Bracketing Returns: Why Shoppers Order Three Sizes to Keep One

A customer orders the same dress in small, medium, and large. They keep one and send two back. Your return rate absorbs the hit. Your reverse logistics team processes three units to sell one. Somewhere, a report gets written about rising apparel returns.

The behavior has a name. Bracketing means ordering the same item in multiple variants, keeping what works and returning the rest. It is standard practice now rather than an edge case. Most apparel brands built their return economics around it without ever naming it.

Stat check: Optoro’s US ecommerce research found that 63% of consumers bought multiple items deliberately. They intended to return some of them. That figure comes from its 2024 State of Returns report. Prevalence varies by study, market, and definition. The research consistently shows bracketing is substantial among online fashion shoppers.

So this guide covers why shoppers bracket and what it costs. It also covers the one metric that separates a real problem from a cost of doing business.

How Common Is Bracketing, Really?

Prevalence figures vary more than most articles admit. Showing the spread beats picking the most dramatic number.

Optoro’s 2024 State of Returns report put bracketing at 63% of US ecommerce consumers. It measured purchasing multiple items with the intention of returning some. A separate study of 1,002 UK fashion shoppers found 62%, with women at 62.7% and men at 62.1%. Other research reports different figures for Gen Z specifically, around 51% in one study and roughly 40% in another. Those results should not be compared directly. The samples, markets, age groups, and question wording all differ. Treat the range as the useful signal and any single percentage as indicative.

Size is the dominant form but not the only one. Around 40% of bracketing shoppers buy multiple sizes of one product. Colors follow at 35%, styles at 24%, and price ranges at 11%.

The category impact is where it becomes commercially serious. Our retail and ecommerce support work sees this pattern across every apparel client. Apparel return rates run roughly 20 to 40%, with fast-fashion segments reaching considerably higher. Research from Radial and Two Boxes surveyed apparel and footwear businesses. Some 56% reported return rates of at least 30%.

Bracketing Versus a Normal Return

The two behave differently enough that treating them as one category obscures both. A normal return is a purchase that did not work out. A bracketed order is a purchase strategy that assumes returns from the start.

Notice the intervention row. A normal return is reduced by making the product page more accurate. A bracketed order is reduced by making the customer confident enough to pick one variant.

Why Shoppers Bracket, According to Shoppers

Here is where most coverage of bracketing goes wrong, and it shapes every decision that follows.

The instinct is to treat bracketing as consumer misbehavior. Shoppers exploiting free returns, gaming generous policies, using your warehouse as a fitting room. That framing leads directly to restrictive policies, return fees and tighter windows. It also happens to be wrong.

Ask shoppers why they do it, and a different picture emerges. Statista data indicates that 48% bracket when sizing options are not clear. A further 36% bracket because they cannot try the item on in a store. Another 26% are unfamiliar with the brand, and 23% are sizing after a weight change.

Read those four reasons together. Not one of them describes someone exploiting a policy. They describe people managing uncertainty that the product page failed to resolve. That reframes the whole problem. Bracketing is not a behavior to police. It is a symptom of an information gap. Retailers can often reduce that gap through better product information, fit guidance, and customer assistance.

Supporting data points in the same direction. DHL research found 54% of shoppers who returned products cited incorrect sizing. Coresight Research puts size and fit at 53% of respondents. That is the leading reason for online apparel returns.

What Bracketing Actually Costs

The obvious cost is the return itself, and that is where most analyses stop. Industry estimates commonly place the all-in processing cost of an ecommerce return between $10 and $65. The range depends on category, handling complexity, and whether the unit needs refurbishment before resale.

On apparel margins, the upper half of that range is not a rounding error. It is a meaningful share of the item price, absorbed on a unit that generated no revenue.

Now apply the bracketing multiplier. A customer who orders three sizes generates one sale and two returns. You paid to ship three units and will pay to receive two back. Then you inspect, repackage and restock both. One transaction, three units of handling, one unit of revenue.

Trace one bracketed order through the operation and the handling becomes visible.

Deliberately no dollar figures here, because the right numbers are yours rather than an industry average. The structure is what matters. One transaction, three units of outbound handling, two units of inbound handling, one unit of revenue.

