Ecommerce returns turn into profit when returned items get back to sellable stock quickly, not when return labels get cheaper. Most Shopify brands under $20M should fix return reasons and processing speed first; in-person drop-off networks like Happy Returns pay off at several thousand returns a month.
A return label costs less than $10. The sweater inside it cost you $40 to $50 and sells for twice that. Most brands spend their energy squeezing the cheap half of that equation.
When I was a merchant success manager at Shopify, one of the fashion brands I worked with out of Los Angeles had an entire section of its warehouse set aside for returns. A 40-foot container of them showed up every week. Not a pallet. A container. Nobody on that team talked about the returns corner as a profit center. It was where inventory went to wait.
That picture came straight back to me watching Izzy Rosenzweig, founder of Portless, interview Eric Banister on The Modern Supply Chain podcast. Eric runs operations at Happy Returns, the box-free returns network UPS owns, and he spent about a decade in forward logistics before switching sides. His advice to anyone touring a warehouse is the line I keep repeating: “ask to see where returns are.” Even in the best operations he has seen, that is the messiest, most backed up corner of the building. Portless turned the conversation into a post arguing that returns can be a profit driver rather than a cost to minimize, and you can watch the full episode with Eric on YouTube.
I agree with most of it. I also think the “millions on the table” framing lands very differently at $3M than at $100M, and the timing matters. It is October 2. Whatever returns process you have today is the one that absorbs your BFCM returns in late December and January. So here is my take, stage by stage, with the math worked out so you can check it against your own store.
The shipping label is the cheapest part of a return; the expensive part is the item inside it, and whether that item gets back to sellable stock before its price drops. Eric put it plainly in the episode: a label is under $10, while the item is $40 or $50 at wholesale and sells for twice that. Most brands I talk to have negotiated their return label rate down to the cent and have never measured how long a returned unit sits before it can be sold again.
The second hidden cost is receiving. Eric called the way most returns arrive, loose boxes and bags strewn across a trailer floor with no manifest, probably the single biggest cost bucket for retailers and their 3PLs. Someone has to open every parcel, figure out what it is, match it to an order, inspect it, and decide where it goes. That labor is slow in July. In January, with your team stretched and your 3PL backlogged, it is where returned inventory goes to lose value.
How much value? The best data point I found comes from SML’s survey of more than 500 US and UK apparel retailers, run in late 2022. Respondents said 47% of returned items resell at full price, 42% sell at a reduced price with an average markdown of about 38%, and 12% are never resold at all. Two caveats before you lean on those numbers. SML sells RFID tagging, so it has a reason to emphasize slow processing, and the sample mixes store and online returns at larger apparel retailers. Treat the split as directional, not as your benchmark.
Eric’s comparison is the one that stuck with me. A small share of returns that cannot be resold from a warehouse costs far less than routing resellable product to a clearance rack. The goal is not a perfect returns operation. It is getting the sellable majority back on the shelf while it is still worth full price.
Happy Returns is a box-free, label-free returns network that verifies items in person at drop-off, consolidates them through regional hubs, and ships them back to merchants in bulk with an advance shipping notice. The shopper starts a return online, gets a QR code, and walks the unpackaged item into a drop-off point, what Happy Returns calls a Return Bar. Eric puts the network at roughly 10,000 locations, including UPS Stores, Staples, and Ulta Beauty.
The verification step is the part most merchants underrate. The associate scans the original hang tag or barcode if it is there. If it is not, a patented “item puzzle” asks them to pick the correct product from a set of options, only one of which matches the order. Because the item is confirmed at the counter, the refund can start processing at drop-off instead of three weeks later at your warehouse. Returns are then bagged, consolidated, sorted at Happy Returns hubs with robotic automation, and shipped to your warehouse palletized, with every bag carrying a scannable code and the shipment carrying a manifest. Eric’s stated goal is for a pallet of returns to look as close as possible to a clean supplier receipt.
The ownership story matters for how you evaluate it. I first had Happy Returns on the show back in 2022, when PayPal owned it. In October 2023, UPS acquired Happy Returns from PayPal, which gave it UPS’s pickup network and, according to Eric, access to much larger retail accounts. On Shopify specifically, Happy Returns named Loop its preferred returns portal partner in 2024, so most Shopify merchants reach the drop-off network through a returns app rather than directly.
Eric’s own explanation of why the model works is the useful part for anyone evaluating competitors. Software-only returns tools generate a label and stop there. Logistics-only providers move boxes without knowing what is inside them. Happy Returns sits in both seats, which is why it can tell your warehouse exactly what is arriving before the truck does.
