Somewhere in Seller Central there's a button that suggests Europe is basically one click away. Build international listings, sync your catalogue, reach hundreds of millions of new customers. The pitch writes itself.
Here's what the button doesn't mention. VAT registrations that take months. An EU Responsible Person requirement that can get your listings pulled overnight. German customers reading product copy that was clearly machine-translated from English and deciding, correctly, that you don't take their market seriously.
Amazon European expansion is genuinely one of the best growth levers available to a UK brand. We took one client, a certified consumer brand, into four new EU marketplaces as part of rebuilding their channel. But it works when it's treated as a set of deliberate decisions, and it burns money when it's treated as a box to tick. Let's take the decisions in order.
EFN or Pan-EU: the first fork in the road
Fulfilment is where the real trade-off sits, and it comes down to two models.
European Fulfilment Network (EFN) keeps your stock in one country and has Amazon ship cross-border to fulfil orders elsewhere. Light on compliance, because your inventory only lives in one place, but heavier on per-unit fees and slower on delivery promises, which hurts conversion in markets where next-day has become the expectation.
Pan-EU flips it. Amazon spreads your inventory across fulfilment centres in multiple countries, you get local fees and local delivery speeds, and your offers become properly competitive. The catch is that stock stored in a country generally means VAT obligations in that country. Stock in seven countries, seven sets of registrations and filings.
Neither is "correct", and the right answer moves with your volume, your margin, your category and how much compliance you can genuinely carry. We've watched the same product justify a different answer eighteen months apart. What goes wrong is rarely the choice itself. The expensive version is the brand that never noticed a choice was being made: a toggle was on, stock drifted across borders and the VAT obligations surfaced after the fact. Tax authorities are not sympathetic to "we didn't realise".
The compliance nobody warns you about
Since Brexit, a UK brand selling into the EU is a third-country brand, and two requirements catch people out constantly.
The first is the EU Responsible Person. For CE-marked products, EU market surveillance rules require an entity established inside the EU whose name and address appear on the product or packaging, who holds your technical documentation and answers to regulators. No Responsible Person, no legal route to market, and Amazon does enforce it. Brands have watched well-reviewed listings vanish overnight in a single enforcement sweep.
The second is VAT, which deserves more respect than it usually gets. Registrations can take weeks or months depending on the country, Germany wants its certificate and filings recur forever. The work isn't hard so much as slow, and every week of it needs to happen before your first euro of revenue, not after.
Then there's the quieter compliance: labelling languages, WEEE registration for anything with a plug or a battery, packaging levies in France and Germany. None of them will sink you on their own, but stacked up they're the reason "just switch on Europe" is a sentence only ever said by someone who hasn't done it.
Don't machine-translate your way into Germany
Run your bullet points through machine translation and a German shopper will spot it in the first sentence. What you save on translation you lose in conversion, permanently, on every visit.
Proper localisation starts with native keyword research, because Germans don't search for a translated English phrase, they search for the word a German would actually use. It means re-doing the argument, not just the words. A claim that lands in the UK can fall flat in France. It sometimes means new images, where text overlays or plug types or certification marks are wrong for the market.
This is the same discipline as anywhere else on Amazon. The listing is a shop the customer can't walk into, and in Munich that shop needs to speak German like a local, not like a tourist with a phrasebook.
One market at a time beats five at once
The Seller Central tooling makes launching five marketplaces as easy as launching one, which is exactly the trap. Five simultaneous launches means five thin ad budgets, five sets of half-localised content, five review counts starting from zero and one team spread across all of it.
Sequence it instead. The certified brand we mentioned earlier added its four EU marketplaces one at a time, each launch localised and funded like it mattered before the next one opened, and the reason it worked is that each market was actually run rather than merely opened. Which market goes first is a proper decision in its own right, weighing size against competition against compliance load, and the obvious pick on the map is wrong more often than you'd expect. We've written more about that ordering in why the order of markets beats the land-grab.
One more thing before any of it: fix home first. Expansion multiplies whatever you already have. If your UK channel has pricing leaks and a messy seller list, Europe gives you the same problems in four languages, and unauthorised sellers follow you across borders faster than you'd think. The price erosion problem doesn't stop at Calais.
The honest maths of European expansion
Done properly, a European launch is a real project: registrations, localisation, compliance, stock and a ring-fenced ad budget per market. Done as a box-ticking exercise, it produces the most common outcome we see in accounts we take over. Five marketplaces technically live, none of them working, and a compliance risk quietly accruing underneath.
The gap between those two outcomes isn't budget. It's whether someone treated each market as a market.
That's roughly how we run it at Orellana: distribution and agency under one roof, one market at a time, compliance boring and done early. If Europe is on your roadmap and the VAT-and-Responsible-Person layer is the part putting you off, that's usually the most fixable part of the whole project.
.png)
.png)
.svg){kind=icon}
