They deploy machine learning detection systems, maintain dedicated abuse teams, and budget for acceptable loss rates as a cost of doing business.

Your situation is different. You're probably handling support alongside inventory management, marketing, order fulfillment, and a dozen other responsibilities. Spending 45 minutes investigating whether someone actually wore that jacket before returning it is 45 minutes you don't have—and 45 minutes not spent on revenue-generating work.

But your small scale also creates advantages. You can spot patterns that algorithms miss, make nuanced judgment calls faster, and build genuine customer relationships that discourage abuse before it starts. The key is having clear frameworks that let you make those calls confidently and consistently.

Preventing Returns Before They Happen

The cheapest return to process is the one that never happens. Before diving into enforcement scripts, consider the upstream fixes that reduce returns—both fraudulent and legitimate—in the first place.

Pre-Purchase Clarity That Reduces "Didn't Fit" Returns

Many returns stem from unclear expectations, not bad intent. Tightening your pre-purchase experience can significantly reduce the volume of returns your team has to evaluate.

Detailed sizing information goes beyond basic measurements. Include fit notes ("runs small in the shoulders," "relaxed fit through the hips"), comparison to well-known brands ("similar to a Uniqlo medium"), and customer-submitted fit data if you can collect it. Some retailers use simple fit quizzes that recommend sizes based on height, weight, and fit preference.

Product photography and video should show items on multiple body types when possible, include close-ups of fabric texture and weight, and demonstrate how items drape or move. A fifteen-second video of someone walking in a dress tells customers more than ten static photos.

Clear policy visibility at checkout sets expectations before purchase. Display your return window, condition requirements, and any restocking fees on product pages and in the cart—not buried in a footer link. Customers who know your policy upfront are less likely to be surprised (and less likely to file disputes) later.

Technology That Supports Prevention

While you may not need enterprise-grade fraud detection, affordable tools can help flag issues before they become expensive.

Return management platforms like Loop, Returnly, or AfterShip automate return authorization, track return rates by customer, and can enforce rules like "store credit only for customers with return rates above X%." These tools pay for themselves quickly if return volume is meaningful to your business.

Address verification catches some fraud patterns early. Customers using freight forwarders, PO boxes in unusual locations, or mismatched billing/shipping addresses warrant closer attention—not automatic denial, but a flag for human review.

Order velocity alerts help identify customers placing unusually frequent orders or making multiple purchases just before a return window closes on previous orders. This behavior sometimes indicates legitimate enthusiasm, but often signals someone testing your policies.

Building a Return Policy That Protects Without Punishing

Your written return policy is your first line of defense—not because it stops determined bad actors (nothing will), but because it gives your support team clear ground to stand on when they need to enforce limits.

Elements of a Fraud-Resistant Policy

Specific condition requirements matter more than generous timeframes. Instead of "items must be in original condition," try:

"Items must be unworn, unwashed, and unaltered, with all original tags attached and intact. Items showing signs of use—including odors, makeup, deodorant marks, stretched or stressed seams, pet hair, or wear patterns inconsistent with brief try-on—cannot be accepted for return."

This language gives your support team objective criteria rather than subjective judgment calls. When an item arrives smelling like perfume, you're not making an accusation—you're applying documented policy.

Internal exception thresholds should exist but remain unpublished. Your public policy might state "we reserve the right to limit returns for customers with unusual return patterns." What you don't publish is your specific threshold—perhaps flagging accounts with return rates above 40% or more than three "item not received" claims in six months. This prevents customers from gaming published limits while giving you documented criteria for enforcement decisions.

Restocking fees for specific categories discourage bracketing without penalizing genuine returns. A 15% restocking fee on opened electronics or removed-tag apparel signals that returns have real costs—while remaining waivable for manufacturing defects or fulfillment errors on your end.

Photo documentation requirements protect you when patterns emerge. Requiring customers to submit photos of defects before authorizing returns creates evidence you can reference if the same customer repeatedly claims quality issues or if their description doesn't match what arrives.

Policy Language That Reads Like a Human Wrote It

The worst return policies sound like they were drafted by an attorney who's never actually spoken to a customer. Compare these approaches:

Legalistic version:"The Company reserves the right to refuse any return request that does not comply with the stated terms and conditions, including but not limited to items showing evidence of customer use, damage, or alteration from original condition."

