Why Are Fashion Ecommerce Return Rates So High for Multi-Item Orders?

Returns are the quiet margin killer in apparel. Most growth teams put their energy into acquisition cost. Meanwhile, the “return loop” chips away at customer lifetime value before the second order ever happens. That loop means shipping, restocking, and markdown costs on every item sent back.

AdScale looked at its own order database across more than 7,000 clothing and apparel stores. The goal: find out where that leak actually starts. The answer: the number of items in the cart. Single-item apparel orders in AdScale’s dataset come back refunded or partially refunded 6.6% of the time. Orders with four or more items come back 10.2% of the time. The rate climbs in a straight line at every step in between. That pattern held across two separate 12-month windows. That stability is what makes it worth building a strategy around, instead of writing it off as noise.

Key Takeaways

• AdScale’s clothing and apparel data shows a clear climb. Refund rates rise from 6.6% on single-item orders to 10.2% on orders with four or more items.
• The gap between single-item and multi-item refund rates has stayed consistent for two straight years, at roughly 3 percentage points.
• Apparel refunds run far above the eCommerce-wide average of 3.75% across all verticals in AdScale’s database.
• Mobile apparel orders return at a higher rate (10.2%) than desktop orders (8.7%).
• Industry research attributes as much as 70% of fashion returns to fit and sizing problems. That lines up with what bracket buying looks like in practice.

What Is Bracket Buying and Why Do Shoppers Order Multiple Sizes?

Bracket buying is when a shopper orders multiple sizes or styles of the same item. The plan is to keep one and return the rest. It is a rational response to fit uncertainty, not carelessness. A shopper might wear a size 10 in one brand. There’s no reliable way to know what that means in the next brand’s cut. It might translate to an 8, a 10, or a 12. Faced with that gap, a rational shopper orders more than they need. Two sizes, sometimes three, with a plan to send the rest back. It shows up in the data as exactly the pattern AdScale found: order size and return likelihood move together.

Why a Stricter Return Policy Doesn’t Fix It

It is tempting to treat this as a policy problem and respond with shorter return windows or restocking fees. That fixes the symptom without touching the cause. The shopper who orders three sizes is not being careless. They are compensating for a missing piece of information. The product page cannot tell them which size will fit their body in this brand’s specific cut. Tighten the return window and the same uncertainty remains at the moment of purchase. The bracket buying does not stop. It just gets more expensive for the shopper. Some of them stop buying from the brand entirely rather than deal with the friction.

The Hidden Cost in Ad Performance Metrics

It also distorts the numbers a merchant is watching in the first place. A strong ROAS on paper can hide a real problem. If a meaningful share of those “won” orders come back as refunds, the campaign was never that efficient. Most ad platforms report the order at checkout and stay silent on what happens to it afterward. A campaign that looks efficient at the point of sale can look very different once refunds are netted out.

Fashion Return Rate Benchmarks: What AdScale’s Database Shows

AdScale pulled data from its clothing and apparel merchant base: 7,149 deduplicated shops across the “Clothing” and “Apparel & Accessories” verticals. The window was a trailing 12 months, using refunded and partially refunded order status as the return signal. That status field is a reliable proxy for returns, not a perfect match for items physically shipped back. AdScale tracks order status rather than a separate reverse-logistics event. Four data points stand out.

The Core Pattern: Refund Rate Rises With Every Item

That is not a small gap between the extremes. It is a consistent step up at every single stage. That kind of pattern is the signature of real shopper behavior, not a fluke in one segment.

The Pattern Holds Across Two Years

The same comparison against the prior 12-month window tells the same story. Single-item orders refunded at 7.6%, and multi-item orders refunded at 10.9%. That is a gap of roughly 3.3 percentage points, nearly identical to the 3.2-point gap in the more recent period. AdScale treats cross-year stability as the bar for whether a finding is worth publishing, and this one clears it.

Apparel Returns More Than Any Other Vertical

AdScale’s full merchant base spans every vertical it tracks. Over the same 12 months, the average refund rate across all of them was 3.75%. Even single-item apparel orders refund at nearly double that rate. Multi-item apparel orders refund at close to three times the all-vertical average. Fashion is not just a category with returns. It is a category where returns run structurally higher than almost anywhere else in eCommerce.

Mobile Orders Return More Than Desktop

Within the same apparel dataset, mobile orders refunded at 10.2% versus 8.7% for desktop. Mobile now accounts for the majority of apparel order volume. That means the channel driving the most sales is also the channel with the least room for fit confidence. A small screen simply has less size-chart real estate to work with.

What Outside Research Says

This lines up with what is already well documented outside AdScale’s own numbers. Fit and sizing issues are commonly cited as the single largest driver of fashion returns. McKinsey research on the topic puts the figure as high as 70%. That number shows up consistently across industry analyses, even when the exact percentage varies. Academic research on virtual fitting tools backs this up. Letting shoppers visualize fit before buying is linked to higher conversion, higher order values, and fewer fit-related returns. That research tracks with what individual retailers have already built to solve the same problem. Zalando’s AI-driven sizing tool matches a shopper’s photos to garment specifications to recommend a size. Fit-guidance services like True Fit exist specifically to talk shoppers into one confident size, instead of a bracket of three. None of that comes from AdScale’s database, but it explains why AdScale’s own numbers look the way they do.

