European Ecommerce by Country: Where to Hold Stock

The single most useful number has two legitimate denominators. In Eurostat dataset isoc_ec_ib20 for 2025, indicator I_BLT12 is 73.56% with unit PC_IND: all individuals aged 16 to 74. The same indicator is 77.82% with unit PC_IND_ILT12: individuals who used the internet within the previous year. The article rounds those figures to 74% and 78%, and the country table consistently uses PC_IND.

The direction of travel is clear either way. The internet-user measure has risen from about 62% in 2015, so the last decade added roughly 16 percentage points. What that top-line hides is how unevenly it is distributed, and the gap is not closing at the same speed everywhere.

Ireland at 95% and the Netherlands at 94% show very high adoption, so additional growth is more likely to depend on purchase frequency and basket size than in lower-adoption markets. Bulgaria at 51%, Italy at 56%, and Portugal and Romania at around 60% have more room to add first-time online buyers. These are directional market observations, not forecasts.

The table lists every Member State by population, with the share of people aged 16 to 74 who bought online in the last 12 months. Both columns are Eurostat: the adoption figures are the 2025 survey, the population figures are 1 January 2026. Reading the two together is the point, because a high percentage of a small country is a small number of people.

Five countries hold roughly 65% of the EU population between them: Germany, France, Italy, Spain and Poland. Three of the five sit at or below the rounded 74% EU adoption average: Italy, Spain and Poland. Germany and France combine scale with adoption above average. The figures show comparative adoption; they do not by themselves prove where future growth or fulfilment demand will occur.

The high-penetration markets are mostly small. Ireland, the Netherlands, Denmark, Sweden and Czechia all sit at or above 84%, but together they are under 52 million people. They are excellent markets to sell into and poor arguments for a warehouse of their own, because the demand rarely justifies the fixed compliance cost of another country.

The practical reading is that adoption tells you how hard you have to work to win a customer, and population tells you how many there are to win. A seller optimising for volume goes where the people are. A seller optimising for conversion goes where buying online is already a habit. Very few of those answers are the same country.

This is where the market map and the compliance map separate. Eligible intra-EU distance sales can be reported through Union OSS without a destination-country registration for each customer market. Local registrations instead follow reportable events such as an import, a transfer of own goods and corresponding intra-Community acquisition, domestic supplies, outgoing transfers or exports, subject to the facts and any applicable exception.

Passive storage after arrival is not a separate taxable transaction. The import or cross-border transfer that brings the goods into a country, and later local supplies, outgoing transfers or exports, are the events to analyse. A pan-European fulfilment programme can therefore affect the registration footprint when it redistributes inventory across borders. We work through those triggers in if Amazon collects VAT, do you need VAT registration?.

The compliance country list is therefore transaction-based: where goods enter the EU, where own goods cross an EU border, where local supplies occur, and where goods leave. Warehouse permissions still matter because they determine which movements can occur, but storage itself is not described here as an independent taxable event.

You will see country-by-country ecommerce turnover quoted to one decimal place. Treat those figures with care, because no official EU statistic produces them. Eurostat measures adoption and behaviour, not national ecommerce sales, so every euro-value country ranking in circulation comes from a commercial estimate with its own definitions.

The scale of the uncertainty is easy to underestimate. Industry estimates of total European B2C turnover have been restated substantially between editions of the same annual report, with one year's figure revised by roughly a hundred billion euros in a later publication. Different providers also count different things: some include services, some only goods, some measure marketplace GMV and others retailer web sales, which are not interchangeable.

That does not make commercial data useless; it makes splicing it dangerous. Take one provider's series for a comparison rather than assembling a ranking from several, do not build a trend line across editions that have been restated, and keep the definition visible next to the number. For a decision about where to put a warehouse, adoption and population from Eurostat carry you most of the way without any of that risk.

Two regulatory shifts change the economics of these choices. Since 1 July 2026 the EUR 150 customs duty relief is gone, replaced by a temporary EUR 3 duty for each different tariff-classified item category in a qualifying low-value parcel. That fixed charge weighs most heavily on low-priced direct imports. The detail is in EU customs duty on low-value parcels.

Pulling the other way, from 1 July 2028 the ViDA transfer-of-own-goods scheme is designed to remove the registrations caused purely by moving your own stock between Member States. If that works as drafted, the marginal cost of adding a fulfilment country falls, and a footprint that looks expensive today looks different in 2028. The conditions are set out in the ViDA timeline.

Taken together, the direction is towards holding stock inside the EU and away from shipping every order across the border. A market-entry plan written on 2024 assumptions is likely to be pointing at the wrong structure.

Separate the two decisions and take them in order. Decide where to sell using demand, which the table above informs, and where to hold stock using logistics and compliance cost, which it does not. Selling into a country is cheap; storing goods in one is not.

Send us the shape of the plan and we will price the compliance side of it before you commit to a warehouse network. Contact our VAT team with the countries in question, or see the pricing page for what registration and filing costs per country.