Bracketing Detection: The Pattern Is in the Order, Not the Return
Bracketing is visible in the order, not the return
Bracketing is buying the same item in two or three sizes with the plan to keep one and send the rest back. For the shopper it is rational: sizing varies by brand, and trying at home beats a fitting room. For the retailer it is expensive, because every bracketed order ships two or three items and gets one or two back, and each return costs an estimated $20 to $30 to process. The important fact for the returns desk is that bracketing is visible in the order, not the return. The return itself looks ordinary: an unworn item, in its original packaging, with the tags on. The order is where the pattern lives: multiple sizes of the same item, one kept, the rest returned unworn.
The scale is not small. Fifty-one percent of Gen Z shoppers bracket, according to RefundSentry, which means the behavior is normal shopping for a large part of the customer base, not an edge case to be punished. The NRF's fraud count excludes bracketing, and it should, because bracketing stays inside the policy: the customer bought the items, tried them, and returned the ones that did not fit. That is what the policy allows. The question is not how to stop it. The question is how to see it, and what to do with what you see.
What the data pattern looks like
The pattern has four parts, and all four have to be present before the label fits. First, the order contains multiple sizes or colors of the same item. Second, one is kept and the rest are returned, and the kept item is the one that fits. Third, the returned items are unworn, with tags attached and no inspection finding. Fourth, the pattern repeats across orders from the same customer, concentrated on the same categories. One order with two sizes is a shopper being careful. Ten orders with three sizes each, all in the same category, is a bracketing customer, and the difference matters, because the response to the first is nothing, and the response to the second is a fit intervention, not a punishment.
The pattern also has to be read against the product, not just the customer. If one product is bracketed by everyone, the problem is the product page, not the customers. If one customer brackets everything, the problem is the customer's sizing uncertainty, and the fix is a fit finder or a measurement guide, not a warning letter. The returns data that matters here is not the return rate. It is the bracket rate by product, which is the share of orders for that product that contain multiple sizes, and the bracket rate by customer, which is the share of that customer's orders that do. Those two numbers, tracked together, separate the product problem from the customer problem, and they are the numbers the returns desk should be reporting.
Fix the fit information first
Charging for returns cuts bracketing, and it also cuts first orders, so most brands fix the fit information first. Better fit data on the product page, a fit finder tied to real measurements, and honest reviews that name the sizing are the three moves that work, and they work because they attack the uncertainty that makes bracketing rational. When the cost of being unsure is zero, the shopper moves the uncertainty from the size chart to the returns desk. When the size chart is right, the shopper does not need to. The post should end on that point, because it is the one the returns desk can act on tomorrow: the bracket rate by product is a list of product pages to fix, in order, starting with the worst one.
Related reading
If this topic is useful, these earlier guides go deeper on connected questions: Return Reason Codes: What Changed My Mind Is Hiding, Returned Inventory Grading: A, B, and C Is a Decision, Not a Description, and One Customer, Four Identities: Why Your Return Counts Undercount People.
Frequently asked questions
What is bracketing in ecommerce returns?
Bracketing is buying the same item in two or three sizes with the plan to keep one and return the rest. It is rational for the shopper, because sizing varies by brand, and it is visible in the order, not the return.
How common is bracketing?
Fifty-one percent of Gen Z shoppers bracket, according to RefundSentry. It is normal shopping behavior for a large part of the customer base, not an edge case.
Is bracketing return fraud?
No. Bracketing stays inside the policy: the customer bought the items, tried them, and returned the ones that did not fit. The NRF's fraud count excludes bracketing, and wardrobing, which is different, because the item was used.
How do you reduce bracketing without hurting sales?
Better fit data on the product page, a fit finder tied to real measurements, and honest reviews that name the sizing. Charging for returns cuts bracketing but also cuts first orders.