How a D2C brand entered ten international markets in under two years

In 2024, TEVEO was one of the strongest social-first activewear brands in the DACH region, but internationally it was starting from zero: no account, no setup, no conversion data outside its home market. The home market itself was largely saturated, and growth planning expected little further gain there. The entire growth target rested on international markets.

When growth has to come from markets where nobody knows the brand, internationalisation stops being a nice-to-have and becomes the growth strategy. That also means it has to be built as a system rather than as a side experiment.

TEVEO is a German activewear brand built through social media and a community that ranks among the strongest in the segment. The collections are created close to this community, and drops sell through the brand’s own D2C infrastructure with localised shops. For us, that means we are taking a brand with a fan base into markets where that fan base does not yet exist, which is a very different task from advertising an interchangeable product.

What forms the foundation of every optimisation in the home market simply does not exist in a new market. No best-performing keywords, because nobody there has ever measured a conversion. No audiences, because there are no existing customers and no remarketing pool. No brand search volume, because nobody there knows the brand. No conversion history for Smart Bidding to work with. And not even reliable benchmarks, because CPC levels and competitive density vary considerably from country to country. The usual reflex of translating the shop, copying the campaigns and adding budget fails for exactly these reasons.

The goal was clear: to build the international expansion as a scalable system instead of a series of individual projects. In concrete terms, a new country goes live within a few days using the same playbook, with tracking in place, a finished feed and a defined launch range, instead of being built up as a separate project over months. And from day one, the system was meant to show honestly which markets carry the growth and which do not.

Most brands internationalise cautiously: one market, perhaps two, followed by years of observation. TEVEO took the opposite route with us. After the first go-live in France in October 2024, the other markets followed at intervals of weeks rather than quarters, at peak pace two new countries a week, until all ten were live. Only then did prioritisation begin, based on real data from each market instead of assumptions made in advance. Performance decides which markets carry the growth, and gut feeling before launch has no say.

The account structure

A separate Google Ads account for each country, linked through an MCC tracking setup: conversion tracking and audiences sit centrally at manager level, while steering happens locally in the country accounts. Every new market starts with a proven measurement setup, and because the same conversion definition applies in every country, the figures from all markets are genuinely comparable. This is not the growth lever itself, but it is the precondition for making clean budget decisions across ten markets running in parallel.

The feed infrastructure

Central feed management with Channable produces a localised product feed for each country, including language, currency, shipping information and availability. A new country feed is therefore created by rules from the existing setup instead of by hand, and the quality of this feed directly determines how well Shopping, and later Performance Max, deliver. Without this setup, each of the ten markets would have been a feed project of its own.

The range

Every market launched with the bestsellers from the DACH region that were known to work. Taking the full range along would have meant spreading the budget across hundreds of products while the algorithm did not yet know what works in the market, burning precisely the budget that is usually tracked most closely during internationalisation. Starting with the bestsellers allows budget to be allocated with precision, and new, market-specific products are tested in a controlled way later. On the campaign side, the entry ran through Search and Shopping, focused on the core category; once enough conversion data had been built up, Performance Max was added.

Ten markets running in parallel create ten different realities. The real data produced priority tiers: markets meant to become established, with matching budgets; test markets with a defined time frame; and defensive markets with minimal investment. France, as the strongest market, was expanded consistently.

Steering was based on contribution margin (CM3) and CAC per market from the attribution view, rather than on ROAS in the ads interface. Platform ROAS only says how much revenue is booked against each advertising euro. It does not say whether anything is left after product costs, shipping and returns. Whether a market pays off is decided by the contribution margin and by what a new customer may cost to acquire.

On top of that comes an effect that grows with success: as long as a market is new, there are hardly any brand searches. Once brand awareness develops, brand campaigns with very cheap conversions are added, making platform ROAS look better and better even though new-customer acquisition behind it has not improved. Exactly when a market starts to look good, the platform view becomes least meaningful.

One of our priority-1 markets, highly competitive and with noticeably more expensive clicks than comparable markets, took more than a year for its efficiency to settle. Three things carried it through: its strategic weight, a structural explanation for the expensive start, and contribution margin metrics that moved measurably in the right direction month after month. Today, this market runs almost three times as efficiently as in its first half-year and is one of the strongest in the set.

Three questions for the market decision

What is the trend, regardless of the level? A market may have poor absolute figures as long as its contribution margin curve rises measurably. Is there a structural explanation for the weak start? High competitive intensity and expensive clicks are a ramp-up problem; no traction despite favourable conditions is a market problem. And what does patience cost? Every euro in the weak market is missing from the strong one. A weak level plus a positive trend plus a structural explanation means build. A flat trend without a structural reason means reallocate.

And sometimes the framework conditions decide: Norway was paused because the tax requirements did not allow the business to operate sensibly there, and performance played no part in it. That, too, belongs to an honest expansion. The account stays in place in case the conditions change.

In under two years, TEVEO went from zero to ten international markets. The first three new markets, France, the Netherlands and Switzerland, now account for around a third of total conversion value. France, the first market and the one expanded most consistently, alone accounts for around 20% of revenue.

The half-year comparison of H1 2025 against H1 2026 shows the curve: conversion value grew 2.6-fold in France, 3.7-fold in the Netherlands and 1.5-fold in Switzerland. This growth was not bought: efficiency rose in parallel, and in the strongest market ROAS has almost tripled compared with its first half-year. The growth the home market could no longer deliver now comes from there.

The system is in place; now it is time to go deeper. The established markets are expanding their range beyond the bestsellers, the strongest late-starting market has already triggered its own follow-on expansion into a neighbouring market, and prioritisation continues: budget flows to where the contribution margin justifies it. Opening a new country is now a rollout of a few days and no longer a fundamental strategic decision. That was exactly the goal.

Anyone who wants to scale internationally should put the system in place before spending the first euro: account structure, MCC tracking and feed infrastructure decide whether market number ten still scales or only creates work. The courage to build every market at once and only then prioritise paid off, because the prioritisation rests on real data instead of assumptions.

Starting with bestsellers reduces the unknowns of market entry to the market itself and protects the budget during the phase in which the algorithm knows nothing yet. And patience is only a virtue when the trend justifies it: anyone who pulls out after three months pays for the learning phase and forgoes the return.