Human version:"We want you to love what you buy. If something's wrong, we'll make it right. We do ask that items come back the way they arrived—unworn, unwashed, tags attached. If an item shows signs of use, we may not be able to accept the return, but we'll always work with you to find a fair solution."

Both versions protect you legally. Only one builds the kind of relationship that makes customers want to follow your policy rather than work around it.

Scripts for Common Return Fraud Scenarios

These scripts assume you're dealing with a potentially problematic return but want to resolve it professionally while leaving the door open for customers who might simply be having a bad day or a genuine misunderstanding.

Scenario 1: Suspected Wardrobing (Worn Item Returned)

What you're seeing: A customer requests a return on an apparel item that arrives with obvious signs of wear—makeup on a collar, deodorant residue, stretched elastic, wrinkles in places that suggest it was worn to an event rather than briefly tried on.

What doesn't work:"This item has clearly been worn. Your return is denied."

This response is accusatory, offers no path forward, and will likely escalate to a chargeback or negative review.

What works better:

Hi [Name],Thanks for sending your return back to us. Our team inspected the [item name], and we found some condition issues that prevent us from processing this as a standard return—specifically, [describe what you found: makeup residue on the neckline / stretched seams at the waist / deodorant marks under the arms].We want to be fair, and we know things aren't always straightforward. Here are the options available:If you believe this was a manufacturing defect or quality issue from our end, could you share any photos from when you first received the item? We take quality issues seriously and want to investigate if something was wrong from the start.We can offer store credit for 50% of the item's value, which you could apply toward a future purchase.We're happy to return the item to you at no additional charge if you'd prefer to keep it.Let me know which option works best for you, and we'll get this resolved.

Why this approach works: You've named the specific issue (no ambiguity about what you found), offered multiple paths forward (you're solving a problem, not just saying no), and left room for legitimate explanations. If the customer is genuinely innocent, they'll likely provide photos or accept the partial credit graciously. If they're gaming the system, they'll either disappear or escalate—and you've documented the item's condition either way.

Scenario 2: Serial Return Pattern Detected

What you're seeing: A customer places their eighth order in three months. They've returned part or all of five previous orders. Their return rate is more than double your customer average.

What doesn't work:Processing each return without comment until you're frustrated, then suddenly denying one without context or explanation.

A proactive approach (on their next return request):

Hi [Name],I noticed you've had some trouble finding items that work well for you in your recent orders. That's frustrating, and I want to help.Before we process this return, would you be open to a quick email exchange about what you're looking for? Sometimes our sizing or styles run differently than other brands, and I'd love to help you find something you'll actually want to keep.If our products just aren't the right fit for you, that's completely okay too—we can process this return and you might want to explore other brands that better match your needs. Let me know which direction makes sense.

Why this works: You've acknowledged the pattern without making accusations. You've offered genuine help (some serial returners legitimately struggle with online sizing and would appreciate guidance). And you've signaled—gently but clearly—that the current pattern isn't sustainable. Legitimate customers respond well to this outreach. Customers who are gaming the system usually read between the lines and take their behavior elsewhere.

If the pattern continues after outreach:

Hi [Name],We've processed your recent return. Going forward, returns on your account will be issued as store credit rather than refunds to your original payment method. This helps us keep our return policy sustainable for all customers.We appreciate your understanding, and we're still available if you'd like help finding products that work better for you.

Scenario 3: "Item Not Received" Claim Despite Confirmed Delivery

What you're seeing: A customer claims their package never arrived. Tracking shows it was delivered to their address three days ago.

What doesn't work:"Tracking shows delivered. You'll need to contact your local carrier."

This response abandons the customer, ignores the possibility of real porch theft, and often triggers a chargeback.

A better response:

Hi [Name],I'm sorry to hear your order hasn't turned up. I checked the tracking, and it shows delivered to [address] on [date]. I know that doesn't help much if the package isn't actually in your hands.A few things that sometimes help locate missing packages:Could you check with neighbors or other people at your address? Carriers sometimes leave packages in unexpected spots (side doors, behind planters, with building staff).Does your building have a mail room, package locker, or front desk that might have accepted it?Occasionally packages get scanned as delivered a bit early—if it doesn't show up in the next day or two, let me know.If the package truly isn't there, here's what we can do: For orders under [$X], we'll send a replacement right away. For larger orders, we'll need to open a claim with the carrier first, which typically takes [Y business days] to resolve.Let me know what you find, and we'll figure this out together.