Timing adds another layer. Returned apparel comes back into a market that has moved, particularly in seasonal categories. That pressure peaks during Q4 and the January returns window. A garment returning in week six of an eight-week season has limited resale life remaining.

The real figure is harder to see than the headline return rate suggests. Moosejaw is an outdoor retailer carrying more than 400 apparel and gear brands. It examined its returns data with True Fit. The result: nearly 15% of returned online purchases traceable to size bracketing. The second half of that case matters more. After implementing size-guidance changes, Moosejaw reduced its bracketing rate by 24% over one year.

That is a useful benchmark precisely because it is measured rather than estimated. Most brands have no equivalent number, which is the subject of the next section.

The Metric Most Brands Never Calculate

Headline return rate tells you very little about bracketing. A 30% return rate could mean widespread bracketing with high satisfaction, or it could mean a product quality problem. The number cannot distinguish between them. The same diagnostic gap appears in electronics, where we cover no-fault-found returns.

Keep rate per bracketed order can. Of the units in a bracketed order, how many survive the return window? That single figure tells you whether bracketing is working as intended or masking something else.

Run the logic. A customer orders three sizes and keeps one. Keep rate is 33%, and the transaction did exactly what the customer intended. It is annoying operationally, but the sale completed. Now suppose that customer orders three sizes and keeps none. Keep rate is zero. The behavior was still bracketing, but the order failed to resolve the customer’s fit or product-selection uncertainty. That points to a fit, quality, or product description problem rather than a policy one.

Many returns dashboards report overall return rate without separately isolating keep rate within bracketed orders. It is the difference between knowing returns are high and knowing why. That changes what you do about it.

Segment it by style, by category, and by size range. A keep rate collapsing on one product line is highly actionable. A return rate rising across the catalog is not.

One further metric belongs beside it. Return-to-exchange conversion rate is the share of eligible return requests converted into an exchange rather than a refund. It measures recovery rather than prevention.

A successful exchange retains the revenue, keeps the customer, and often fixes the sizing problem that caused the return. A refund does none of those things. Tracking both numbers tells you how well you prevent bracketing and how well you recover from it.

Why Your Bracketing Rate Is Probably Wrong

Before acting on any bracketing number, check how it was measured. Two common errors push the figure in opposite directions, and most brands have at least one of them.

Household and gift orders inflate it. Two sizes of the same style in one order can mean two different people rather than one uncertain shopper. A family ordering for two children looks identical to bracketing in the data. Strip out multi-recipient orders where you can identify them.

Split checkouts deflate it. Some customers order a medium, think about it, then order a large in a separate transaction minutes later. Any system measuring bracketing within a single order ID misses that entirely.

The fix for the second error is a configurable cross-order detection window. Something in the 24 to 72 hour range catches most split checkouts. Validate it against your own purchasing patterns rather than adopting a number. Some shoppers reorder within ten minutes, others after reading a review days later. Brands measuring only within a single order ID understate their true bracketing rate, sometimes substantially. That matters because the understated number makes the problem look small. Nobody funds a fix for a problem that looks small.

What Actually Reduces Bracketing

Given that 48% bracket because sizing is unclear, the intervention follows logically. Make sizing clear.

Publish real measurements, not just size labels: A size chart listing S, M, and L tells a shopper nothing they did not already know. Garment measurements in inches or centimeters, by size, give them something to compare against clothing they already own.

Name your fit deliberately: Tell customers whether an item runs small, large, or true to size. Say it on the product page rather than leaving it to reviews. Brand-to-brand sizing inconsistency is one of the largest drivers of bracketing and one of the cheapest to address.

Use model reference points: Height, usual size, and the size worn in the photograph. That gives shoppers a concrete comparison rather than an abstract chart.

Surface fit feedback from reviews: Customers already tell each other whether something runs small. Pulling that signal onto the product page turns scattered comments into a decision aid.

Consider personalized size recommendation where the catalog justifies it: This covers rule-based recommendations, fit quizzes, measurement profiles, purchase-history models, and virtual try-on. The Moosejaw result cited earlier came from exactly this category of intervention. Judge any tool by whether it reduces the customer’s uncertainty rather than whether it increases engagement. Those are different outcomes. Only one of them lowers your return rate.