Recoverable returns money scales with revenue and return rate, so the “millions on the table” claim is literally true for a $100M apparel brand and closer to tens of thousands of dollars for a $5M one. That is not a knock on the claim. It is a reason to size the opportunity before you buy a solution for it.
Here is a worked example you can rerun. It is illustrative, not benchmarked. Assume a 20% return rate, which sits inside the 15% to 25% range Eric described for growing apparel brands. Apply the SML resale split above. For every dollar of retail value that comes back, you recover about 47 cents at full price and about 26 cents through markdowns, which means roughly 27 cents of every returned dollar is gone. Now assume faster processing moves half of the marked down units back to full price. That recovers about 8 cents per returned dollar.
These are retail values, not profit, and your markdown depth and return rate will differ, so swap in your own. The pattern holds anyway. At $5M, $80K a year is real money, but it is a process fix and a better returns app, not a new logistics partner. At $100M, $1.6M justifies a dedicated reverse logistics strategy.
Eric gave a useful starting point for where Happy Returns fits: 5,000 to 10,000 returns a month, though he said they work with merchants doing a few thousand. Translate that into revenue with illustrative numbers, an $80 average order and a 20% return rate, and 5,000 returns a month means about 25,000 orders a month, or roughly $24M a year. That is my read of where in-person drop-off math starts to work: low eight figures in apparel, not $2M.
Processing speed protects more margin than chasing a lower return rate, because every week an item sits unprocessed pushes it closer to markdown and pushes the customer’s refund further out. Return rate reduction matters, especially through better sizing and product detail, but it is slow work. Cutting your dock-to-stock time from three weeks to five days is an operational change you can make this quarter.
The customer side of the same delay is just as expensive. My wife returned something to a well-known brand and could not understand why it took three weeks to process. That is the experience most returns create by default, and shoppers notice. In the NRF and Happy Returns 2025 Retail Returns Landscape, 76% of consumers said they are more likely to choose a return option that gives an instant refund or exchange, and 71% said a poor returns experience makes them less likely to shop with that retailer again, up from 67% the year before.
Izzy made a point in his third appearance on my show this summer that fits here. He separates returns into two jobs. One is the RMA layer, the portal and labels, where brands use Loop, Narvar, or AfterShip. The other is physical: receiving, photographing, grading, and restocking, which Portless handles through eight return centers around the world for brands shipping internationally. His warning was that the ship-out muscle and the return muscle are different, and brands that bolt returns onto their outbound team are usually the ones with multi-week delays.
My suggestion is simple. Measure one number: days from a return arriving to the item being sellable again. If it is over seven days in a normal month, that is your returns project for Q4. Not a fraud tool. Not a new policy. Speed.
Return fraud is real, but for most Shopify brands it is smaller than the headline number, so measure it before you design a policy that punishes your best customers. The NRF report estimates that 9% of all returns are fraudulent. That figure comes from surveying ecommerce professionals at US merchants with more than $500 million in revenue, and it is a retailer estimate, not a counted rate.
Compare it with what Happy Returns sees in its own network. According to Supply Chain Dive’s reporting on Happy Returns’ AI fraud pilot, less than 1% of Return Bar returns get flagged for review. Flagged items go through Return Vision, which photographs the returned product and compares it with the retailer’s catalog images to catch decoy returns, like a cheaper look-alike dress sent back in place of the expensive one. Audits are completed within a day of the item reaching a hub, and retailers average $218 in prevented loss per flagged return. In-person verification at drop-off is doing most of the work before the AI ever sees anything.
Eric also drew a distinction worth borrowing. True fraud is the empty box or the box of rocks. Policy abuse, like wardrobing, is a customer bending the spirit of your policy, and it is a much harder call. He admitted openly that Happy Returns is still working out the fine line between managing abuse and making returns inconvenient for big shoppers. If the people who run millions of returns a month are still calibrating, a $2M brand should not be bolting on restocking fees and fraud apps before it knows its own numbers.
Start with a report. Tag returns by reason and by customer, then look at the top 1% of returners by count. If they are also your top 1% by lifetime value, you do not have a fraud problem. You have loyal customers who are bracketing sizes, and the fix is better size guidance.
Lock three decisions in October: your holiday return window, your refund timing promise, and who processes returns in the first two weeks of January. Retailers in the NRF survey expect 17% of holiday sales to come back, and the heavy returns period runs from just after Christmas into mid-January. Decisions made in a panic on January 3 are always worse than decisions made on October 2.
