Crossing a border adds four constraints a domestic adjacent market does not have: local credibility, pricing that does not transpose, a longer decision path, and the fixed cost of a structure. Pick the first country on how close its buying path is to yours, never on market size. Hire locally once you have ten local customers, not before: before, it is a bet, after, it is a consequence.
International rarely arrives as a decision. It arrives as an inbound request from somewhere else, an existing customer opening abroad, or an investor asking where the next growth is. And it almost always starts the same way: translate the site, run a few campaigns, wait.
Opening a country is a special case of adjacent market, and the general method still holds: interviews first, narrow segment, deals sold by hand, industrialisation afterwards. We described it in launching an offer into an adjacent market. What follows is only what the border adds, and what does not exist when you open a neighbouring sector at home.
The four constraints specific to a border
Credibility does not cross. Your references are unknown, your brand means nothing, and your foreign base is read as a service-continuity risk. It is the heaviest constraint and the least anticipated: it is fixed neither by product nor by price, but by local proof, meaning the country’s first customers.
Pricing does not transpose. The same grid can be out of market in one country and undervalued in another, because salary levels, available alternatives and buying habits differ. Converting a price into another currency is not a pricing decision, it is a conversion. The real question stays the same as anywhere: what value does the customer get, and how do they measure it locally, a subject we cover in value before price.
The decision path lengthens. A purchase signed alone at home can go through legal, a security review and a group-level approval elsewhere, especially when headquarters sits in a third country. Forecasts built on your home market’s durations will be wrong, and you will draw the wrong conclusion from them.
Fixed costs arrive before revenue. Entity, local accounting, compliance, sometimes in-country data hosting. These do not scale with volume, so every month without a customer weighs more. That is what makes the order of steps decisive.
Choosing the first country
The temptation is to pick the largest market, usually the United States seen from Europe. It is the most expensive choice: the market is vast, but competition is entrenched, service expectations are high, and acquisition cost without local recognition is brutal.
The criterion that saves time is how close the buying path is. In which country does the buyer hold the same title, control the same budget, follow the same approval process as your current customers? That country takes half the learning of a market three times larger but organised differently. Language matters less than process: two countries sharing a language can buy in very different ways.
Also look at what you already have. Your visitors, your unsolicited inbound, the countries where sign-ups appear without you prompting them. The market has often already voted, and nobody has read the result. It is the same reflex as starting from your real customers rather than your theoretical target, which we describe in why an ICP that is too broad kills growth.
Remote first, on the ground second
The sequence that works is almost always the same, and it starts with nobody on site.
Sell remotely up to ten customers. It is uncomfortable, less efficient, and it teaches you exactly what you need: which objections recur, which vocabulary lands, which proof is missing. Ten customers won from your home country are worth more than a local director hired on a promise.
Then hire, and hire a seller, not an executive. The classic mistake is recruiting a senior profile to “build the subsidiary”. They arrive with no local reference, no material in the right language, no process proven in that market, and spend their first year doing what you should have done before them. A seller who inherits ten existing customers and a validated argument produces far faster.
Open the entity only when it becomes necessary. Hiring locally, answering a public tender, or sparing your customers a painful indirect tax. Those three reasons are real. “Looking serious” is not one.
What to translate, and in what order
Translating the whole site before having a customer is an investment that expires: your message will change after the first ten interviews, and you will pay for the translation twice.
Start with what serves the sales conversation: the page that explains the value, the proof, and the documents a rep sends after a meeting. The rest can wait until traffic justifies it.
One technical point deserves attention, because it is expensive to repair: a partially translated site must declare its language equivalences properly, and never announce a version that does not exist. A prospect who switches language without warning mid-journey wonders who they are dealing with, and search engines draw the same conclusion.
The signals that a country will not take
They are the same as on any adjacent market, with one nuance.
If all your local deals come from companies headquartered in your home country, you have not opened a market: you have followed your customers. That is useful, often profitable, but it proves nothing about your ability to convince a purely local company, and it does not transpose to the next country.
If you win only by being cheaper than entrenched players, you are funding market share that evaporates at the first price increase. And if every deal requires a different regulatory adaptation, you are not on a market, you are on a series of special cases.
What it changes in your steering
Track the country separately from day one, with its own indicators and its own trajectory. Folded into the general dashboard, a new country is invisible for a year: too small to weigh, expensive enough to degrade the average without anyone knowing why. That separation assumes your data can tell the two apart, which belongs to the foundation described in building your revenue operations foundation.
Announce the expected degradation before starting, too. A higher acquisition cost and a longer cycle during the first quarters are normal. They only become a problem the day nobody predicted them, and the discussion moves onto people instead of onto the market.
Where to start
Pull the list of your inbound requests from the last twelve months and sort them by country. Then check, for the top three, whether the buyer there holds the same title as at your current customers.
That cross-check takes an hour and it names your first country better than a market study. Then call five of those inbound requests, even old ones, even lost ones. They will tell you what was missing, and that is exactly the list of what to build before going there for real.
Frequently asked questions
- How do you choose your first country?
- On how close its buying path is to yours, not on market size. A country where the buyer holds the same title, the same budget and the same approval process as your existing customers takes half the learning of a market three times larger but organised differently. Also look at where your visitors and unsolicited inbound requests come from: the market often names its own candidate.
- Do you need a local legal entity?
- Not to sell. Most B2B companies invoice from their home country for the first few years without it blocking a signature. The entity becomes necessary when you hire locally, when a public tender requires it, or when indirect taxation makes cross-border invoicing too heavy for your customers.
- When should you hire a local salesperson?
- Once you already have about ten customers in the country, won remotely. Hiring earlier means asking someone to validate your promise and sell at the same time, in a market where you have neither references nor recognition. That profile exists, it is expensive, and it leaves after a year when results do not come.
- Should you translate the site before selling?
- Translate first what serves the sales conversation: the pages that explain the value and the proof. The rest can wait. A half-translated site, where the prospect switches language without warning mid-journey, costs more than a site entirely in your original language.