Why this works: You've taken the claim seriously (important for legitimate cases), provided practical troubleshooting steps, and set clear expectations about what happens next. Your tiered response—immediate replacement for smaller orders, carrier investigation for larger ones—protects you from significant losses while still helping customers quickly when the stakes are lower.

Pattern recognition note: Track "item not received" claims by customer. A single claim is often legitimate (porch theft happens). Three or more claims from the same customer in six months warrants closer scrutiny and potentially a policy conversation.

Scenario 4: Requesting an Exception to Published Policy

What you're seeing: A customer wants to return something outside your return window, without original tags, or otherwise beyond your stated policy limits.

For a customer with positive history:

Hi [Name],I understand you'd like to return the [item] even though it's past our standard 30-day window.Looking at your history with us—you've been ordering since [year/time period]—I'd like to make an exception. I can offer store credit for the full value of the item as a one-time accommodation. Would that work for you?For reference, our standard policy is [brief description], but I wanted to find a way to help in this case.

Why this works: You're granting an exception while explicitly framing it as an exception, tied to their positive relationship with your business. This rewards loyalty without creating an expectation that you'll waive policies for everyone.

For a customer without that history or with concerning patterns:

Hi [Name],I wish I could accommodate this return, but the request falls outside our policy in a few ways: [specific reasons—timing, condition, missing tags, etc.].We apply our policies consistently so we can keep prices competitive and provide a good experience for all our customers. I know that's probably not the answer you were hoping for.Is there anything else I can help you with?

Why this works: You've explained the "no" with specific reasons, framed consistency as benefiting all customers (not just convenient for you), and remained professional without over-apologizing or inviting negotiation.

Building Systems to Catch Patterns Before They Cost You

Scripts handle individual conversations. Systems catch patterns before they become expensive problems.

What to Track for Every Customer

At minimum, flag these data points:

If you're using return management software like Loop, Returnly, or AfterShip, most of this tracking happens automatically. If you're managing returns manually, even a simple spreadsheet that logs customer email, return rate, and notes on condition issues will surface patterns within a few months.

Setting Your Thresholds

Your specific thresholds will depend on your product category, margins, and baseline return rates. As a starting framework:

Document everything. If you eventually need to deny a return or close an account, your records should tell a clear, defensible story of the pattern and your attempts to address it.

The Human Review Layer

Automated flags are useful; automated enforcement actions are risky. Every threshold trigger should route to a human who can review context before taking action.

That customer with a 55% return rate? Maybe they've been buying gifts and the recipients are returning them. Maybe they had one terrible experience that skewed their numbers. Maybe they're genuinely abusing your policy. A human can often tell the difference with five minutes of review. An algorithm rarely can.

Training Your Team on Consistent Enforcement

Inconsistency destroys policy credibility. If one support agent grants exceptions freely while another holds firm, customers learn to keep asking until they get the answer they want—and your team burns out making judgment calls without clear guidance.

Build a Decision Framework

For common scenarios, document:

This framework doesn't remove human judgment—it creates guardrails so your team can make decisions confidently and consistently.

Reframe How You Think About "No"

Many support agents struggle to enforce limits because they've internalized "good service means making customers happy." Reframe it: good service means fair treatment for all customers.

When you honor a fraudulent return, you're effectively transferring costs to your honest customers through higher prices, tighter future policies, or reduced investment in product quality. Saying no to abuse isn't failing at customer service—it's protecting the customers who play by the rules.

Practice the Uncomfortable Conversations

Before going live with enforcement approaches, practice them. Have team members role-play the customer who pushes back hard on a wardrobing denial. Work through the awkward pauses. Get comfortable with the language and the silence that sometimes follows a firm "no."

The goal isn't to make these conversations enjoyable—they won't be. The goal is to make them survivable and consistent.

When to Part Ways with a Customer

Some customers cost more to keep than to lose. This is difficult for small businesses to accept when every customer feels precious and your total count is in the hundreds rather than millions.

But consider: a customer who returns 70% of what they buy, escalates every denial, and leaves negative reviews when they don't get exceptions is actively costing you money, time, and mental energy you could spend on customers who value what you sell.

The Account Closure Script

When you've documented a clear pattern and intervention hasn't changed behavior:

Hi [Name],After reviewing your account history, we've made the difficult decision to close your account with us.Over the past [timeframe], we've had [X returns], [Y condition issues], and [Z exception requests]—which unfortunately isn't sustainable for our small business.We've processed a refund for your most recent return, and any remaining store credit has been converted to a refund to your original payment method.We genuinely hope you find a retailer that's a better fit for your needs. This decision is final, and we won't be able to reopen the account.We wish you well.

