EU's €2 Line-Item Fee and Mandatory Product IDs Arrive 1 November

The Analysis

Something landed in my feed on 7 October that a lot of e-commerce sellers and parcel forwarders are going to feel in their margins long before they feel it in their dashboards. The EU put out its guidance on the 1 November customs changes, and the headline number that matters for anyone pushing low-value goods into the bloc is a flat EUR 2 charged per declaration item, every time a declaration is lodged from 1 November 2026. Not per parcel, per item line on the declaration, and that single distinction is where most of the money and most of the confusion is going to live.

I have been on the receiving end of fee changes like this before, and the pattern is always the same. The press writes about a round number, the brokers talk about how the system actually books it, and the people paying never quite meet in the middle until the first invoice lands. What sits in my stomach about this one is not the two euros. It is the quiet second half of the rule, the bit about product identifiers and the bit about the fee sliding into the import VAT base, and those two together rewrite the maths in a way the coverage has not really spelled out.

So before anyone files this under a two-euro problem, sit with what the guidance actually says. For consignments of remote-sold goods up to EUR 150, the EUR 2 is joined by a further EUR 3 per item as a temporary tariff. The affected low-value B2C declarations have to carry product identifiers C127, C128 or C129, and where there is no standard code to hang on an item you fall back to Y081. Three different tariff items in one parcel therefore run to EUR 15 of fixed cost before a cent of duty on the goods themselves, and that is the friendly version of a much messier story I want to walk through.

The first thing anyone running an import desk has to nail down is what actually counts as an item for the purpose of that EUR 2. The guidance says per declaration item, but a declaration item is not the same thing as a parcel, and it is not the same thing as a SKU, and it is not automatically the same thing as an HS subheading either. In practice the line that gets charged is the line the customs broker actually keys into the declaration system, and how many of those lines a parcel ends up producing depends on how the shipment is described, how the goods are classified, and whether the platform or the consolidator has already grouped things for the commercial invoice.

A single box with a phone case, a charging cable and a screen protector can be one combined line if the broker is happy to group them, or three lines if each piece gets its own heading, and the fixed fee on that one box ranges from EUR 2 to EUR 6 on the processing side alone before the tariff layer is even switched on. Stretch that spread across a catalogue of a few thousand lines and you are not looking at rounding error, you are looking at a cost line that can triple or shrink to a third depending on a data-entry decision someone makes at a screen on a Tuesday.

That ambiguity is worth dwelling on, because the people who set these systems up rarely think about it until it is too late. The declaration is assembled by the broker from instructions the seller or the forwarder passes down, and the seller usually thinks in parcels and SKUs while the broker thinks in tariff lines. Somewhere between those two languages the line count is decided, and the fee follows. A consolidator who merges ten small accessories into one commercial line to keep the invoice tidy is also merging ten potential fee lines into one, whether or not that matches how the goods should be classified. The reverse happens too: a platform that pushes one SKU per line because its own warehouse system is built that way can turn a three-accessory order into three declaration lines and three lots of EUR 2 without anyone intending to. The point is that the count is a choice made inside a workflow, and right now almost nobody owns that choice deliberately.

Now pile the second layer on top, because this is the part I think the trade press has undercooked. For the low-value remote-sales lane, anything up to EUR 150, that EUR 2 is not standing alone. There is a further EUR 3 per item as a temporary tariff, so each declaration line on that lane costs EUR 5 of fixed charge rather than EUR 2. And then the part that really moves the needle: the processing fee enters the import VAT base. That means the two euros is not the two euros. At a 21 per cent import VAT rate, every EUR 2 line actually costs EUR 2 multiplied by 1.21 by the time VAT is assessed on it, so EUR 2.42, and the EUR 5 combined charge becomes EUR 6.05 by the same logic.

The fee is being taxed, which is a rounding item on one line and a serious line item across millions of lines. The structure the rule quietly builds is a chain. A classification decision determines how many declaration lines a parcel becomes. The line count determines the fixed fee. The fee then feeds the VAT base. The VAT base feeds the total landed cost. Classification sits upstream of everything else, and that is the piece most importers will not see until their landed-cost model drifts a point or two and nobody can explain why.

