Return Rates by Category: Nineteen Percent Is the Number, Your Category Is the Question

Nineteen percent is the number. Your category is the question.

Retailers expected 19.3 percent of online sales to be returned in 2025, against 15.8 percent of all retail sales, according to the National Retail Federation and Happy Returns. The online rate rose from 17.6 percent in 2024, and total returns were estimated at $849.9 billion, down slightly from $890 billion in 2024. Holiday returns run at 17 percent of holiday sales. Those are the headline numbers, and they are useful in exactly one way: they tell you what normal looks like for the industry as a whole. They do not tell you what normal looks like for your store, because the rate is not one number. It is a different number in every category, and apparel, which has to be tried on, returns at a multiple of the rate for goods that do not.

The mistake is to compare your rate to the headline and either panic or relax. A store at 25 percent is not necessarily broken, and a store at 12 percent is not necessarily healthy, because the categories set the baseline and the mix sets the rest. A store that sells mostly apparel will run hot against the 19.3 percent benchmark and be fine. A store that sells mostly consumables will run cool and still have a problem, if the returns it does get are all the same product with the same defect. The benchmark is the starting point. The question is always what your rate is made of.

How the number is built matters

The NRF number is built from a survey, not from audited totals, and the post should say so, because the method decides what the number means. The 2025 report drew on 2,006 shoppers who had made an online return in the past year and 358 e-commerce professionals at large US merchants with more than $500 million in revenue, surveyed in the summer of 2025. That is a large-merchant sample, and large merchants dominate the dollar figures. A small store's rate will differ, and the difference is not an error. It is the sample.

Rates also cannot be compared across bases, and this is where most benchmark posts go wrong. The NRF figure is a share of sales value. Other studies count parcels, items, or orders, and a rate measured in parcels is not the same as a rate measured in dollars, because a returned order with three items counts once in one method and three times in another. Appriss Retail, which models transaction-level data rather than surveying, puts total returns at 13.21 percent of all retail sales and fraud at 15.14 percent of returns, or $103 billion in 2024, against the NRF's 9 percent. The two studies do not agree, and the disagreement is informative: it tells you the range, and it tells you to pick one method and track it consistently, rather than shopping for the number that flatters you.

What to do with your own rate

Track three splits before you track the headline. First, the rate by reason: fit, quality, damaged, wrong item, and changed mind are different problems with different fixes, and a rate that is mostly fit is a sizing problem, not a returns problem. Second, the rate by product: the worst ten products usually drive a disproportionate share of returns, and fixing one product page can move the whole rate. Third, the rate by channel: online returns run higher than in-store, and a store whose online rate is 21 percent higher than its overall rate is exactly in line with what the NRF found in October 2024. Once those splits exist, the benchmark has a job: it tells you whether your rate is high for your category, and the splits tell you why. Without the splits, the benchmark is just a number to worry about.

Frequently asked questions

What is the average ecommerce return rate?

Retailers expected 19.3 percent of online sales to be returned in 2025, against 15.8 percent of all retail sales, according to the National Retail Federation and Happy Returns.

How much merchandise is returned each year in the US?

An estimated $849.9 billion in 2025, down slightly from $890 billion in 2024. Holiday returns run at 17 percent of holiday sales.

Why do return rate studies disagree?

Different methods and different bases. The NRF surveys large merchants and reports a share of sales value. Appriss Retail models transaction data and reports 13.21 percent of all retail sales, with fraud at 15.14 percent of returns.

Is a 25 percent return rate bad?

Not necessarily. Apparel returns at a multiple of the rate for goods that do not need fitting. Compare your rate to your category, then split it by reason, product, and channel before judging it.