Somebody in a meeting said Finland, and then Sweden, and possibly Germany because the market is large.
That list is now in your plan. Nobody checked it against search data, because the decision felt like a commercial one rather than a marketing one, and by the time anybody looks at volume the translation quote has already been approved.
I am not saying the list is wrong. I am saying it was chosen in a room, and the cheapest thing you can do this morning is find out whether the data agrees with the room.
With a domestic population of 1.3 million, almost all of your organic opportunity is in somebody else’s language and somebody else’s results page. Which makes market selection the single highest stakes decision in the whole plan, and it is usually the one nobody researched.
Validate before you build anything
This is about an hour of work per candidate market and it goes in this order.
Write down the ten to fifteen queries that actually indicate somebody is about to buy. In English first. Not topics. The specific strings a buyer types, including the boring ones with the word price or the phrase alternatives to.
Get those translated by a person from that market, not by a tool. This step is where the whole exercise usually goes wrong. The literal translation of your product category is frequently not the term practitioners in Helsinki or Stockholm actually use, and a keyword list built from machine translation will show you no volume for a market that has plenty, because you are measuring the wrong words.
Pull volume for the translated set, per country. You are not looking for large numbers. You are looking for whether commercial intent exists at all.
Look at who ranks. Three possible answers, and they mean different things. Established local competitors means the demand is real and expensive. Aggregators and directories means the demand is real and the buyers are comparing, which is usually good news for you. Nobody in particular means either the demand is not real or you have found something.
Then check the language of the results Google returns in that country for those queries. This is the most useful signal in the whole process and it takes two minutes. If page one is entirely in Finnish, an English page will not compete there whatever anybody tells you about English proficiency. If page one is a mix, you have room.
Do that for each candidate market before committing a cent, and the list from the meeting will usually change.
Translate or write natively, honestly
Machine translation on its own does not rank. Google’s quality systems identify it reliably, and the failure is not grammatical, it is that translated copy misses how people in that market actually phrase the problem.
The pattern that does work in practice is a machine or AI first draft followed by real editing by somebody native to the market. Published guidance puts that editing at something like two to three hours per fifteen hundred words, which is a real cost and a much smaller one than a translation agency.
And be careful with the reassurance you will hear about English. Nordic English proficiency is genuinely excellent, and that is not the same as searching in English when spending company money. The evidence on Nordic B2B is fairly consistent that local language content converts better, even among people whose English is better than mine. This is also the register problem I went through when writing about selling into Finland and Sweden, which is a separate matter from the language itself.
The one page test, before the fifty page build
Here is the discipline that saves the most money, and hardly anybody does it.
For each market that survived the validation above, publish one page. Your single best commercial page, properly localised, at the right URL, with the tagging correct.
Then wait eight to twelve weeks and look at impressions rather than clicks. Clicks at that stage will be almost nothing and will tell you nothing. Impressions tell you whether Google is willing to show you in that country at all, and for which queries, which is the only question you asked.
If impressions climb and the queries look commercial, build the rest. If Google shows the page for nothing after three months, you have learned that for roughly one page of cost instead of forty.
The temptation is to skip this because one page feels too small to prove anything. It is not proving that the market will work. It is proving that the market exists in search, which is a smaller and far cheaper question.
Structure, briefly, because people overthink it
Subfolders. For almost everybody, at your stage, subfolders on your existing domain.
They keep your authority in one place, which matters enormously when that authority is small. Google has been explicit that subdirectories work for international targeting, with the tradeoff that you do more work to establish geographic relevance.
Country domains send the strongest geographic signal and require you to build authority from nothing, per country. That is a reasonable trade for a market that already produces real revenue and a bad one for a market you are testing. You can always add a country domain later for your best market while the rest stay in folders.
The language and region tagging that tells Google which version to show whom is a separate job, it is easy to get subtly wrong, and a site with mismatched tags can end up with the wrong pages competing against each other. Get it right when you set up the first folder rather than retrofitting it across forty pages.
What not to do
Do not translate your whole site into four languages simultaneously. That is the single most common way an Estonian company spends a year of content budget on pages nobody in any market will find.
Do not build a market page because a salesperson has one customer there. One customer is a relationship, not a market.
And do not let the domestic Estonian version of the site absorb effort in proportion to how it feels. It is your home, it is the version your team reads, and it is the smallest search market you will ever work in. Keep it correct and put the hours elsewhere, which is the same discipline as not copying what worked for a company ten times your size.
Before lunch
An hour, in this order.
Write the fifteen buying-intent queries for your category in English.
Message somebody you know in your first target market and ask them what they would actually type. Not what the words translate to. What they would type. That conversation is worth more than any tool.
Then search three of those translated queries with the country set to that market, and look at what language page one is in.
You will either confirm the plan or discover it was built on an assumption, and both of those are a good morning’s work.
My notes on SEO cover the technical side once the market choice is settled. If you’d like to think through your own market list, drop me a line on email, WhatsApp or LinkedIn and we can have a quick chat. I’m contracted full time so this isn’t a pitch.