Here is where it bites the people who design how goods are packed and how orders get bundled. Per-line charging inverts the instinct that has governed parcel economics for years, which was to combine as much as possible to save on postage and pick-and-pack handling. Combining can now save on fee lines, but only up to a point, and combining also drags you toward classification choices that carry their own risk. A three-piece set declared as one line is cheaper on fees than the same set declared as three lines, but merging different tariff articles into a single line can trip misclassification exposure, and the moment a member-state customs starts querying combined lines, the saving evaporates and a penalty conversation begins. The cost logic and the compliance logic point in opposite directions, and that tension is the real job for anyone mapping how their SKUs turn into declaration lines. It has stopped being a pure fulfilment question and has become a classification-and-fee question wearing a packing question's clothes.

Let me put numbers on it with the assumptions stated openly, because hand-waving helps nobody. Take a single B2C parcel bound for the EU. Assume it carries three distinct HS codes. Assume the goods value is EUR 120. Assume the destination member state applies a 21 per cent import VAT rate. Run the parcel two ways. Split into three declaration lines: the fixed charge is EUR 5 per line, that being the EUR 2 processing plus the EUR 3 temporary tariff, times three, which gives EUR 15.

The VAT base becomes the EUR 120 of goods plus the EUR 15 fee, so EUR 135, and VAT at 21 per cent is EUR 28.35. Total landed on this parcel is EUR 120 plus EUR 15 plus EUR 28.35, or EUR 163.35. Merge into one declaration line instead: the fixed charge is EUR 5 once, so EUR 5, the VAT base is EUR 125, VAT is EUR 26.25, and total landed is EUR 151.25. The gap between the two ways of describing the same physical box is EUR 12.10 per parcel, and EUR 10 of that is pure fee-line strategy while the rest is the VAT flowing off the larger fee base.

Scale it up. At an annual volume of 100,000 parcels, picking the merged-line treatment over the split-line treatment is worth about EUR 1.21 million a year on identical physical goods. That is the lever the rule drops in your lap, and it is also the lever that turns classification discipline into the difference between a margin and a miss.

I would be doing you a favour to stop there and call it a saving, but that EUR 12.10 is not free money and the calculation falls apart under more than one condition. This is a worked example built on the current guidance wording as published on 7 October, not a legal opinion, and three things can break it. The line count is only yours to choose inside the bounds of correct classification, so if the three articles genuinely belong in three separate HS headings, merging them is not a saving, it is a liability waiting for a customs query.

The IOSS and non-IOSS paths handle the VAT collection differently, and a consignment collected outside the Import One-Stop Shop may not carry the fee into the base the same way, or may surface the cost at a different point in the chain where your model is not looking. Member states have a long habit of reading union-level guidance with local colour, and a border that treats declaration item strictly as a declared line will behave differently from one that reads it against the commercial breakdown. The honest framing is this: under the guidance as written, line-counting is a real and sizeable lever, but it sits on top of classification correctness, and the moment you bend classification to chase the fee, the saving is at risk and quite possibly unlawful. The move that survives contact is to model both treatments now, lock the one that is defensible rather than merely cheap, and treat any further saving as upside you did not plan around.

For the importer and exporter actually moving this volume, the work between now and 1 November is not panic, it is a mapping exercise. Pull your top SKUs, see how many declaration lines each one currently produces, and run the two treatments side by side before the switch flicks. The product-identifier requirement, C127, C128, C129 with Y081 as the fallback, has to be sitting in your declaration data before the first post-change filing, not patched in after the first rejection bounces back. And the VAT-base effect means your landed-cost model has to absorb the fee as a taxable element, or your margin maths will be wrong in a direction you only notice at quarter-end when the spreadsheet will not reconcile. Someone inside the business has to own the line count, because right now nobody does and the fee is quietly shopping for a home.

