Expanding between the US and EU on Amazon: the cross-marketplace PPC playbook.

The single most expensive mistake when expanding between the US and EU on Amazon is copying campaigns across marketplaces and assuming the bid math travels. It does not. Different CPCs, different shopper behaviour, different competitive density, different languages — a US campaign cloned into Germany carries a documented +28% ACoS and −42% CTR penalty. Here is the cross-marketplace PPC playbook, in both directions.

Why "just translate it" fails

Localisation is the easy part to see and the small part of the problem. The hard part is that the economics differ per marketplace. The break-even CPC for the same product is different in the US, the UK, and Germany because conversion rates, competition, and price points differ. A bidding tool that runs one model and applies it everywhere mis-prices clicks in every market except the one it was tuned for. We documented the German case in detail in why German campaigns underperform UK ones, and the broader principle in EU-native is not a translation layer.

US sellers expanding into the EU

What changes when you cross the Atlantic eastward:

• Five-plus distinct marketplaces, not one. DE, FR, IT, ES, and the UK each have their own auction dynamics. Pan-EU FBA pools inventory; it does not pool bid strategy.
• Language is per-marketplace, and so is intent. German shoppers search differently from French ones — and Amazon's COSMO/Rufus layer reads that intent natively now, so machine- translated copy reads as machine-translated to the algorithm too.
• Compliance is real. VAT, EPR, and GDPR are procurement gates. Your tooling should be an EU-hosted, GDPR-compliant processor, not an afterthought.
• Germany is the prize and the trap. The largest EU marketplace is also where the untranslated-copy penalty bites hardest.

EU sellers expanding into the US (and Canada / Mexico)

What changes going westward:

• One huge, fiercely contested marketplace. Higher click volume, higher competitive density, and CPCs that run hot on competitive terms — past $2.50 in beauty and supplements.
• North America is three marketplaces. The US, Canada, and Mexico each have their own auction. A unified North America strategy still needs per-marketplace bid math underneath.
• Currency display vs billing. Your customers see USD; your costs and reporting may be in EUR. Keep the distinction clean so margin analysis stays honest.
• Different seasonality. US peak events and promotional calendars do not line up with EU ones — your budget rhythm has to change, not just scale.

The playbook, in five steps

1. Compute break-even CPC per marketplace, per ASIN. Never reuse the home-market number. Margin and conversion rate move across borders.
2. Localise intent, not just words. Build keyword sets from how shoppers in that marketplace actually search — and write listings clear enough for the AI layer to recommend.
3. Run per-marketplace bid math. One model per marketplace, not one model translated.
4. Set per-marketplace safety limits. A new market is where a runaway bidder does the most damage. Cap spend per marketplace with circuit breakers — see the safety architecture.
5. Compare TACoS across marketplaces, normalised by margin. If one market runs 3+ points hotter than the others, you have found your localisation tax.

Why this is hard to do by hand

Each marketplace you add multiplies the surface area: more break-even calculations, more keyword sets, more daily inventory checks, more auctions to watch. The work scales linearly with markets while your attention does not. This is the case Mirox was built for — multi-marketplace native across DE, FR, IT, ES, NL, BE, PL, SE, UK, US, CA, and MX, with per-marketplace bid math and one trace per decision.

See how the per-marketplace model works, or watch it run across your marketplaces in Simulation Mode before you commit budget to a new one.