16 Retail Returns and Refund Statistics Every Owner Should Know in 2026

We read the National Retail Federation and Happy Returns 2025 Retail Returns Landscape, the NRF release behind its headline figures, the American Customer Satisfaction Index's 2026 holiday-return analysis, and the Census Bureau's latest e-commerce release. The sources show that returns are not only a customer-service cost. They affect conversion, cash timing, inventory, fraud exposure, margin recovery, and the probability that a customer shops with the retailer again.

The figures below keep estimated return value, survey responses, customer satisfaction scores, and quarterly sales data in separate evidence classes. That distinction matters because a returned-sales estimate cannot be used as if it were a government count of individual returned items.

Back to Retail statistics

Retailers expected $849.9 billion in merchandise returns in 2025

The National Retail Federation and Happy Returns estimated that consumers would return $849.9 billion in merchandise during 2025. This is the value of merchandise expected to be returned, not the processing cost, the value permanently lost, or the gross profit destroyed.

NRF and Happy Returns collected retailer perspectives from 358 professionals involved in e-commerce at large U.S. merchants with more than $500 million in revenue. The figure is therefore an industry estimate based on merchant reporting and modeling, not an audited total of every retail return.

An estimated 15.8% of annual retail sales was returned in 2025

The same report says retailers expected 15.8% of their annual sales to be returned in 2025. NRF reported that this was down from 16.9% in 2024, when returns totaled about $890 billion.

The denominator is annual sales value. It is not a unit return rate, an order return rate, or a percentage of profit. A retailer comparing its own result with 15.8% must calculate the same measure: returned sales divided by shipped or completed sales, using the same period and treatment of exchanges.

Online returns were expected to reach 19.3% of online sales in 2025

NRF estimated that 19.3% of online sales would be returned in 2025. The online rate was higher than the 15.8% all-channel annual-sales estimate, which is consistent with the extra uncertainty customers face when they cannot inspect, try on, or compare a product in person before buying.

The figures should not be combined by adding 15.8% and 19.3%. The first is an all-channel sales-value estimate and the second is an online-sales estimate. They use different denominators.

82% of consumers said free returns were an important online shopping consideration

In the NRF and Happy Returns consumer survey, 82% of respondents said free returns were an important consideration when shopping online. That was up from 76% the prior year.

The finding describes purchase consideration, not the percentage of orders that are returned. It shows why a return policy can influence conversion before a customer ever places an order. The policy is part of the product promise, especially in categories where fit, color, size, or product condition is difficult to judge digitally.

76% of consumers were more likely to choose an instant refund or exchange

The NRF survey found that 76% of consumers were more likely to choose a return option that provided an instant refund or exchange. Instant resolution can improve perceived convenience, but it also shifts risk to the retailer if the item has not yet been inspected.

Owners should separate three events in their reporting: the date the customer handed over the item, the date the refund or exchange was authorized, and the date the returned merchandise was inspected and dispositioned. A fast customer refund and a fast inventory recovery are different operating outcomes.

71% of consumers were less likely to shop with a retailer after a poor return experience

About 71% of consumers said they were less likely to shop with a retailer again after a poor returns experience. That was up from 67% in 2024. Four out of five respondents also said they would share a negative experience with friends and family.

This turns a return from a single-order event into a retention and reputation issue. A return policy can be financially generous and still create a poor experience if the customer cannot find the policy, waits too long for a refund, receives an unexpected fee, or cannot get a clear answer about the status of the return.

9% of all returns were estimated to be fraudulent

The NRF and Happy Returns report found that 9% of all returns were fraudulent. Among the retailers tracking specific abuse patterns, 71% reported overstated quantities, 65% reported empty-box or "box of rocks" returns, and 64% reported decoy returns such as counterfeit items.

These are report estimates and merchant-reported abuse categories, not a legal finding about every customer flagged by a retailer. A fraud control should therefore record the reason code, evidence, review decision, and outcome instead of treating every unusual return as confirmed fraud.

45% of shoppers said it was acceptable to bend the rules on returns

Almost half of shoppers, 45%, said it was acceptable to "bend the rules" when returning items. The result shows why return abuse cannot be understood only as organized fraud. Policy ambiguity, generous exceptions, wardrobing, missing packaging, and disputed product condition can all create a gray area between a legitimate return and deliberate abuse.

The operational response should be proportionate. A retailer can keep a clear public policy while using purchase history, product category, return frequency, item condition, and fraud signals to decide when additional verification is justified.

85% of surveyed retailers were using AI to detect or prevent return fraud

The NRF release says 85% of surveyed retailers were employing artificial intelligence to detect or prevent return fraud. It is a technology-adoption figure from the merchant survey, not evidence that AI prevents 85% of fraud.

NRF also reported that nearly two-thirds of merchants, 64%, considered updating their returns process in the next six months a priority. Other planned responses included greater focus on third-party logistics partners at 49%, seasonal returns staff at 43%, and extended return windows at 37%.