A word on who actually eats this. The sub-EUR 150 lane is where the high-volume, low-margin sellers live, the ones already running on thin contribution after advertising and returns. A EUR 12 swing per parcel is the kind of number that turns a profitable SKU unprofitable without changing the price on the screen, and the platforms that pass costs to the shopper will do so, while the ones that absorb them will watch margin compress. Neither outcome is a catastrophe, but both are reasons to get the line logic right before the rule is live rather than after the first bad invoice. Get the classification honest, get the identifiers in, and model the VAT like it is part of the goods. That is the whole game for 1 November, and it is a data-and-classification job wearing a fee's clothes.

Step back from the fee itself and look at what it sits inside, because the context changes how much you should invest in re-engineering your declaration logic. That EUR 3 per item temporary tariff is not a random number. It is a placeholder while the union works out the permanent regime that replaces the low-value duty relief it has been signalling the end of for some time. The line-counting lever this article describes exists under a temporary rule that may be redesigned within a couple of years, and that fact should temper how much engineering you pour into your SKU-to-line mapping. There is a difference between riding out a temporary charge with sensible classification and rebuilding your whole declaration architecture for a rule that might look different by the time you finish building it. The cheap, defensible moves, get the identifiers in, model the VAT honestly, treat merged lines only where classification truly allows them, those survive any redesign. The expensive re-platforming does not.

The person who should own the line count is almost never the person who can change it, and that gap is where the EUR 12 a parcel leaks away. The seller's order system thinks in SKUs. The fulfilment partner's warehouse software thinks in picks and cartons. The broker's declaration platform thinks in tariff lines. The fee lands in none of those budgets explicitly, so nobody's dashboard flags it as their problem. Until someone is handed the line-count decision as a named responsibility with the cost attached to their number, the saving from this rule stays theoretical. I have watched this exact pattern with other per-line charges: the maths is obvious in a spreadsheet, and then it dissolves into the gap between three systems that none of them owns. Put a name and a metric on it before 1 November, or the lever stays on the page.

Do not treat the C127, C128 and C129 requirement as a compliance checkbox you tick once. Those product identifiers are how customs will triage the low-value B2C flow, and a missing or wrong code is a rejected declaration. Under the old world a rejection meant a delay. Under this new per-item structure a sitting parcel accrues the very fixed costs this article is built around, plus demurrage and a customer-service cost that shows up in returns and ratings rather than in the freight line. So the identifier upload is a throughput gate, not a formality, and the teams that load it cleanly before the first post-change filing will move while the teams that patch it afterwards watch parcels pile up in a queue they created themselves. Get the data right once, at the source, where the SKU is born, not at the broker, where it is too late to ask.

There is a competitive edge buried in here that the cost-focused reading misses. The seller who gets the line logic right can hold shelf price while a competitor eats the fee, and at the volume the example used, that EUR 1.21 million a year is not just savings, it is the room to win share or protect margin in a lane where everyone is already thin. The fee is uniform, so it does not discriminate between a disciplined declarant and a careless one, which means the saving goes entirely to the operator who bothered to model it. In a low-margin import game that difference is the whole quarter, and the teams that treat 1 November as a data project rather than a tax will be the ones still standing when the next change lands.

One more practical trap sits in the broker's software rather than in the rule itself. The per-item fee only becomes visible if the declaration system actually counts lines and bills them, and a lot of legacy broker platforms still price and report per shipment, not per line. An importer who assumes the EUR 2 is being tracked is often wrong, because the broker's invoice lumps it into a handling line and the per-line lever stays invisible until someone asks for the breakdown. Before 1 November, pull a sample declaration from your broker and confirm the line count is shown and the fee is itemised, because you cannot manage a cost you cannot see. The teams that get this right are the ones who ask the boring IT question first, not the ones who trust the headline.

I have said it a few times now because it is the only thing that matters here. This is not a two-euro story. It is a story about who decides how many lines a parcel becomes, and what that decision costs once the fee and the VAT base start talking to each other. The teams that treat it as the former will come out ahead, and the ones that treat it as the latter will spend the first quarter of next year explaining variances they could have seen coming. From where I sit, having watched a few of these changes land, the winners are always the ones who read the second paragraph of the guidance, not the headline.作者 Leo

↑ Back to the key points

— By Leo