Recruitment agency expanding to a new country

Choose the market by counting, not by feel

Most agency expansions are decided by proximity, a personal connection or a single promising conversation at a trade fair. Those are real inputs and they are not evidence. The market you can actually count is the one you should enter.

Counting means three numbers per candidate country, gathered before any commitment. How many companies exist that fit your client profile. How many of them are hiring the roles you supply, right now. How many agencies already serve them. All three are obtainable in a week from public sources, and together they eliminate most candidate markets without spending anything.

What this usually reveals is that the obvious market is obvious to everyone. Germany is the first choice of most Central European staffing agencies, which means it has the highest demand and the deepest incumbent competition on the continent. That is not a reason to avoid it. It is a reason to enter it with a narrower specialisation than you would need somewhere thinner.

It also reveals the opposite case: countries where demand is modest but agency density is low, and where a competent supplier is unusual rather than one of forty. Those markets rarely make it onto a shortlist because they feel small, and they are frequently where the first cross-border contract actually gets signed.

The three tests a candidate market has to pass

Demand. Count live vacancies in your roles, by employer, over a period rather than on one day. The useful figure is not total vacancies, it is vacancies per employer. A market with many employers each hiring one person is a placement market. A market with fewer employers each hiring ten is a volume market, and it needs a different agency shape to serve.

Competition. Count registered staffing and recruitment companies serving those roles in that country, and look at how many are specialists rather than generalists. A market with many generalists and few specialists is open to a specialist. A market that already has six specialists in your niche is not, unless you arrive with something structurally different.

Permission. Establish what you are legally allowed to do, both to reach clients and to supply workers. Temporary agency work is licensed in many countries, and the licence takes time. This test is last in sequence and first in importance, because failing it invalidates the other two entirely.

Run all three before choosing. Agencies that run demand only, which is most of them, discover competition in month two and permission in month four, by which point the budget is committed.

Channel legality differs more than anything else

European B2B outreach is often described as one regime because GDPR is one regulation. GDPR governs personal data. Whether you may send a commercial message is governed by national law implementing the ePrivacy directive, and those diverge sharply.

Germany is the case that matters, because it is the market most agencies pick and the one where the default playbook is least usable. The workable approach is to lead with LinkedIn, which is not caught by the same rule, use the phone within its own limits, and keep email as a follow-up to a contact that already exists rather than as the opening move. Slower, entirely functional, and the reason a German campaign should be resourced differently from a Polish one.

None of this is legal advice, and an agency committing to a market should take some locally. The planning point is that the channel decision follows the country, and a plan that assumes one sequence works everywhere will underperform in exactly the market it cares most about.

Language is a commercial decision

The question is not whether the buyer speaks English. Many do. The question is what sending English says about how seriously you intend to operate there.

In the Nordics and the Netherlands, English is normal in business and an English approach costs little. In Germany, France, Poland, Italy and Spain, English marks the sender as external and raises an unspoken question about who would service the account day to day. That question is fatal in staffing specifically, because the client is buying local delivery.

The middle path that works is native language in the outbound, with the honest position that delivery is coordinated centrally and the local relationship is real. What fails is a native-language email followed by a call in English, which reveals the arrangement in the least reassuring possible way. Decide the language of the whole funnel, not the first message.

Entity, or no entity

Not at the start, in most cases. An agency can sell into a neighbouring country, contract cross-border and deliver workers under posting rules without incorporating, and doing so for two or three quarters answers the only question that matters: whether clients there will actually sign.

A local entity becomes necessary at identifiable moments. When the country licenses temporary agency work and the licence requires local establishment. When clients require a domestic counterparty for contracting or payroll. When posted-worker administration becomes a permanent cost rather than an occasional one. Or when the volume justifies local staff who need a local employer.

Incorporating before any of those is true adds fixed cost, accounting obligations and management attention while removing no constraint. The reverse mistake, delaying past the point where clients are asking, is less common and more damaging, because it looks like reluctance to commit.

What the first ninety days should contain

A cross-border entry that is going to work shows evidence early. A useful ninety-day plan is narrow on purpose.

• Weeks one to two. Finish the counting. Confirm demand, competition and permission. Choose one role family and one region, not a country.
• Weeks three to four. Build the list properly from the national register plus place data, and settle the channel per the legal position. Set up sending infrastructure separate from the domestic one.
• Weeks five to ten. Run the campaign at low volume and high specificity. The objective is not meetings, it is learning which objection appears most often.
• Weeks eleven to thirteen. Deliver one placement, even a small one, even at a poor margin. A reference in-market changes every subsequent conversation and nothing else does.

Notice what is absent: an office, a local hire, a translated website and a brand launch. All four are commonly done first, all four cost money before anything is proven, and none of them wins the first client.

The four things that break

The list, because the data model is different. Registers differ by country in what they publish, how they classify activity and how they name entities. A list-building process tuned to one register produces poor coverage against another, and the failure is silent.

Deliverability, because the domestic domain gets reused. Sending a new market’s campaign from the established domain risks the domain that already works. New market, new domain, warmed separately.

Follow-up, because of the time difference in decision speed. Buying rhythms differ. A German client that has not replied in three weeks may still be evaluating. A Polish client that has not replied in three days has moved on. A single cadence applied to both mistimes one of them.

Delivery expectations, because the client assumes local. The first cross-border placement usually surfaces an assumption nobody stated: that somebody would be physically present. Agreeing what presence means, in the contract, before the first placement, prevents the most common early loss.

When to stop and try elsewhere

Set the stopping rule before starting, because in month five nobody wants to be the person who says the market was wrong.

Reasonable rules for a first entry: no qualified conversations after eight weeks of correctly executed outreach, or qualified conversations that consistently stall on the same structural objection, such as the absence of a local entity or a licence you do not hold. The first says the targeting or the channel is wrong and is worth one correction. The second says the market requires an investment you have not made, and no amount of persistence substitutes for it.

We run market entry campaigns as a separate build: national register data, place data, channel chosen to fit the local rules, separate sending infrastructure, and native-language copy. Flat pricing, 3,750 EUR the first month and 2,850 EUR a month after that, cancel any time.