When Returning Beats Keeping: The Math of Free Returns

American retailers processed roughly $890 billion in returns last year, according to the National Retail Federation's 2025 estimate, or about 17.6 percent of all retail sales. That number includes everything from a sweater that didn't fit to a Costco couch that turned out to be the wrong color, and it has been climbing steadily since the e-commerce boom of the early 2020s. Free returns, once a competitive differentiator, are now table stakes — and a meaningful fraction of the apparel and footwear economy is built on the assumption that a non-trivial percentage of what ships will come back.

For shoppers, this changes the math. A return policy is not just a backstop against making a mistake. It is, used carefully, an active tool for getting better outcomes on the things you keep. The question is how to use it without crossing the line that gets you flagged, banned, or stuck eating restocking fees.

This is a guide to the line.

The legitimate strategies

Not all return-adjacent shopping behavior is gray-area. Some of it is explicitly endorsed by the retailers themselves.

Buy-multiple-sizes-return-most. Zappos has, since its founding, openly encouraged shoppers to order multiple sizes of the same shoe and return the ones that don't fit. The reasoning was sound: shoes vary in fit by brand and even by individual style within a brand, and a customer who can iterate quickly is more likely to find a pair they love. The company's no-questions-asked 365-day return window was the foundation of that strategy. Twenty years on, the policy is unchanged, and the operational cost of free returns is just baked into the unit economics. If you order three sizes of a Brooks running shoe from Zappos and return two, you are using the service exactly as designed.

The same is true, with somewhat tighter windows, at Nordstrom (no fixed window, case-by-case, but generally generous), REI (one year for members on most items), and a handful of DTC apparel brands that explicitly market their fit-trial programs.

The 3-trip rule for off-price fashion. Off-price retailers — Macy's, Nordstrom Rack, Saks Off 5th, Bloomingdale's Outlet — operate on a different return cadence than their full-price counterparts. Stock turns weekly. A piece you saw on Tuesday may be markdown-cycled by Sunday. Experienced shoppers in this segment treat the return window as a price-discovery tool: buy the piece at the current price, then check back weekly until either it goes to a deeper markdown (in which case you return and rebuy at the new price, where the policy allows it) or it sells out (in which case your original purchase is the floor). Macy's price-adjustment policy used to make this explicit; in 2026 most of these retailers have moved to a "return and rebuy" workflow rather than automatic adjustments, but the math is the same.

The 3-trip framing is just a heuristic for how often it's worth checking: any more frequently and you're spending more in time and gas than you'll save; any less and you'll miss the window on most markdowns.

Costco's separate-cost-of-goods analysis. Costco's return policy is famous — almost everything, almost any time, almost any reason. The widely circulated stories about people returning years-old televisions and half-eaten roasts are not myths, though Costco has tightened the policy on electronics (90 days) and on cases of repeat abusers. The reason Costco's policy works is that the membership fee, not the markup on goods, funds a meaningful portion of the company's profits. A returned item is a hit to cost of goods sold but does not fundamentally threaten the business model the way it would at a thinner-margin retailer.

For shoppers, this changes how to think about a Costco purchase. The membership pays for itself over the year through grocery and gas savings; the return policy is, on top of that, an insurance policy against buyer's remorse on the bigger-ticket items — the patio set, the air purifier, the kayak. You should still make the purchase decision as if you intended to keep the item. But you should price the safety net at zero, not at "a return-shipping fee plus restocking."

The gray-area strategies

A few patterns are not officially endorsed but are widely practiced and not, by most measures, harmful.

Wearing-once-then-returning. This is "wardrobing," and most retailers have policies against it (typically requiring tags attached and the item unworn). Enforcement varies. A formal dress worn to a wedding and returned the following Monday with the tags retied is, in most stores' eyes, a return — but if you do it twice, you'll likely be flagged. The retailers' return-rate analytics are more sophisticated than they were a decade ago: they track item categories, pattern frequency, and the gap between purchase and return, and they cross-reference across orders.

Our position on wardrobing is straightforward: we don't recommend it, and not for moralistic reasons. The downside risk is being banned from a retailer you actually shop at, and the upside is one event's worth of free clothing rental at the cost of a lot of friction. There are dedicated rental services — Rent the Runway, Nuuly, the various tuxedo programs — that solve the same problem cleanly.

Returning when the price drops. Many retailers' return windows overlap with their markdown cycles. If you buy a coat at full price and it goes on sale 12 days later, and you're inside a 30-day return window, you can return-and-rebuy. Most stores allow this; some explicitly accommodate it as a price adjustment. As long as you're not abusing the pattern, it's a legitimate use of the policy.

