This month your international orders are up 38 units and you feel good about it. Then the carrier invoices land. Eleven parcels came back refused, because the customer at the door was asked for $14 in import charges and said no. You paid return freight on all eleven — about $41 a parcel, before the outbound label you already burned — and the duties you already settled on those orders are gone. Shopify will not give them back once the package has shipped.
That is the part almost nobody models. International profit is not price minus product cost minus shipping. It is that, minus first-mile, minus the freight you actually pay, minus duties you never recover, minus the odds of refusal times the cost of a refusal. Skip the last term and you are treating risk as zero.
Why does an international order that shipped still lose money?
Three costs fire after the parcel leaves your building, and only one of them shows up anywhere you look.
The first is the freight gap. You quote a rate, the carrier bills the real cross-border rate, and on light parcels to Europe or Canada the gap runs $5 to $9 per order, buried inside "shipping.
The second is duty and tax you prepaid. Under delivered duty paid (DDP), duties settle before the parcel arrives — exactly what stops the doorstep surprise, and exactly why the money has already left your account. If the order comes back, that money does not.
The third is the refusal. A parcel refused at the door clears customs in reverse and the return freight falls on you. You lose the sale, pay to bring the goods home, and usually eat a negative review.
Add those three and a profitable overseas order flips sign without any change in your pricing.
What does a refused parcel actually cost?
Seller-side tracking keeps landing on about $41 per refused parcel in return freight and handling, before the outbound label you already burned. Of 212 UK customers who refused a parcel in one quarter, exactly 9 reordered within six months.
| Cost line | Who eats it | Typical amount |
|---|---|---|
| Outbound cross-border freight | You, already spent | $12–20 |
| Freight quoted vs freight paid | You | $5–9 gap |
| Duty and VAT already settled | You, permanently | $8–25 |
| Return freight and handling | You | ~$41 |
| Goods value if abandoned | You, entirely | full COGS |
That bottom row is what ends businesses. US-to-Canada shipments via DHL eCommerce get returned to the sender, with fees possibly attached. For essentially every other destination, an undeliverable parcel is classified as abandoned. No return, no resale, no salvage.
Honest range: $50 to $90 per refused parcel, depending on whether the goods come back.
Why won't Shopify refund the duties?
Because customs consumed the duty, not Shopify. This is a rule, not a support ticket you argue your way out of.
Shopify's documentation is blunt: once a Managed Markets order has been fulfilled, refunds are not provided for duties, customs fees, or VAT. If you refund the entire order before fulfillment, duties and customs are returned to you on the following payout date. That is the only window where you get the money back.
Markets Pro follows the same logic. Taxes and duties are charged at the moment the package is fulfilled. Refund an order you already fulfilled and shipped, and Markets Pro refunds none of the fees, duties, or taxes.
That changes how you handle returns. The instinctive move — customer wants a refund, give them a refund — silently converts a duty you paid into a permanent loss. Refund policy and duty policy have to be designed together, before the first overseas order ships.
Which markets should you treat as high risk?
The threshold that used to shelter small parcels is gone. Both the EU and the US dropped their duty-free thresholds to zero. A parcel that had a near-zero chance of triggering a customs invoice a few years ago now triggers one routinely. That is why international margins started collapsing this year rather than five years ago: the probability of a customs event moved from exception to default. The old rule of thumb — small parcels slip through — is now actively wrong.
The refusal rate is not uniform, and it is not small. In high-tax markets such as Canada and the UK, refusal rates on duty-unpaid shipments reach 10 to 20 percent. That is one parcel in five to ten, not a rounding error you can absorb.
The mechanism starts earlier than the doorstep: unexpected extra costs drive up to 39 percent of cart abandonment, and the same charge that empties a cart fills a return bin. You pay twice.
So a 10 percent refusal rate is not an edge case. It is a structural cost that belongs in your unit economics as a fixed percentage, the way you already treat a refund rate.
How do you put duty into a calculator with no customs field?
You fold it in. This is where most sellers get stuck, then conclude the tool is broken.
MerchCalc's eight-field Shopify profit calculator takes eight inputs — price, COGS, shipping to customer, payment processing %, app cost per order, ad spend per unit, refund rate %, first-mile shipping — plus two scenario dials for price and COGS. It returns gross profit, gross margin, net profit, net profit margin, and a revenue breakdown.
