Returns Data by Size: Spotting the SKUs That Invite Bracketing

Most returns dashboards stop at the product level. This jacket returns at 30 percent, those jeans at 22. Useful, but it hides the real story. Bracketing lives at the size level. Two sizes of the same product can behave like different products entirely, and the average buries the one size that is causing the damage.

A size-level cut takes an afternoon to build and it changes the fix list completely. Instead of a vague goal like reduce returns, you get a short list of specific SKUs and specific sizes, each with an obvious cause and an obvious remedy.

Build the size-level view

Start with three fields per return: the SKU, the size ordered, and whether the same order contained another size of the same SKU. That last flag is the bracketing signal. Aggregate by SKU and size and two patterns jump out: sizes with high return rates, and sizes that appear in multi-size orders far more than their sales share explains.

Add the keep rate per size if you can. A size that ships often in pairs and is kept half the time is a coin flip product. Shoppers are telling you, in behavioral data, that they cannot predict how that size will fit. That is a product information problem, not a customer behavior problem.

The usual suspects

The worst offenders repeat across brands. Waist sizes that sit between standard inseam offerings. Tops where the medium-large boundary falls at a common body size. Shoes in half sizes that run narrow. Dresses where one size covers a four-inch bust range while its neighbors cover two.

New fits are frequent offenders too. A relaunched cut that kept the old size labels inherits the old expectations and breaks them. Shoppers order their usual size plus one neighbor, hedging against a change nobody announced. The size-level data catches this within weeks of launch, long before the product-level average moves.

Match the fix to the signal

A size that brackets because its measurements are missing gets measurements: the actual garment dimensions for that size, on the product page, next to the size selector. A size that brackets because it genuinely fits ambiguously gets a fit note: this style runs small between sizes, most shoppers take the larger one. Specific guidance beats a generic size chart because it answers the exact doubt the data exposed.

Sometimes the fix is the size run itself. If one size covers too wide a range, splitting it, or adding the missing half size, removes the ambiguity at the source. That is a product decision, and the size-level data is what makes the case for it in a range review.

Make it a monthly habit

Bracketing magnets change as the range changes. New launches, fabric swaps, and supplier changes all shift fit quietly. A monthly size-level review, thirty minutes with the same report, catches new offenders while they are still small and confirms whether last month's fixes actually moved the number.

Track the fix, not just the finding. When you add measurements or a fit note to a flagged SKU, note the date and compare its multi-size order rate over the following month. The SKUs that respond prove the method. The ones that do not are telling you the problem is the garment, not the information.

Frequently asked questions

What returns data do you need to spot bracketing by size?

Three fields per order line: the SKU, the size ordered, and a flag for whether the same order held another size of that SKU. Add keep rate per size if your data allows. That is enough to rank SKUs and sizes by bracketing intensity.

Why is product-level returns data not enough?

Averages hide the problem. A product can show a moderate return rate while one of its sizes brackets heavily and the rest behave normally. The fix for a single bad size is cheap and specific, but you cannot see it until you cut the data by size.

What are the most common bracketing magnet SKUs?

Sizes that sit between standards, sizes covering a wider measurement range than their neighbors, half sizes that run narrow, and relaunched fits that kept old size labels. New fits in their first weeks are frequent offenders as shoppers hedge against unannounced changes.

How often should you review size-level returns?

Monthly works for most brands. Fit shifts with new launches, fabric changes, and supplier changes, so a standing monthly review catches new bracketing magnets early and verifies that previous fixes, like added measurements or fit notes, actually worked.