The Imagination Gap: Why Fit Uncertainty Drives Fashion Returns

Here is the reframe worth sitting with: the return is not the failure. The guess is.

Every multi-item apparel order in this dataset represents the same shopper problem. They could not tell, from the product page alone, which single item to buy. They filled that gap with their wallet instead of information, ordering enough options that one would probably work. Call it the imagination gap. It is the gap between what a shopper can picture for their body and what the page actually shows. A single stock photo on a single model closes almost none of that gap. It doesn’t help anyone who doesn’t happen to share that model’s proportions.

The imagination gap is not solved by a stricter return policy. A stricter policy does not give the shopper any more information at the moment they decide what to order. It just moves the cost of that uncertainty somewhere else, from the retailer’s refund ledger to the shopper’s patience. Patience is the one thing an eCommerce brand cannot get back once it is spent.

How Can Merchants Reduce Fit-Related Fashion Returns?

The practical question for a merchant is not whether fit uncertainty is real, the data settles that. It is what to actually do about it, without overselling any single tactic as a guaranteed cure. A few approaches show up consistently in the research on this problem. They map directly onto what AdScale’s database suggests is happening.

Close the Information Gap on the Product Page

Size charts that reflect the specific garment, not a generic size grid, close part of the gap immediately. A grid built for that specific garment gives a shopper something real to measure against. For example: “Medium fits a 34-36 inch chest for this exact style.” That beats trusting a label that means something different in every brand. Fit-focused customer reviews do similar work. One review that says “I’m 5’6″, 145 lbs, runs slightly small” is worth more than five reviews about fabric quality. Mobile carries the higher return rate in AdScale’s own database. That size and fit information needs to work on a small screen. It can’t just exist somewhere the desktop shopper might scroll to find it.

Close the Expectation Gap in Ad Creative

Ad creative plays a role here too, earlier in the funnel than most merchants think about it. A shopper who arrives at the product page already expecting a certain fit needs less convincing once they land. That is because the ad that brought them there already set realistic expectations. AdScale’s own Agentic Ad Creatives tool exists partly for this reason. Testing multiple creative angles, including ones that show fit and scale more literally, gets far easier with AI-assisted generation. A traditional photoshoot budget can’t stretch across every SKU the same way. That kind of testing matters more than it used to. The Meta Andromeda update raised the bar on how often creative needs to refresh before fatigue sets in.

Steps to Reduce Apparel Return Rates This Month

1. Pull refund rate by item count for your own store. Segment orders into single-item and multi-item buckets and compare refund rates. This is the exact analysis AdScale ran, and it takes one query against order data most merchants already have.
2. Rebuild size charts at the garment level, not the brand level. A single generic size chart across an entire catalog hides the exact inconsistencies that drive bracket buying. Different cuts need different charts.
3. Surface fit-specific reviews near the size selector. Reviews that mention body measurements and the size ordered do more to close the imagination gap than star ratings alone.
4. Audit the mobile size-chart experience specifically. If the desktop chart is a wide table, check whether it collapses into something unreadable on a phone. Mobile is where the higher return rate is showing up.
5. Test ad creative that shows garments at a more literal scale. Use AI creative tools where a full multi-model photoshoot is not in the budget for every SKU.
6. Track refund rate by order size monthly, not just as one aggregate number. An aggregate return rate can look stable while the multi-item segment quietly gets worse.
7. Consider exchange-first flows for multi-item apparel orders. A shopper who bracket-bought two sizes can trade the wrong one for credit or a different size, not a refund. That preserves revenue while still resolving the same underlying uncertainty.

Frequently Asked Questions

The Bottom Line on Fashion Ecommerce Return Rates

A return is not proof that a shopper made a mistake. It is proof that they were guessing, and the store gave them nothing better to work with. AdScale’s own database shows exactly where that guessing concentrates: in the second, third, and fourth item on the order, the sizes and styles a shopper added because the page could not tell them which one they actually needed. Close the imagination gap, and the guessing has less reason to happen in the first place.

Keep Learning

• Clothing Industry Ad Benchmarks: What 100M+ eCommerce Orders Reveal About Google vs. Meta: AOV and platform performance benchmarks across AdScale’s clothing and apparel merchant base.
• Understanding Poor Ad Performance: Why Your ROAS Is Low and How to Fix It: A framework for diagnosing weak ad performance before blaming the platform.
• Agentic Ad Creatives: The AI Ad Creative Generator Built Into AdScale: How AI-generated creative variants make fit-focused ad testing realistic at scale.
• Meta Andromeda Update: New Creative Strategy for Facebook and Instagram Ads: What the Andromeda update means for creative testing frequency.