Amazon taught shoppers to expect free, frictionless returns. For a DTC brand competing against that, matching it feels like table stakes. The math doesn't agree.
What a single return actually costs
Most Shopify dashboards log a return as one number: the refund. The real P&L impact is five or six line items stacked on top of each other.
Direct costs
| Cost | Typical range | Notes |
|---|---|---|
| Return shipping label | $5 to $12 | More for heavy or oversized items |
| Receiving and inspection | $3 to $8 | Labor, handling time |
| Quality check | $2 to $5 | Damage, completeness, missing tags |
| Repackaging, if resalable | $1 to $4 | New tags, polybags, boxes |
| Payment processing fees | 2.9% + $0.30 | Not recovered on the refund |
Indirect costs
| Cost | Typical range | Notes |
|---|---|---|
| Inventory depreciation | 10% to 50% of original price | Open-box pricing or liquidation |
| Customer service time | $5 to $15 per interaction | Most returns trigger at least one ticket |
| Lost attach revenue | Varies | The accessories stay sold, but the anchor item is gone |
| CAC write-off | $15 to $50 | Spent to acquire a customer with zero LTV |
On a real order, the stack looks like this. A $75 apparel item, returned, reshelved as "final sale" at 40% off:
- Return shipping: $7
- Processing: $6
- Inventory loss: $30
- Original CAC: $22
- Payment fees: $2.48
That's a $67 loss on a $75 order. Not a refund. A net-negative transaction.
When free returns pay for themselves
They aren't universally bad. Under the right conditions the math works.
If your AOV is $200+ and returns run under 15%, free returns act as cheap conversion insurance. The margin absorbs it. The same logic applies when LTV is high: if repeat customers spend 3 to 5x their first purchase over a year, eating shipping on order one is just acquisition cost. The trap is assuming returners will repeat. Many don't.
Consumables, home goods, and personalized products return at under 10%. A free-return badge on the product page is mostly a trust signal at that rate.
And in some categories the norm is already set. Eyewear, mattresses, shoes. Fighting that expectation usually costs more in lost conversion than the returns themselves.
When free returns eat margin
Apparel is the obvious one. E-commerce apparel returns at 20% to 30%, and trend-driven or occasion pieces push past 40%. Returned items often resell at half price, if they resell at all.
Low-margin SKUs are the second trap. Below 40% product margin, a 15% return rate is usually enough to put you underwater after the full cost stack above.
Bracketing categories (shoes, jeans, jewelry) return at 50%+ of bracketed orders by design. The customer orders three sizes knowing they'll return two.
And anything with strong resale value (electronics, luxury, limited drops) attracts disproportionate fraud on top of the legitimate returns.
What actually happens to conversion when you add friction
The fear is that the moment you charge for returns, conversion collapses. That's not what the data shows.
Small moves rarely cost much:
- $5 to $7 flat return fee: 2% to 5% conversion drop, net-positive for most merchants
- Free returns but only for exchanges: almost no measurable drop when framed well
- Restocking fees on opened items: deters wardrobing, barely touches legit buyers
Big moves cost real conversion:
- Killing returns entirely: 15% to 25% drop (don't)
- Paid returns in a category where free is the norm: 8% to 12% (usually not worth it)
- Hiding the return policy in fine print: worse for trust than just saying "returns cost $6"
The biggest lever isn't the fee itself, it's the language. "Store-credit returns are free. Refunds incur a $6 processing fee" reads as fair. The same policy buried three clicks deep reads as hostile.
Strategies that work
Tier the policy by customer
Not every customer deserves the same experience. A first-time buyer gets the standard policy with a restocking fee. A repeat customer on their second or third order gets free returns. Your top 10% by LTV gets prepaid labels, no questions, instant processing.
That needs a system that actually tags customers by risk and loyalty, which Shopify doesn't do out of the box.
Make exchanges the easy path
Exchanges preserve revenue and often lead to a higher-value replacement. Price the policy accordingly: free shipping on exchanges but paid on refunds, a 5% to 10% bonus credit for picking exchange over refund, and instant exchange issuance instead of a multi-day refund wait.
Prevent returns before checkout
The cheapest return is the one that doesn't happen. Size guides with real measurements, fit-finder quizzes, 360° photography, reviews with fit notes, and pre-purchase chat all cut post-purchase regret. None of it is glamorous, all of it compounds.
Screen high-risk returns
A first-purchase return from a brand-new account is not the same as a loyal customer's third return this year. Require photos on "item not as described" claims. Hold refunds on high-risk flags until the item is inspected. Flag empty-box patterns early, not after the third one.
Use restocking fees as a nudge, not a revenue stream
A $5 to $10 restocking fee won't cover your real processing cost. That's fine. Its job is to make a casual returner pause. Frame it as transparent cost-sharing rather than a penalty.
Measuring your return economics
Before changing anything, know the baseline.
| Metric | How to calculate | Benchmark |
|---|---|---|
| Return rate | Returns / orders | 15% to 30% apparel, 5% to 10% hard goods (see NRF 2024 returns report) |
| Return cost ratio | Total return costs / GMV | Keep under 5% |
| Returner repeat rate | Returners who buy again / total returners | 30%+ makes returns worthwhile |
| Fraud and abuse rate | Fraudulent returns / total returns | NRF estimates ~13.7% of returns industry-wide |
Run a cohort analysis. Compare LTV of customers who return with customers who don't. If returners match or beat non-returners, free returns are probably paying for themselves. If they're materially lower, your policy is selecting for low-value customers.
A worked example
What this looks like on a hypothetical $3M GMV apparel brand starting from a free-returns-no-questions baseline. Use this as a template for modeling your own numbers, not as data from a specific store.
Starting position:
- 28% return rate
- Return cost ratio: 7.2% of GMV
- Returner repeat rate: 22%
Policy changes layered in over six months: free exchanges, refund processing fee around $6 to $7, 10% exchange credit bonus, VIP restocking-fee waiver at 3+ orders.
Modeled outcome:
- Return rate: 24%
- Share of returns that became exchanges: 35%, up from 15%
- Return cost ratio: 4.1% of GMV
- Returner repeat rate: 31%
On a $3M GMV business, that's roughly $180K in annual margin recovered. Your numbers will move differently depending on category, AOV, and starting return rate. Run the math on a six-month look-back from your own Shopify export before changing policy.
Where RefundSentry fits
Tiered return policies need real intelligence about who's who. RefundSentry scores every return 0 to 100 across 60+ fraud and abuse signals, tags customers into risk zones that sync to Shopify segments, and flags wardrobing, bracketing, and empty-box patterns automatically. Generous where it earns loyalty, strict where it's being exploited, without manually reviewing every return.
Takeaways
Free returns typically cost $30 to $60 per incident, not the $7 shipping label. The right policy depends on your category, margin, and LTV mix. Tiered beats flat. Exchanges beat refunds. The brands winning in 2026 aren't the ones with the most generous policies, they're the ones with the smartest ones.