Consumers aged 18 to 30 made an average of 7.7 online returns in the prior year

The NRF survey found that consumers aged 18 to 30 made an average of 7.7 returns of online purchases in the previous 12 months, more than any other generation in the survey.

This is a survey average among consumers who answered the study's return questions, not a universal return rate for every young shopper. The useful implication is segmentation: return exposure can vary by customer age, category, purchase behavior, and the number of items ordered per transaction.

Customers who made a holiday return scored seven points lower on ACSI satisfaction

The American Customer Satisfaction Index surveyed 1,962 customers in January 2026 across general merchandise, specialty retail, and online retail about holiday returns. Roughly 21% had made a return. Customers who did not make a return recorded an ACSI satisfaction score of 77 out of 100, compared with 70 for customers who did.

The seven-point gap does not prove that every return causes the entire decline. Customers who return may have started with a harder product or service problem. It does show that the return population is a high-stakes group whose experience deserves separate measurement.

The ACSI return process score was 78, but restocking fees pushed returners to 66

Among the ACSI holiday-return respondents, the return process itself scored 78 out of 100. However, 36% of customers who made returns encountered a restocking fee, and the average ACSI score for that group fell to 66.

The result is a warning against evaluating returns only by speed or internal processing cost. A process can be operationally competent while an unexpected fee, unclear policy, or poor product experience still leaves the customer dissatisfied.

E-commerce represented 16.9% of adjusted retail sales in Q1 2026

The Census Bureau estimated $326.7 billion in seasonally adjusted U.S. e-commerce sales in Q1 2026, equal to 16.9% of adjusted total retail sales of $1,929.0 billion. E-commerce grew 9.8% year over year, compared with 3.9% for total adjusted retail sales.

This is sales-channel context, not a returns statistic. It explains why return policy, refund timing, and reverse logistics increasingly affect a large digital sales base. A retailer should calculate its own online return rate alongside online sales growth rather than inferring returns from market-level e-commerce data.

A return rate is only useful when its denominator is explicit

Retailers commonly use several return measures:

NRF's 15.8% and 19.3% figures are sales-value estimates. They cannot be directly compared with a unit return rate from a store POS system without converting both measures to the same denominator.

A returned dollar is not the same as a recovered dollar

The financial outcome of a return depends on what happens after the refund. A product may be returned to sellable inventory, exchanged for another product, marked down, liquidated, donated, recycled, written off, or lost to fraud. Every path has a different recovery value and cost.

Use a net return contribution calculation:

• Returned sales value = sum of the sales value refunded or exchanged.
• Processing cost = labor, customer support, packaging, inspection, payment fees, and reverse shipping.
• Disposition loss = original expected resale value minus actual recovered value after markdown, liquidation, donation, or write-off.
• Fraud loss = value of the refund or replacement minus any recovered merchandise value for confirmed fraudulent returns.
• Net return contribution = recovered value minus processing cost, disposition loss, fraud loss, and refund-related fees.

An exchange may avoid some lost customer value while creating a different inventory movement. An instant refund may improve the experience while extending the period in which the retailer carries uncertainty about the returned item. These outcomes should be reported separately rather than collapsed into one return count.

Build a return table that keeps estimates, surveys, and actual operations separate

Track returns by reason, channel, condition, and customer outcome

The minimum return dashboard should retain the order ID, SKU, category, channel, purchase date, return initiation date, carrier scan date, refund date, reason code, customer-selected reason, product condition, packaging condition, original price, discount, shipping cost, return shipping cost, inspection result, disposition, fraud review, exchange status, and recovered value.

Break the results out by size or fit, not as described, damaged, wrong item, changed mind, late delivery, duplicate purchase, gift, promotion, and suspected abuse. Then connect each reason to the next customer outcome: repurchase, exchange, refund only, complaint, support contact, and churn.

Use the return dashboard to calculate five operating benchmarks

Start with five local numbers that can be compared over time:

• Return rate by sales value = returned sales value divided by completed sales value.
• Net recovery rate = recovered value after disposition divided by returned value.
• Refund service level = refunds completed within the promised time divided by all refunds.
• Preventable return rate = returns linked to avoidable product information, sizing, picking, packing, or quality issues divided by all returns.
• Repeat purchase after return = customers who purchase again within a defined period after a return divided by customers who returned.

Use the same period, population, and channel in numerator and denominator. Report the NRF estimates as market context, the ACSI scores as customer-experience evidence, and the retailer's own dashboard as the source for operating decisions.

Sources

• NRF: 2025 Retail Returns Landscape
• NRF: Consumers Expected to Return Nearly $850 Billion in Merchandise in 2025
• NRF and Happy Returns: 2025 Retail Returns Landscape report PDF
• ACSI: Every Return Tells a Story Retailers Should Be Reading
• Census Bureau: Quarterly retail e-commerce sales, Q1 2026