Send it. Move on. Document the decision and the reasoning. Don't second-guess yourself.

Protecting Margins While Protecting Your Brand

Everything in this guide assumes you want to protect your business while still sounding like a business worth buying from. The scripts avoid accusatory language. They offer alternatives. They treat customers as adults capable of understanding reasonable limits.

That approach isn't weakness—it's strategy. A customer who gets denied fairly and respectfully is far less likely to leave a scorched-earth review than one who feels dismissed or insulted. And the customers watching how you handle problems often become your most loyal buyers.

But there's a limit. Being professional doesn't mean being a pushover. Your return policy exists for a reason. Your thresholds exist for a reason. Enforce them.

Managing returns consistently takes time—especially when patterns emerge that require investigation and documentation. Every wardrobing case you evaluate is an hour you're not spending on product development, marketing, or growth.

If your returns queue is consuming more of your week than it should, it might be time to bring in support. A dedicated team can handle day-to-day enforcement, flag patterns for your review, and maintain the human voice that keeps good customers coming back—while you focus on building the business.

Ready to take returns off your plate? Book a call with Evergreen Support to see how we handle policy enforcement without losing the personal touch your customers expect.

Frequently Asked Questions

What's the difference between wardrobing and a legitimate return?

Wardrobing specifically describes purchasing an item with intent to use it once and return it—wearing a dress to an event, using a tool for a single project, or "borrowing" electronics for a trip. Legitimate returns happen when customers genuinely change their minds, receive defective products, or find items don't meet reasonable expectations. The key indicators are physical signs of use: odors, stains, wear patterns, stretched materials, or missing components inconsistent with brief try-on. Documenting these conditions with photos protects you when declining returns.

How should I respond when a customer threatens a chargeback after I deny their return?

Don't panic—chargebacks aren't automatic wins for customers. Respond calmly, reiterate your documented policy, explain the specific reasons for denial, and state your final alternative (store credit, partial refund, or returning the item to them). If they file a chargeback anyway, submit your documentation: return policy they agreed to at purchase, photos of item condition, complete communication history. Merchants with clear records win these disputes more often than you'll expect. The threat itself shouldn't change your policy enforcement.

At what point should I consider banning a customer for return abuse?

Consider account closure when you've documented a sustained pattern (typically 50%+ return rate across multiple orders), your intervention attempts haven't changed behavior, and the customer continues the pattern despite clear communication about limits. One frustrating interaction doesn't warrant a ban—but ongoing abuse after you've communicated consequences is reasonable grounds for parting ways. Always document your reasoning thoroughly in case of disputes through payment processors or public complaints.

Should I implement restocking fees on all returns?

Blanket restocking fees can deter legitimate customers and hurt conversion rates. Restocking fees work best for specific categories where returns create meaningful costs—electronics requiring testing, personalized or customized items, or products that can't be resold at full price after opening. Consider waiving fees for exchanges or store credit, which keeps revenue while covering handling costs. Test different approaches with your specific customer base and monitor whether fees reduce problematic returns without hurting overall sales.

How do I tell if high return rates indicate fraud or product problems?

Track return reasons by product SKU, not just by customer. If a specific item has disproportionate "not as described" or "didn't fit" returns, your listing, photography, or sizing information might be the issue—not customer behavior. Fraud tends to cluster around customer behaviors (consistently claiming items never arrived, repeatedly returning items in used condition). Product issues cluster around specific SKUs. Your best customers will tell you when something's wrong with the product itself—fraudsters rarely offer that kind of constructive feedback.

Our Approach to Return Policy Guidance

This guide draws on direct experience managing e-commerce support operations where return fraud patterns emerge clearly across thousands of customer interactions. We've handled returns for apparel, electronics, home goods, and specialty products—each with distinct fraud patterns and enforcement challenges. Our approach prioritizes documented, consistent policy application that protects margins while preserving customer relationships. Small e-commerce businesses deserve the same fraud protection strategies that large retailers deploy, adapted for teams without dedicated fraud departments or enterprise software budgets.

Works Cited

[1] National Retail Federation — "2023 Consumer Returns in the Retail Industry." https://nrf.com/research/2023-consumer-returns-retail-industry