If you ship into the EU from outside it, the story of this summer is not that a €3 customs charge exists. It’s that a €3 charge has quietly changed the economics of the parcel-by-parcel import model you may have built your business on (and dont think that bulk order import will be the longterm solution, first checks have been reported already). Two months of data are in, and the clearest effect is not simply higher costs. It’s that direct-from-Asia parcel fulfilment got relatively less attractive and EU-local inventory got relatively more attractive, right before the quarter that decides your year.

This piece walks through what the rule actually did, what the July and August data showed once the dust settled, where the real margin damage sits, and how to decide, SKU by SKU, whether direct import still makes money for you in Q4.

What actually changed on 1 July

On 1 July 2026 the EU removed the €150 customs-duty relief threshold and replaced it with a temporary flat duty of €3 per item, running until 1 July 2028 while the full Customs Reform is built.

Three points decide how it hits you:

  1. It applies to B2C distance sales of goods in consignments with an intrinsic value up to €150, shipped into the EU from outside it. B2B shipments to VAT-registered importers keep standard duty rates instead.
  2. “Item” means a distinct tariff line, not a physical unit. The Commission’s own worked example is a parcel holding one silk blouse and two wool blouses: two tariff sub-headings, so €6, not €9. Five identical T-shirts are one line and €3.
  3. It applies whether or not you use IOSS. The one exception worth checking is preferential-origin goods under a free trade agreement, which can keep their reduced rate only if they are not sold under IOSS and are declared on a standard H1 declaration. Sold under IOSS, the same goods pay the €3. Test your specific customs flow rather than assume every sub-€150 consignment is treated identically.

What the data showed in July and August

The headline reaction was a sharp fall in China and Hong Kong air freight into Europe. But the numbers being quoted around this measure a handful of different things, and mixing them is how sellers end up planning against the wrong risk. Here is what each indicator actually says.
• Ecommerce imports into Europe (volume), down about 24% in July versus June, The direct-parcel channel shrank fast.

• China/HK to Europe air tonnage, July: Hong Kong down 19% on June and 24% year on year; mainland China down 3% and 6%. Week of 17-23 Aug: up 1% week on week, the first weekly rise since early June, demand may be finding a floor at a lower level.

• Direct China/HK to Europe freighter capacity, about 28% below June, concentrated at ecommerce gateways (Madrid down 78%, Budapest 58%, Ürümqi 72%) The cheap dedicated lanes you relied on thinned out.

• Total Asia Pacific to Europe freighter capacity, about 18% higher than late June, not a corridor-wide shortage, a redeployment.

• China-Europe spot rates, fell roughly 30% from the late-June peak of $5.43/kg to early-August lows, then rose three weeks running to $4.14/kg: about 24% below the June peak, but 13% above the same week last year, cheap freight was temporary capacity seems to have shifted to different routes.

Sources: Rotate via Air Cargo News; Rotate via The Loadstar; WorldACD via Air Cargo News.

Read together, these do not describe a capacity crunch. They describe a repricing. Dedicated China and Hong Kong ecommerce capacity has been withdrawn or redeployed while capacity elsewhere in Asia-Europe has grown, so total corridor capacity actually rose even as the specific routes low-value ecommerce depended on thinned out and got dearer. Rates have rebounded from their late-July lows, though current spot prices are still below June’s pre-deadline peak. So the Q4 risk is not today’s freight price. If demand rebounds before that dedicated capacity returns, rates could tighten unusually quickly. Plan for a rising trend and tighter dedicated capacity, not for empty aircraft.

The margin problem you should not model as “just €3”

The €3 looks trivial per order. It stops looking trivial the moment you discount, and it stops looking recoverable the moment a customer returns something.
Discounting is the first trap, because the duty is fixed while your price is not. Cut a €30 item to €15 for Black Friday and the customs cost does not move, so it takes twice the share of a thinner margin. Multi-item bundles compound it: a gift set can generate a separate €3 for each distinct tariff line inside it, so a four-category set can carry up to €12 in duty on a single order. Some genuine retail sets qualify for one tariff heading, so the real figure depends on classification. Check that before you build a Black Friday gift range, not after.

Returns are the second trap, and the more damaging one. Once a parcel clears customs, the €3 does not come back on an ordinary change-of-mind return. Ireland’s Revenue states it plainly: from 1 July, customs duty on these consignments is refunded only where the goods are faulty, not where the customer simply changes their mind. The EU has also removed the simplified declaration-invalidation route that distance-sale returns used to rely on. For ordinary change-of-mind returns, treat the €3 as irrecoverable. Faulty or non-conforming goods are different and can qualify for repayment.

So stop modelling this against gross margin and revenue percentages. Model it against contribution margin, and model three scenarios, because who pays the duty changes everything:

• You absorb it: margin compression.
• The customer pays it at checkout or the door: conversion and AOV pressure, plus refused parcels.
• A marketplace subsidises it: commission and contribution pressure.

Then put two lines in the spreadsheet:
Duty per order = €3 × distinct tariff lines per consignment
Customs-duty burden per retained sale = duty per order ÷ (1 − non-refundable return rate)

Take a fashion retailer averaging 1.4 tariff lines per order, so €4.20 of average duty per order. Across 100 orders that is €420 of customs duty. If 30 of those orders come back as non-refundable change-of-mind returns, the same €420 is ultimately supported by 70 retained sales: €6.00 of customs duty per retained sale, before reverse postage, inspection and markdown. That is the number that decides whether a promotion makes money, not “€3 is 20% of €15”.

