The number retailers track hides the one that matters
Returns management is still treated by most finance teams as a single blended percentage carried as a provision line, the same number for every category, rarely revisited. That approach worked when return rates were low and uniform. It stopped working once returns became a category-specific, refund-heavy problem that shows up in the wrong month's numbers.
A report titled "Da Compra à Confiança," released July 28, 2026 by Confi through Neotrust and AfterSale in partnership with E-Commerce Brasil at the Fórum E-Commerce Brasil, put a category breakdown on the table for the first time at this scale. Footwear returns 16-20% of orders, activewear 15-18%, fashion 12-18%, and children's apparel 12-15%. Electronics sit at 2-5%, pet products at 2-4%, and books and stationery at 1-2% [1].
Globally, the picture points the same direction. Retailers responding to the National Retail Federation and Happy Returns expect 15.8% of annual 2025 sales to come back, totaling $849.9 billion, with the online-specific rate closer to 19.3% [2]. Apparel and footwear pull that average up; electronics and consumables pull it down.
Refund and exchange are not the same transaction
Does a return automatically mean lost revenue? No. A return only becomes lost revenue when the business defaults to a refund. When the same request is routed to an exchange or store credit, the merchandise value stays booked and only the logistics cost repeats.
“Returns are no longer the end point of a transaction. They provide an opportunity for retailers to create a positive experience for customers and can translate to brand loyalty.”
In Brazil's H1 2026 data, 58% of approved return requests ended as a refund and 42% converted into store credit [1]. Those are two different financial events attached to the same returned box: one reverses the sale entirely and still charges for the reverse shipment, the other keeps the revenue for the same logistics cost.
The gap between the two outcomes is largely a checkout-flow design choice, not a warehouse constraint.
The same report found that sizing issues, not buyer's remorse, drive most of the volume: 55% of returns trace to size, versus 15% for regret, 12% for defects, 8% for unmet expectations, and 5% each for logistics errors and wrong items shipped [1].
A customer requesting a different size still wants the product. Defaulting that request to a refund treats a retained customer like a lost one.
What Returns Management Metric Actually Predicts Margin Loss?
What should retailers measure instead of a raw return rate? Return revenue retention: the value that stays in the business, through exchanges and store credit, divided by the total value of approved returns in the same period. It isolates the decision a retailer controls from the one it doesn't.
Brazil's national floor sits at 42%, the H1 2026 average [1]. A well-designed post-purchase flow can reach roughly 55%, because that share of returns is size-related and structurally exchangeable [1]. Below the floor, the fix is rarely the customer. It's the order in which options appear on the returns screen and how fast an exchange gets approved.
Order volume alone is a weaker signal, because it blends categories with opposite economics and drops in a slow sales month, making a bad month look deceptively healthy.
What a single return actually costs
Reverse shipping is the visible cost. It is rarely the whole cost. Brazil's H1 2026 report found a returned item can consume up to 65% of its own value once return freight, triage, reprocessing and lost resale value are added together [1]. A $70 item can cost roughly $45 to bring back, more than the operation recovers by reselling it.
External data confirm the multiplier effect. ACI Worldwide's Global Annual Ecommerce Report, covering billions of 2025 transactions and published in February 2026, found that every $1 million refunded carries a total cost to the retailer of about $1.3 million once processing and reverse logistics are added [1]. The cash that leaves the register understates the real hit by roughly 30%.
| Category (Brazil, H1 2026) | Return rate |
|---|---|
| Footwear | 16-20% [1] |
| Activewear | 15-18% [1] |
| Fashion | 12-18% [1] |
| Children's apparel | 12-15% [1] |
| Electronics | 2-5% [1] |
| Pet products | 2-4% [1] |
| Books & stationery | 1-2% [1] |
Why the timing makes it worse
Returns don't arrive evenly across the year. Globally, holiday-season returns had already reached 12.2% of online orders by early January 2026, with nearly half of all returns tied to items customers considered significantly different from what was described [3].
Brazil's ACI-sourced data shows about 20% of a full year's refunds concentrate in November and December alone, with December 2025 running at a 2.89% refund rate against a 2.25% average for the rest of the year, roughly 28% higher [1].
The squeeze compounds because baskets are shrinking while reverse logistics is billed per shipment, not per dollar sold. Brazil's average order value for the semester fell 13.3% to R$ 275,50, with 1.97 items per purchase, while Correios raised domestic parcel rates 4,2643% starting April 12, 2026 [1].
Smaller baskets and pricier return shipments move margin in the same direction at once, a dynamic not unique to Brazil.
Turning refunds into exchanges: the math
A modeled scenario drawn from the report's own figures illustrates the opportunity, without claiming to represent any single retailer. At a 20% revenue retention rate, only 120 approved returns convert into store credit while R$144,000 in merchandise leaves as pure refunds in the same period [1].
Redirecting size-related cases toward instant exchange, with a credit bonus and a 12-month expiration for cases that still choose credit, can lift retention from that 20% floor toward the 50-55% ceiling the report considers structurally achievable for size-driven returns [1]. Closing that gap turns refund leakage into retained revenue at the cost of logistics only, not merchandise value.
What to do before peak season
Retailers running this playbook typically move through four stages inside a single quarter:
The window to act is short and dated: global refund volume already concentrates around the holiday period [3], and Brazil's data show the same seasonal spike arriving in November and December [1]. A returns management flow rebuilt in September operates through peak; one left for December gets built under pressure, during the exact weeks refund volume is highest.
- Audit the last 90 days of approved returns by destination (exchange, credit, refund) and calculate revenue retention in dollars, not order count.
- Cross-reference returns by reason and SKU to isolate the products driving size-related volume, and pause paid media on any item returning at more than double its category average.
- Rebuild the returns screen so exchange for the same item appears first, backed by real-time inventory, ahead of credit and refund.
- Track retention weekly by the month the sale happened, not the month the credit posted, since a return from one month often lands in the next month's numbers.
Sources
- [1]E-Commerce Brasil, Devolução: o vazamento que come a margem do e-commerce em 2026 — ecommercebrasil.com.br
- [2]National Retail Federation, Consumers Expected to Return Nearly $850 Billion in Merchandise in 2025 — nrf.com
- [3]Digital Commerce 360, 2025-2026 holiday-season returns exceed 10% globally — digitalcommerce360.com
