Everyone wants to grow across borders. New countries, new markets, more sales.
But where ambition is high, the system often stalls. International growth rarely stalls for lack of market opportunity; it stalls because the structure behind it doesn't grow with it. Campaigns get too little data, algorithms learn too slowly and decisions are made based on assumptions rather than signals.
In this blog, we show you the fundamental mistakes that stall international campaigns; from incorrect PMax structures to unstable tracking. And how, with the right setup, data and local approach, you do achieve predictable growth.
Why one PMax campaign can block your international opportunities
One Performance Max campaign for all of Europe seems efficient, but it can also backfire. Large markets like Germany can grab a lot of budget, while smaller countries build up too little data to learn from. And that's a shame, because those smaller markets are precisely where lower acquisition costs and untapped growth opportunities often lie.
That doesn't mean you should always drive each country separately. When a market still has little data, isolation actually works against you: the algorithm lacks context, learns too slowly and distributes the budget inefficiently. At that stage, it is smarter to bundle countries temporarily, so that new markets can ride on the data stream of larger or similar countries.
However, once a market has enough signals, such as click volume, conversion data and feed quality, it pays to steer separately. A separate feed gives the algorithm room to learn what works locally and where growth potential lies.
Choosing the right structure is important for international growth; knowing when bundling strengthens and when unbundling pays off.
Without local feeds, Europe doesn't understand you
Consumers in Europe look alike, but behind that surface are distinct differences by market. From our white paper on Germany shows that German shoppers primarily value reliability, clear delivery information and secure payment options, while French consumers rather be driven by price, service and convenience.
These differences are consistent with what we also see more broadly in European research. A comparative study in Behavioral Sciences (Madlenak et al., 2025) confirms that online purchasing priorities vary widely from country to country. While in some markets price and speed are decisive, in others security and trust are key. What convinces in one market is counterproductive in another. This is why one generic feed or campaign for the whole of Europe rarely works optimally.
Tracking is often more unstable than dashboards suggest
Many Web shops assume that their data collection is in order and that every interaction is measured according to best-practice. In reality, in international markets, much of the add-to carts, checkouts and purchases are merely modeled. Modeled data fill in missing conversions with predictions. They appear accurate, but do not always reflect actual purchases.
POAS is risky when conversions and order values are modeled, because then you optimize on predicted profit instead of real profit. In doing so, you steer the algorithm on expected signals, making it less able to learn which audience or market is really paying off.
In markets such as Germany and France, users are less likely to consent to tracking, leaving parts of behavior invisible. Server-side tracking does not prevent this, but it does ensure that the measurements that are allowed remain more reliable. Because data is transmitted through the server rather than the browser, fewer signals are lost due to adblockers or cookie restrictions.
First learn, then scale
International growth often stalls not because of lack of potential, but because scaling up is done too quickly without a stable base. A new country gets budget, produces little and then is written off. Not because the market does not work, but because it was never examined whether the offer, the message and the data were sufficiently underpinned to grow. Data quality thus remains the biggest bottleneck.
Sometimes bundling is necessary precisely to build up sufficient data volume, but scaling without a solid foundation remains risky. Once a market has enough signals, steering locally becomes truly effective.
At Tomahawk, we apply this in phases. For example, in the collaboration with Karhu examined what worked for each market before we scaled. Only when the campaigns were set up locally did performance stabilize and ROAS improve significantly.
The success factors for international growth in 2026
Anyone who wants to make international growth predictable must get the basics right structurally. Specifically, this means the following
- Measure every intent and purchase through server-side tracking
With less data loss and more reliable measurements - Provide direct data flows between ad and analytics tools
So that you drive on complete, unfiltered performance data - Use country-specific feeds that are truly local
With market-specific USPs, attributes and currencies - Follow a scaling process that grows logically
First learn what works, only then scale. - Apply POAS when markets are mature
With stable volumes and reliable margin information
These steps are not complex. They are just often skipped, resulting in predictable losses. Growth across borders does not start with more campaigns, but with a system that is ready to learn.
