One-click returns ROI: does frictionless actually raise your return rate?

The fear is reasonable. Make returning easier and more people return, so the logic goes, and you have engineered your own margin problem. The peer-reviewed evidence says something more precise and more useful: the dimension of your return policy that drives sales is not the same dimension that drives returns, and one-click sits squarely in the first. This is what the research actually found, what the 2026 retailer data shows about the opposite strategy, and how to model the ROI on your own numbers.

In this article:

• Five dimensions of leniency, and why only scope raises returns
• What happened to fashion brands that added fees: 53% saw returns fall, 44% saw fraud rise
• A seven-step model for your own numbers, including the €3 that never comes back

Why this matters now

Two forces are pulling in opposite directions.

Return costs went up. Handling a single return runs roughly $20 to $30 across shipping, inspection, restocking and support, and for EU-bound cross-border orders the €3 import duty introduced on 1 July 2026 is not recoverable on an ordinary change-of-mind return. Every return that crosses a border now costs strictly more than it did in June.

So retailers started charging. 72% of US retailers now charge a fee for at least one return option, up from 66% a year earlier. Zara charges $4.95, H&M $3.99, J.Crew $7.50, Abercrombie & Fitch $7, and Macy's charges non-members $9.99.

Meanwhile the demand side went the other way. Around 46% of shoppers have abandoned a purchase because they were dissatisfied with the return process or options, and 67% say a negative return experience would put them off buying from that retailer again.

That is the ROI question in one sentence: does removing friction cost you more in extra returns than it earns you in extra orders and retained customers?

What the research actually found

The best available answer is a meta-analysis published in the Journal of Retailing: Janakiraman, Syrdal and Freling, "The Effect of Return Policy Leniency on Consumer Purchase and Return Decisions" (2016, vol. 92 no. 2, pp. 226 to 235). It pooled roughly two dozen academic papers and, critically, broke "leniency" into five separate dimensions rather than treating it as one dial.

• Time: a 60-day window versus 30 days
• Money: full refund versus partial, free return versus paid
• Effort: no forms required versus forms required
• Scope: sale items returnable versus not
• Exchange: cash back versus store credit

The headline result is that leniency increases purchase more than it increases returns. But the dimension-level findings are where the operating decisions live.

Money and effort leniency are what increase purchases. The research team's summary is explicit that the return policy factors that increase purchases are money and effort leniency. One-click returns are effort leniency, near enough by definition.

Time and exchange leniency are what reduce returns. Longer deadlines and store credit both pull return rates down. The explanation offered for the time effect is the endowment effect: the longer someone owns a thing, the more attached to it they become, and the less likely they are to send it back.

Scope leniency is what increases returns. Freling's example: if a sale item is returnable, a shopper reasons that it is a good value and they can always send it back if the colour or fit is wrong.

So the honest answer to the title question is this. The clean, replicated evidence for a higher return rate attaches to scope, not to effort. The clean evidence for a lower return rate attaches to time and exchange. One-click sits in the column that reliably lifts purchases, and the two most dependable levers for suppressing returns are ones you can run alongside it rather than instead of it.

A field experiment referenced in the same literature goes further: Petersen and Kumar found a lenient policy both increased purchases and decreased subsequent returns from those purchases.

Worth a mention Ask Narayan Janakiraman where lenient policies genuinely backfire and he does not talk about clothing. He talks about televisions. During Super Bowl season, a predictable population buys a big-screen TV, watches the game on it, and returns it afterwards. His conclusion is that those retailers should tighten up, because "selling TVs during that time of the year isn't the same as selling a pair of pants." The lesson is not that leniency is dangerous. It is that leniency is category-specific, and the category most brands worry about, apparel, is not the one where it goes wrong.

What happened to the retailers who went the other way

Since 2022 a long list of fashion brands added return fees: Zara, H&M, Boohoo, Uniqlo, Next, ASOS, Urban Outfitters, J.Crew. Boohoo introduced a £1.99 charge in July 2022, following Zara's £1.95 fee for third-party drop-offs. H&M was among the last of the majors to move.

The results, from the NRF and Happy Returns survey of retailers who did it, are genuinely strange, and worth reading twice.

On the upside: 53% reported lower overall return rates, 52% reported higher exchange rates, 51% saw shoppers switch to a different free return method, 43% recouped revenue, 42% saw less return fraud.

On the downside: 47% saw more customer complaints, 37% lost customers, 34% saw lower average order value, 24% saw reduced sales, 44% saw more return fraud, and 37% saw lower exchange rates.

Read those two lists side by side. Return fraud went down for 42% and up for 44%. Exchange rates went up for 52% and down for 37%. The same intervention produced opposite outcomes at roughly the same frequency, which is what you see when the variable being measured is not the variable that matters.

Worth a mention When Boohoo announced its return fee, one of the most-shared customer responses was not about the money at all. It was that the sizing was inconsistent enough that ordering several sizes was the only way to get a fit, so charging for the returns amounted to charging customers for the brand's own problem. That is the whole returns debate compressed into one complaint: a fee taxes the symptom, and the cause sits upstream in product data, sizing and photography.

The dimension nobody prices: what happens after they click

The other half of the ROI is what the return becomes once it is initiated, and this is where the money actually moves.

Loop reports that its merchant network has retained over $843 million in revenue through exchange incentives and upsells, with some brands retaining more than 50% of total return-initiated revenue. The mechanic is simple: make exchanges free and refunds paid, and add bonus credit, typically 5 to 15% above the original value, so a $100 return becomes $110 in store credit.