Policy changes deserve a cautious note. Return fees and shortened windows do reduce bracketing, and they also reduce purchasing. NRF research found 82% of consumers consider free returns important. Separate ICSC research found 67% become less likely to purchase when return fees apply.

So restricting returns can reduce some return behavior while introducing conversion friction. Evaluate that trade-off against your own return economics rather than treating it as a universal rule. It may still be right for some catalogs, as a deliberate decision rather than a reflex.

Where Customer Support Fits

Product page improvements handle the customers who read them. A meaningful share do not, and they contact you instead.

Pre-purchase sizing questions can be among the highest-value contacts an apparel brand receives. The customer is still deciding whether and what to buy. Someone asking whether a jacket runs small is a customer trying not to bracket. Answer accurately, and you convert one sale instead of shipping three units to complete one. Answer vaguely, and they order two sizes to be safe.

Channel matters here. Sizing questions arrive mid-decision, often with the product page open, which makes live chat considerably more effective than email. Our work on live chat support covers why that window is so narrow.

Agent capability matters more than most brands assume. An agent reading the size chart back to a customer adds nothing. An agent who knows the style runs small, and the fabric has no stretch, is different. Knowing most customers size up prevents the bracketed order entirely.

The same conversation works at the return request. A customer returning a medium can often be offered a large rather than a refund. That converts a return into an exchange, which is the single most valuable move in apparel returns. Our fashion and apparel support programs are built around both moments.

Timing is worth planning for too. Bracketed returns concentrate in the same January window as holiday returns. Our analysis of the post-holiday returns wave covers that timing.

Frequently Asked Questions

What is bracketing in ecommerce?

Bracketing means ordering the same item in multiple variants, then keeping one and returning the rest. Size is the most common form at around 40% of bracketing behavior. Colors follow at 35%, styles at 24%, and price ranges at 11%.

How many shoppers bracket?

Estimates vary by methodology. Optoro’s 2024 State of Returns report found 63% of US ecommerce consumers purchased multiple items intending to return some. A UK study of 1,002 fashion shoppers found 62%, split evenly between women and men. Gen Z-specific studies report different figures. Those results should not be compared directly, because the studies use different populations, markets, and definitions.

Why do shoppers bracket?

Mostly because sizing information is inadequate; Statista data shows 48% bracket when sizing options are not clear. A further 36% cannot try items on in a store. Another 26% are unfamiliar with the brand, and 23% follow a weight change.

How much do bracketed returns cost?

Industry estimates place all-in processing costs between $10 and $65 per return, depending on category and handling complexity. A three-size bracketed order means shipping three units out, receiving two back, then inspecting and restocking both. One transaction produces three units of handling and one unit of revenue.

What is the keep rate per bracketed order?

The share of units in a bracketed order that survive the return window. A customer ordering three sizes and keeping one produces a 33% keep rate, which is bracketing working as intended. A keep rate near zero suggests the order failed to resolve the customer’s fit or selection uncertainty.

Why might our bracketing rate be inaccurate?

Two common measurement errors. Household and gift orders inflate it, because two sizes can mean two people. Split checkouts deflate it, because customers ordering a second size in a separate transaction are missed. Use a configurable cross-order window, commonly 24 to 72 hours, validated against your own purchasing patterns.

Do return fees reduce bracketing?

They can reduce some return behavior while introducing conversion friction. NRF research found 82% of consumers consider free returns important. ICSC research found 67% become less likely to purchase when return fees apply. Evaluate the trade-off against your own return economics.

Conclusion: Bracketing Is Feedback, Not Misbehavior

Most retailers treat bracketing as something customers do to them. Free returns get exploited, and warehouses become fitting rooms. The natural response is to tighten policy until the behavior stops. The shopper data does not support that reading. Nearly half bracket because sizing was unclear. A further third had no way to try the item first. Neither group is gaming anything. Both are solving a problem the product page created.

Seen that way, your bracketing rate is a measurement of how much uncertainty your listings leave unresolved. So it is expensive feedback delivered through your reverse logistics operation. It is telling you something specific about your product pages.

So the question worth raising at your next merchandising review is narrow. Of the styles with the highest bracketing rate last quarter, how many carry garment measurements? How many state their fit, or give a model reference? If the answer is few of them do, you already know where to start.