Bracketing: When Shoppers Order to Return

A customer orders the same dress in a small, a medium, and a large, keeps the one that fits, and sends the other two back. It is tempting to log that as a return problem, or worse, as abuse. It is neither. It is bracketing, and it is the single most rational thing a shopper can do when they cannot try the item on before buying and returns are free. The customer is not gaming you; they are solving a sizing problem you created by selling clothing online without a fitting room. Misreading bracketing as fraud leads merchants to reach for punitive fixes — fees, blocks, shortened windows — that punish exactly the high-intent buyers who were always going to keep one item. The better move is to measure it precisely, understand its economics, and attack the root cause, which is sizing uncertainty.

What bracketing actually is

Bracketing is the practice of ordering multiple variants of the same product — usually adjacent sizes, sometimes colors — with the pre-formed intention of keeping one and returning the rest. It is concentrated in apparel and footwear, the categories where fit is hardest to predict from a photo and a size chart, and it has grown in lockstep with the normalization of free, easy returns. The behavior is a direct, predictable response to two facts: the shopper cannot feel the fabric or try the garment on, and the cost of guessing wrong has been reduced to near zero. Remove either fact and bracketing shrinks. That is the whole mechanism, and it is why the instinct to treat bracketers as bad actors is misplaced.

Bracketing is also distinct from the behaviors that genuinely are abuse. Wardrobing — wearing an item to an event and returning it — and serial-return fraud both involve using the product and then extracting a refund. Bracketing involves buying more than you will keep and returning the unworn surplus in sellable condition. The returned units come back new, restockable, and resellable at full price. Conflating the two leads to policies that treat a customer who bought three sizes to find the right one the same as a customer who wore a jacket to a wedding, and those two people deserve very different responses.

How to measure it before you react to it

You cannot manage bracketing until you can see it in your data separately from ordinary returns, and most analytics setups blur the two. The tell is not the return itself; it is the shape of the order that produced it. A bracketed order contains the same style in more than one size, and it resolves with a high single-keep rate — one unit retained, the rest returned in condition. Instrument for that pattern directly rather than inferring it from your blended return rate, and the four signals below will separate deliberate bracketing from genuine dissatisfaction.

The economics: not automatically bad

Here is the part that gets lost when bracketing is filed under 'return problem.' Every bracketed order contains a guaranteed sale — the kept item — that might not have happened at all if the shopper had been too unsure of the size to buy. Bracketing converts hesitation into a purchase. The cost is the reverse logistics on the returned surplus: two extra shipping legs, the handling to inspect and restock, and the working capital tied up between purchase and refund. Whether bracketing is a net positive or a net drain depends entirely on whether that keep-one sale would have happened anyway and on how cheaply you process the returned units. This is the same calculus behind the true cost of free returns: free returns manufacture incremental orders and manufacture incremental return cost at the same time, and the only question that matters is the ratio between them. UX research groups such as Baymard Institute have documented for years how much sizing uncertainty suppresses apparel conversion, which is the flip side of the same coin — the uncertainty that drives bracketing also blocks purchases outright when returns feel risky.

Bracketing is not a customer abusing your policy. It is a customer paying return-shipping tax to solve a sizing problem you have not solved for them.

Mitigations that don't punish good customers

Because bracketing is a rational response to sizing uncertainty, the durable fixes reduce the uncertainty rather than penalize the response. The highest-leverage move is better fit guidance at the point of decision: accurate, product-specific size charts, fit notes drawn from real return data, and a storefront size recommendation that tells a shopper which single size to order instead of leaving them to hedge across three. When a customer trusts the recommendation, they order one size, and the bracket collapses to a single kept unit. Feeding your fit-related return reasons back into the size chart closes the loop — every 'too small on this style' return becomes data that sharpens the next shopper's recommendation. Punitive levers, by contrast, tend to backfire: a restocking fee on bracketed orders taxes your most engaged apparel buyers, and a hard block on multi-size orders simply pushes the same customer to place three separate orders or shop a competitor with a friendlier policy.

A softer class of incentive can also nudge shoppers to pre-commit without punishing anyone. Small perks for ordering a single size — free expedited exchange if the first pick is wrong, a modest loyalty credit for a one-item order that is kept — reward confidence instead of taxing caution. This is the posture ResReturn is built around: rather than fighting bracketing at the policy layer, it captures structured return reasons on every returned unit, surfaces the fit signal in returns intelligence so merchandising can fix the size chart or the pattern, and makes the exchange path frictionless so a mis-sized keep converts to the right size instead of a refund. You end up shrinking the bracket at its source — sizing doubt — while keeping the return experience good for the customers who were always going to buy just one.

• Measure bracketing directly: track duplicate-style, multi-size orders and their single-keep rate, separately from your blended return rate.
• Separate bracketing from wardrobing and fraud in your reporting, because unworn, restockable surplus is a different problem than used-and-returned goods.
• Attack the root cause with product-specific size charts, fit notes from return data, and a size recommendation that names one size to order.
• Feed fit-tagged return reasons back into merchandising so recurring size misses on a style get fixed at the pattern, not the policy.
• Reward pre-commitment with perks for single-size orders rather than punishing multi-size orders with fees that hit your best apparel buyers.