Using the return window to test compatibility. A new dishwasher, a new mattress, a new pair of running shoes — these all have meaningful break-in or trial periods. The retailers know this. Mattress retailers in particular have built business models around 100-day trial periods. Use them; that's their purpose.

The strategies that bite

Several patterns reliably get shoppers in trouble.

The 30 percent return rate. Major retailers — Amazon, Best Buy, Sephora, Macy's, and increasingly mid-tier players — track customer return rates in aggregate. The exact thresholds vary and are not publicly disclosed, but reporting in 2024 and 2025 surfaced internal numbers in the 30 to 40 percent return rate range as the trigger for warnings, and substantially above 40 percent for outright bans. A "ban" in this context typically means closure of the customer's account, refusal of future returns, or in extreme cases a refusal of further service entirely.

If you order $1,000 of stuff from Amazon in a month and return $400 of it, you are in the warning zone. If you return $700, you are in the ban zone. The math compounds across categories — a high return rate on apparel is more tolerated than a high return rate on electronics — but the topline rate is what gets tracked.

The "no return label" Amazon refund. Amazon, since around 2023, has increasingly issued refunds without requesting the item back. The shopper sees a message like "Keep the item, refund issued." This feels like a windfall. It is not, contractually, a transfer of ownership of the item to the shopper.

Read the fine print: in most of these cases, the shopper is being asked to dispose of the item, donate it, or destroy it. The shopper does not legally own the returned merchandise. Amazon is making a unit-economic decision that the cost of return logistics exceeds the cost of writing off the item. From a strict legal standpoint, you are now a bailee — holding goods that aren't yours.

In practice, most shoppers keep these items, and Amazon does not pursue them. But two things are true. First, the items often end up in landfills (you didn't want them; that's why you returned them; you're not going to use them now), which is the actual cost of the policy that Amazon is externalizing. Second, repeated "keep the item" refunds are themselves a flagged pattern in Amazon's account-health systems — they correlate with fraud profiles, and accounts that accumulate too many can be restricted.

The honest move, if you find yourself with a "keep the item" refund: donate it promptly to a charity that can use it, or pass it to a friend. You'll feel better about it than you will about it sitting in a closet for a year before getting thrown out.

The restocking-fee surprise. Most retailers have moved away from broad restocking fees on apparel and small goods, but they remain common in two categories: large appliances (15 to 25 percent restocking fees are typical for refrigerators, washers, dryers if returned non-defective) and specialty electronics (cameras, drones, some computer hardware). A "free returns" headline almost always excludes these categories. Read the policy before buying anything in the $400-and-up bracket; the cost of a bad fit is much higher than it is for a $40 sweater.

The "final sale" trap. Final-sale items are non-returnable, full stop. Off-price retailers and sample-sale events use this aggressively. A code that brings a $200 dress down to $40 is great; a code that brings it down to $40 and marks it final-sale is a different proposition. Treat the latter as a $40 commitment, not a $40 purchase.

How to use returns without becoming a flag

A simple framework, after twelve years of watching this dynamic in retail:

Treat the return policy as an insurance product, not a substitute for decision-making. You should still buy with the intent of keeping. The return is a backstop, not a strategy.

Concentrate your "trial" purchases at retailers that explicitly invite the behavior. Zappos and REI absorb returns in stride because their unit economics are built for it. A small DTC brand's return policy may technically be the same, but the human cost of returning a $90 shirt to a small business is not zero, and small brands track patterns.

Watch your overall return rate. If you're buying enough across enough retailers that you're not sure whether you're at 5 percent or 30 percent, you're at risk of triggering a flag at the high end without realizing it. A simple spreadsheet, or just a once-a-quarter mental tally, is enough.

Don't treat "free returns" as the same thing as "no consequences." The shipping is free. The carbon, the labor, the eventual landfill, and your standing with the retailer are all real costs, even when they're not on your invoice. The retailers are getting better at metering those costs back to the shoppers who incur the most of them.

Do treat the return policy as a real benefit when you're making a genuine fit decision — a coat you can't try on in person, shoes from a brand you haven't worn before, a piece of furniture whose dimensions are right but whose color in person might be wrong. That is what the policy is for, and using it for that purpose is exactly the trade the retailer offered.

The shoppers who get the most value out of free returns are not the ones who exploit the policy hardest. They are the ones who use it to make better keep-or-don't-keep decisions on the things they were going to buy anyway. Those shoppers also tend to have good standing with their retailers, which compounds: they get earlier access to private sales, better customer-service treatment when something goes wrong, and the kind of long-term relationship that the retail loyalty programs are actually trying to build. The math, in other words, is on the side of using the policy sparingly enough that you don't become someone the retailer is trying to discourage.