Notice that this is exchange leniency, the same dimension the 2016 meta-analysis identified as reducing return rates. Two independent lines of evidence, a decade apart, pointing at the same lever.

The margin sensitivity is worth having in front of you when you build the case. The Product Returns Research Group found that for an average company, a five percentage point improvement in return rate can deliver around 200 basis points of net margin. Returns are not a rounding error on the P&L. They are one of the few remaining levers of that size.

How to model the ROI on your own numbers

Do not model this as a single return-rate number. Model it as four separate flows, because they move independently.

Step 1. Establish your baseline honestly. Return rate, cost per return including reverse shipping, inspection, restocking, support contacts and, for cross-border, the unrecoverable import duty. Most brands understate this by ignoring the support and restocking legs.

Step 2. Measure return completion, not just returns received. How many customers start a return and never send the parcel. That population is not a saving. They are holding an unwanted product and will not buy again. If you cannot produce this number, generating it is the highest-value thing you will do this quarter.

Step 3. Estimate the conversion effect separately from the return effect. They are different dimensions with different signs. Removing effort should lift the first without necessarily moving the second. Whether it does in your category is testable with a market-level rollout rather than a guess.

Step 4. Add the exchange lever at the same time. This is what turns the model from break-even to positive. Free exchange, paid refund, bonus credit. Both the academic and the vendor data point the same way here.

Step 5. Consider extending the window rather than shortening it. Counterintuitive, and one of the few return-suppressing moves with peer-reviewed support behind it. A brand shortening its window to control returns may be pulling the lever backwards.

Step 6. Price the border separately. For EU-bound flows, a return that crosses a customs border carries costs a domestic return does not: transit, clearance, re-import admin and a €3 outbound duty that never comes back. A domestic return address changes the cost per return without touching your policy at all, which makes it the cleanest variable in the whole model.

Step 7. Re-measure per market. European return rates run 18 to 22% of orders with some markets between 25 and 40%, and German online fashion sits far above the EU average. A blended number will flatter your weak markets and punish your strong ones.

What this looks like inside the business

The sale that never showed up in the data. Nothing tells you a shopper read the returns page and left. It surfaces as a conversion rate you attribute to price or product. When 46% of shoppers say they have abandoned a purchase over the returns process, that is a large invisible population.

The return fee that worked and the one that did not. Two brands in the same category, same fee, opposite results, as the NRF split shows. The differentiator is almost never the fee. It is whether the underlying reason for returning was addressed, and whether an exchange was made easier than a refund.

The window you shortened to save money. If the endowment effect is real in your category, a shorter window increased your return rate while also making the policy less attractive at checkout. That is the worst available trade, and it is a common one.

The parcel that crossed two borders for a €12 accessory. Reverse logistics on a cross-border return can exceed the value of the item. The one-click experience your customer sees and the route the parcel takes are separate decisions, and only one of them is visible to them.

Worth a mention The single most counterintuitive finding in this literature is that longer return windows lower return rates. Retailers instinctively shorten the window when returns rise, which is exactly backwards according to the meta-analysis. The proposed mechanism is ownership: give someone 60 days with a jacket rather than 14 and they stop treating it as a trial and start treating it as theirs. It is the rare policy change that is simultaneously cheaper, more generous and better for margin.

What to check before you scale

• Your return completion rate, meaning returns initiated versus parcels actually received.
• Your true cost per return, including support contacts and unrecoverable duty on cross-border flows.
• Whether exchanges are currently easier or harder than refunds in your flow. If harder, that is the first fix.
• Your return window, and whether it was shortened at some point to control returns.
• Which returns are driven by sizing, fit or photography, and therefore belong in a product fix rather than a policy fix.
• Return rate by market rather than blended, especially Germany.
• How many borders a returned parcel crosses, and what each crossing costs.
• Whether the €3 duty on outbound EU orders is being tracked as landed cost rather than as a recoverable.

How ShopReturns helps

One-click is the part your customer sees. The cost sits in what happens after the click, and that is the part we run. Local return addresses in nine European markets mean the parcel goes to a domestic address with a carrier the customer already uses, so it never crosses a border on the way back and no customs event occurs. Items are verified in the Wrocław hub within 48 hours with a barcode scan and photo-documented check, then handled the way you decide: back into stock, resold locally, donated or disposed of. Only consolidated bulk travels home, and only when the arithmetic supports it. The effect on the model is direct: the frictionless experience stays, and the cost per return falls.

FAQ

So does one-click raise the return rate or not? The evidence for a higher return rate attaches most clearly to scope leniency, not to effort leniency. Effort leniency is a purchase driver in the research. Expect a modest increase in completed returns, since some of those returns were happening anyway as abandoned customers, alongside a larger increase in purchases.

What is the fastest way to lower a return rate with evidence behind it? Extend the window and make exchanges easier than refunds. Both are supported by the meta-analysis, and the exchange lever is separately supported by merchant data.

Should we charge for returns? The retailer data is genuinely split, with roughly as many reporting worse fraud and worse exchange rates as better. If you charge, the version that performs is a paid refund alongside a free exchange, not a flat fee on everything.

How much of this evidence is European? The meta-analysis is academic and not market-specific. The fee data is largely US, from NRF and Happy Returns. The return-rate ranges and market-level splits are European. The mechanisms travel; the magnitudes should be tested locally.

Does the €3 EU duty change the calculation? Yes, for cross-border flows. It is not refundable on change-of-mind returns, so every EU return that crosses a border now has a fixed cost floor it did not have before July.

Not sure what your current setup is costing you?

Send us your numbers and we'll run them against the new regime - where the duty hits twice, what return freight costs against the goods you recover, and how much is avoidable.