What actually breaks when a European company enters the US

Not the strategy. The seams — and the operating debt you were already carrying at home.

On paper it is the easy move. Bigger market, a business culture that looks familiar, a playbook that already works, more or less the same language. Nobody writes a board paper saying the US will be hard.

It is one of the most underestimated operational challenges I have seen. And the gap is rarely strategic. It is behavioural, cultural, and structural.

First, a number worth being careful with

A figure gets quoted constantly in this context — that seventy per cent of transformations fail. I have used it myself. Most versions of it trace back to nothing you can check: a 1993 book whose authors called their own estimate unscientific,[2] a 2015 consultancy article with no method attached,[3] a column that multiplied the number by a spending forecast. A peer-reviewed review of the literature found no valid empirical basis for the figure at all.[4]

The version you can check is BCG’s, from 2020:[1] internal data on seventy companies plus a survey of 825 senior executives. Thirty per cent met or beat their target. Forty-four per cent created some value but missed. Twenty-six per cent delivered less than half of what they set out to.

The interesting number is not the seventy. It is the forty-four. Most cross-border expansions do not blow up. They quietly underdeliver for years while everybody stays busy. That is far harder to see, and far harder to stop.

Five things I watch break

1. Customer expectations

US customers expect faster responses, clearer accountability and more proactive service. What counts as exceptional in Europe is often the baseline here. Companies find this out late, when the friction is already showing up in churn.

2. The size of the adaptation

Teams assume the model transfers with light adjustment. It rarely does. The US is not a plug-and-play extension of Europe, and it is not Canada either.

3. The talent premium

US operators cost more, and waiting to hire them is the expensive choice. Local expertise is not discretionary spend — it is operating infrastructure. The premium actually worth paying is for people who have worked across both cultures, not simply for people who are here.

4. Data fragmentation

Two countries become two truths quickly. Misaligned definitions, inconsistent controls, delayed remediation. It surfaces downstream in reporting, in risk, and in what the customer sees. This does not mean the head-office definition wins by default; that is its own failure mode.

5. Nobody owns the seams

Everyone touches cross-border integration. No one owns it. And the seams are exactly where execution breaks — the handoffs, the assumptions, the nine hours between San Francisco and Paris. Integration has to be designed. It does not happen because people are willing.

For scale-ups these show up precisely when growth accelerates, not before. Which is the worst possible timing.

A case where all five showed up at once

N26 entered the US in 2019 with a strong European product and real momentum. It reached roughly 500,000 US accounts.[5] On 18 November 2021 it announced it was leaving; US accounts closed on 11 January 2022.[6]

The public reason was a refocus on Europe. The more useful version is what co-CEO Max Tayenthal said afterwards: “We had to build the bank from scratch with different licenses, agreements, and payment rails.”[6] In Europe, N26 held a full banking licence. In the US it operated through a partnership with Axos Bank — less control, less differentiation, competing against Chime, which had thirteen million accounts.[5]

And the home market was not quiet. In September 2021 BaFin fined N26 €4.25 million for taking too long to file suspicious activity reports. Weeks later — in the same month the US exit was announced — the regulator capped new customer sign-ups at 70,000 a month, later tightened to 50,000, and installed a special commissioner over the bank’s anti-money-laundering controls. That cap was not lifted until June 2024.[7]

Whatever the US business needed, it was competing for executive attention against a regulator at home. That is the pattern worth taking away: the US did not create N26’s problems. It amplified the ones that were already there, and put a price on them.

The part that is not in the operating model

One more thing, because it is the one people skip.

You can design the matrix, the federation, the hybrid. You can define decision rights, escalation paths, governance, dashboards and operating cadences. If the people across the two countries do not actually know each other, none of it works.

Cross-border performance runs on trust, and trust runs on human connection, which does not build well over video. So it has to be built on purpose: onsite weeks, rotations, shared rituals, periodic in-person alignment. On a budget line these read as perks. They are infrastructure.

Most cross-border friction is not conflict. It is misinterpretation — and the only fix for misinterpretation is knowing the person at the other end well enough to read what they did not say.

After more than a decade working for foreign firms in the US, I have seen every permutation of this. Organisational design does not fix weak relationships or power struggles. The operating model is the engine. The relationships are the oil.