Check your VAT route before you finalise landed cost

One technical detail changes the landed cost and catches sellers out. Under IOSS, no import VAT is due on the €3 and it stays out of the taxable amount, because VAT is settled at the point of sale before the duty arises. Under the Special Arrangements or the standard import procedure, the €3 forms part of the import VAT taxable amount, so VAT is charged on the duty itself (European Commission VAT addendum). Same duty, different landed cost depending on how your goods clear. One more point for anyone modelling the future handling fee: that fee is outside the scope of VAT, so if you build a scenario around it, do not add VAT on top of it.

When EU inventory wins, and when it doesn’t

The structure of the reform itself points towards more EU-local fulfilment. It is built to make bulk flows through supervised EU customs warehouses relatively more attractive than millions of individual direct-to-consumer parcels, which is the whole rationale for the lighter treatment the reform proposes for warehouse-based trusted traders.

That does not mean everyone should move everything. For an €80 electronics SKU with a 3% return rate, direct shipment can stay perfectly rational. For a €19 fashion accessory with a 35% return rate and several tariff headings per basket, the case for direct shipment becomes much harder to justify. The right output is a decision rule, not a blanket answer.

Segment your SKUs by average order value, contribution margin, tariff lines per basket, return rate, weekly EU order velocity, direct-air freight cost and EU warehousing cost. Then apply one test:

Localise the SKUs where the avoided parcel duty, avoided air-freight premium, conversion gain from faster delivery and lower return friction together exceed the cost of holding and fulfilling that stock inside the EU.

Everything else can stay direct, at least until the temporary regime ends in 2028 and the next customs framework changes the maths again.

Your Q4 audit, in eight steps

  • Export your last 90 days of EU orders and flag every one currently fulfilled from outside the EU.
  • Map a commodity code to every SKU and calculate your average chargeable tariff lines per basket.
  • Recalculate contribution margin at Black Friday prices, including €3 × tariff lines, outbound freight, return rate and reverse-logistics cost.
  • Rank the 20 SKUs producing the largest customs-margin leakage. That is your shortlist, not your whole catalogue.
  • Quote EU-local fulfilment for those SKUs only and compare against the leakage
  • Stress-test freight at current capacity and rates, not June’s. The dedicated lanes are thinner, and rates have already turned upward from their August lows.
  • Confirm your 1 November product-identifier data fields with your carrier or customs broker.
  • Run a no-fee, a €2-per-item, and a higher-fee downside scenario for the coming handling fee, rather than budgeting a single number. Only assume the reduced €0.50 rate if you actually qualify as a Trust and Check trader operating a customs warehouse for distance sales.

One confirmed November change, and one still pending

One requirement is confirmed for 1 November. A second measure could follow from November, but its final terms are still pending. Both land about three weeks before Black Friday.

Product Identifiers: confirmed. From 1 November 2026, PIDs become mandatory for goods sold via imported distance sales, and can already be declared voluntarily since 1 July. Get your data fields confirmed in September, not October.

The handling fee: planned, not fixed. A separate Union handling fee is expected to apply no earlier than November 2026, with the amount to be set by Commission delegated act. Earlier Commission and Parliament proposals envisaged a €2 fee per item, with a lower €0.50 rate reserved for certified Trust and Check traders operating a customs warehouse for distance sales, a much narrower case than simply holding stock in the EU. Treat none of these figures as final.

Frequently asked questions

Does the €3 duty apply to UK-to-EU orders?
The €3 applies to goods entering the EU from outside it, which includes shipments from Great Britain. The UK runs a separate £135 low-value import regime that is unaffected by the EU change, so if you sell into both markets you manage each separately.

What if a parcel has several SKUs under the same tariff heading?
They count as one item. The €3 is charged per distinct tariff line, so five units of the same classification carry one charge, while five different classifications carry five.

Does IOSS collect the €3 duty?
No. IOSS handles import VAT at the point of sale. The €3 is a separate customs duty accounted for by the declarant, and it applies whether or not you are registered for IOSS.

Do EU-warehouse shipments pay the €3 duty?
Not on the consumer order. Holding stock in the EU does not make customs disappear, the inventory still enters the EU, usually in bulk, and may carry standard duty and import VAT at that point. What changes is that the outbound order to the customer is no longer an individual low-value import triggering the €3 parcel regime, and a return no longer crosses the border a second time.

Are preferential-origin goods treated differently?
They can be. FTA-origin goods can keep their preferential duty rate, but generally only if they are not sold under IOSS and are declared on a standard H1 declaration. Under IOSS, the €3 flat duty applies instead. Confirm which path your shipments actually take.

The winners this peak season will not be the merchants with the lowest freight rate. They will be the ones who know, SKU by SKU, where the direct-import model still makes money and where it no longer does. Working that out is a modelling exercise, not a rebuild, and it’s exactly the problem Salesupply is built to solve. If you want a second pair of eyes on where your Q4 exposure sits, get in touch.

Olga Foto New

Olga Pijanowska