The return window is a margin decision disguised as a policy decision. An extended holiday window drives gift purchases, but it also moves returns into the period when your full-price season has ended. Set it deliberately, and give exchanges a longer window than cash refunds if your returns app supports it. The refund promise is the experience decision: say when the customer gets their money, then staff to hit it. The January staffing plan is the cash decision, because every day of backlog is markdown risk on inventory you already paid for.
How far to go depends on your stage. Under $1M, use Shopify’s built-in self-serve returns and return reports, write a clear policy, and fix your top return reason. That is enough. From $1M to $5M, an exchange-first returns app earns its fee, and your 3PL’s returns turnaround should be in writing before December. If your outbound fulfillment is not ready for peak either, ShipBob’s Q4 fulfillment playbook covers what to pressure test. From $5M to $20M in apparel or footwear, run the worked example above with your real numbers and price out a drop-off network through your returns app. Above $20M, push your returns partner for palletized, manifested receipts like Eric described, because receiving labor is where your biggest returns cost hides.
This is the pattern I see constantly at the $500K to $2M stage: premature complexity. Brands add a returns app, a fraud tool, and a drop-off network before they have measured dock-to-stock time once. Measure first. Then buy.
The returns-as-profit argument holds best for US apparel and soft goods brands with real return volume, and it weakens quickly outside that box. Everyone in this conversation has a commercial stake, including the people I am quoting. Happy Returns sells the network. Portless handles physical returns for its clients. The NRF report was co-produced with Happy Returns and surveyed only merchants above $500 million. SML sells RFID. None of that makes them wrong, but it should make you rerun the math with your own numbers before you act on theirs.
Geography is the next limit. The Return Bar network is a US network, and as a Canadian I notice that gap. If you sell heavily into Canada, the UK, or Australia, in-person drop-off is not your near-term lever, and Izzy’s point about graded returns centers abroad matters more.
Category is the third. Brands selling consumables, supplements, or final sale goods have low return rates and little to recover. Eric was candid that beauty is an opportunity Happy Returns is still exploring, partly because a used beauty product can be regulated as hazardous material in transit. If that is your category, the playbook is not built yet.
For transparency: Izzy has been on my show three times and has written for the site, and Portless is not a sponsor. I went deeper on his supply chain model, and where it does not fit, in my July breakdown of Portless and the tariff cliff. I have tried to hold this piece to the same standard.
Calculate returns cost as the resale value you lose, not just the shipping you pay. Start with last year’s returned merchandise at retail value from your Shopify returns report or returns app. Estimate what share resold at full price, what share sold at a markdown and how deep, and what share was never resold. Subtract recovered value from returned value to get your lost resale value. Then add return shipping, receiving labor, and support time. Most merchants find the lost resale number is several times larger than the shipping number, which tells you whether to focus on processing speed or label costs.
Happy Returns is usually worth evaluating once you are handling a few thousand returns a month, which for an apparel brand typically means low eight figures in annual revenue. Its operations lead described 5,000 to 10,000 returns a month as a good starting point, though smaller merchants are served. At an $80 average order and a 20% return rate, 5,000 returns a month is roughly $24M in annual revenue. Below that, an exchange-first returns app and faster in-house processing usually recover more value per dollar spent. Shopify merchants typically reach the drop-off network through a returns portal such as Loop.
Offer free returns during BFCM if your margins can absorb your expected return rate, because shoppers weigh return policies heavily when buying gifts. In the 2025 NRF and Happy Returns survey, 82% of consumers said free returns were a major consideration in a purchase. The safer structure for most Shopify brands is free exchanges or store credit with a modest fee for cash refunds, plus an extended holiday window that you set deliberately. Run the numbers on last year’s Q4 return rate first. If returns already eat most of your contribution margin, free returns will amplify the problem.
Aim to restock sellable returns within about seven days of arrival, and faster during peak season. That target is a practical rule of thumb rather than an industry standard, but the logic is consistent: every week a returned item sits unprocessed moves it closer to markdown and delays the customer’s refund. Measure your current dock-to-stock time first, since most merchants have never tracked it. If you use a 3PL, put a returns turnaround commitment in writing before December, because January backlogs are where seasonal inventory loses the most value.
Return fraud is common enough to monitor but rarely large enough to justify punishing customers before you measure it. The 2025 NRF and Happy Returns report estimates 9% of returns are fraudulent, based on a survey of retailers with more than $500 million in revenue. Inside Happy Returns’ own network, less than 1% of in-person returns are flagged for review after item verification at drop-off. For most Shopify brands, the bigger issue is policy abuse like wardrobing and bracketing sizes, which is better handled through clear policies, size guidance, and reviewing your top returners than through blanket restocking fees.