There is no duty field, no country field, and no channel field. That is a design choice: duty is part of landed cost, so it belongs inside COGS before you type anything.
| What you are modeling | Field it belongs in | What you actually type |
|---|---|---|
| Duty, VAT, brokerage, clearance fees | COGS or Product Cost | One number: goods plus duty, not goods alone |
| Freight you pay above what you charged | Shipping Cost to Customer, or First-Mile | Your real cross-border rate, not the quoted one |
| Return freight on refused parcels | Refund Rate, or COGS | Your observed refusal rate, or annual loss per unit |
| Carrier handling and cross-border surcharges | First-Mile or Inbound Shipping Cost | Total inbound freight divided by units |
| Card fees on the overseas transaction | Payment Processing Fee (%) | The percentage, plus any fixed per-order fee |
The refusal cost is the awkward one. Raise Refund Rate to your observed refusal rate and let the calculator treat it like a refund, or compute your annual refusal loss and add it to COGS per unit. Both work. Leaving it out does not.
What does a refused parcel do to your margin?
Run one order through it.
| Line | Clean order | Refused parcel |
|---|---|---|
| Selling price kept | $64 | $0 — fully refunded |
| COGS, including $14 of duty | −$22 | $0 — goods back on the shelf |
| Duty that no refund restores | inside COGS | −$14 |
| First-mile | −$3 | −$3 |
| Cross-border freight actually paid | −$16 | −$16 |
| Return freight and handling | $0 | −$41 |
| Net per order | +$23 | −$74 |
Now it gets refused. Revenue goes to zero, the goods come back to your shelf, and you still ate the $3 first-mile, the $16 freight, the $41 return, and $14 of duty no refund will ever restore. That is $74 gone on one parcel. If the destination treats it as abandoned instead of returning it, add the $8 of goods cost the duty did not cover: $82.
Against a clean-order contribution of $23, one refusal erases 3.2 clean orders. An abandoned one erases 3.6. Not one. Three-plus.
That reframes the pricing conversation. A 10 percent refusal rate is not a 10 percent loss on those orders — it is three orders' margin, eaten every time it happens. Price freight and duty against a refusal-adjusted margin, not a clean order.
What should you do this week?
Measure your refusal rate. Pull 90 days of international orders, filter undeliverable or refused, divide by total international orders. Without that number you are guessing at a $50–90 cost per event.
Stop refunding duties you cannot recover. Store credit, partial refund, or a flat restocking fee — pick one deliberately.
Charge the freight you actually pay. A $7 gap on every order is a margin leak, not a shipping cost.
Put duty into COGS before you calculate anything. Run it through a Shopify profit calculator only after duty, freight and goods are already combined into landed cost. Anything less is a gross profit number that flatters you.
Move high-risk markets to DDP. Prepaid duty removes the doorstep surprise that triggers refusals. It costs margin per order and saves the $50–90 tail.
Then run it through a calculator that forces you to count refunds and first-mile honestly. That is what MerchCalc's Shopify profit calculator does — eight fields, no fiction. The Etsy and Shopify profit comparison runs the same math across two platforms. All tools live at MerchCalc.
FAQ
Who pays return shipping when a customer refuses a customs charge?
The seller, almost always. The parcel clears customs in reverse and return freight lands on your account. Outbound freight is spent and prepaid duty is not refunded, so a refusal costs both directions plus the tax.
Can I get the duty back if I refund the order?
Only before fulfillment. Shopify's documentation is explicit: after a Managed Markets order is fulfilled, duties, customs fees and VAT are not refunded. Refund before it ships and the duty returns on the next payout; after it ships, it is gone.
Is DDU or DDP better for a small store?
DDP is safer for margin and reviews, because prepaid duty removes the surprise that drives refusals. DDU keeps the sticker price lower and shifts the charge to the customer — workable in low-tax markets, costly in Canada and the UK, where refusal rates reach 10 to 20 percent.
How many refused orders before a market stops being worth it?
Run the margin test: at $23 clean contribution and $74–82 refusal cost, a 10 percent refusal rate eats 32–36 percent of every ten orders' contribution. Past your net margin, the market is unprofitable however good revenue looks.
This calculator and content is for estimation reference only and does not constitute financial advice.
Sources: Shopify Help Center, "Fulfilling and managing Managed Markets orders" and "Payouts" (Markets Pro); Shopify Help Center, "Requirements and considerations for shipping with Managed Markets"; Ecommerce Fastlane, "The Doorstep Bill Quietly Killing Your International Sales," July 2026; monkeyman.agency, "Shopify DDP vs DDU," July 2026; Fulfillmen, "DDP vs DDU," April 2026; Factlen, "Settling Tariffs at Checkout Under DDP," October 2026.