Running GTM for a CEE B2B SaaS across five markets at once
A Central European vendor selling into the US, UK, LATAM and the Gulf pays a credibility discount before the first call. What worked at SentiOne.
GTM Architect & Growth Operator · Now · 12 November 2026 · 5 min read
On this page
A B2B SaaS company in Warsaw builds something good and sells it well at home. Then it decides to go international. The plan is usually a list of countries and a hiring roadmap.
The obstacle is not on that list. A buyer in Chicago has no mental category for a vendor from Central Europe. Not a negative one, an absent one. Everything you claim gets quietly discounted until something local vouches for you, and no amount of outbound volume fixes an absent reference point.
I ran this as Global Head of Growth at SentiOne from 2020 to 2024, across the EU, the UK, the US, LATAM and account-based work in the UAE. Over those four years new customers grew 80%, the brand 60% and organic traffic 50%.
The discount nobody puts in the plan
A vendor from an unfamiliar region pays a credibility tax on every first conversation. You can either design around it or spend three years paying it in longer cycles and lower win rates.
Five markets, one cadence
The instinct is to build a team per country. It feels like respect for local difference. In practice it produces five strategies, five sets of messaging, five reporting formats, and nothing that compounds. Market three learns nothing from market one because there is no shared surface for the lesson to live on.
What worked was the opposite shape: one operating cadence for every market, with a shared weekly rhythm, shared pipeline definitions and one experiment log. We adapted locally at three layers and nowhere else.
How the operation is shaped
A team per market
Each region writes its own positioning, runs its own experiments and reports in its own format. Nothing transfers. The head office arbitrates between five versions of the truth.
Costs scale, learning does not
One cadence, three local layers
Shared positioning, product narrative, reporting and rhythm. Local proof, local problem language, local buying process. A test that works in the UK is testable in LATAM next week.
Learning compounds across markets
Proof, problem language and buying process stay local
Proof. A German enterprise buyer wants a German reference. This is the layer where the credibility discount is actually paid down, and it is why the first customer in a market matters more than the next ten. Everything else is easier once someone local has said yes on the record.
The language of the problem. This is a deeper job than translation. The same product solves a problem that different markets describe differently, and buyers search for their own description of it. This is where category content and analyst education did the work at SentiOne: being present in the vocabulary the market already uses.
The buying process. Procurement in a Gulf enterprise, a US mid-market deal and a Nordic public tender are three different machines. The material has to fit each one, even when the pitch behind it is identical.
Everything outside those three stays shared. Positioning, product narrative, pricing logic, reporting definitions and the weekly rhythm. The moment those fragment, the markets stop being comparable, and comparability is the only advantage a small international team has.
The order it went in
- Fix the shared layer first. Positioning, narrative and reporting definitions before any market opens. Retrofitting these across five regions costs several times what setting them once does.
- Win one reference per market before scaling spend. The first local customer is a positioning asset, not just revenue.
- Enter the vocabulary, not just the search results. Category content and analyst relations move the reference point in a way performance marketing cannot.
- Keep one experiment log across all regions. A landing page or offer test that works in one market is a candidate everywhere. The landing page and offer tests behind the 20% sales lift ran through that log.
- Use ABM where the market is relationship-shaped. The Gulf work was account-based because the buying process there rewards depth over reach. That decision belongs to the buying-process layer and leaves the strategy untouched.
- Report every market the same way, weekly. Different numbers, identical definitions.
Do
- Treat the first local reference as a positioning deliverable
- Keep positioning, narrative and reporting identical across markets
- Localise proof, problem language and buying process, and nothing else
- Run one experiment log the whole operation reads
- Match motion to how the market buys, including account-based where it fits
Don't
- Open a market with outbound volume and no local proof
- Let each region write its own positioning
- Treat localisation as translation of existing pages
- Compare markets whose metrics are defined differently
- Assume the motion that works at home travels unchanged
What to watch
The fourth line is the one worth defending in a budget conversation. An operation that transfers eight experiments a quarter between regions is compounding. One that transfers none is five companies wearing the same logo.
If you are planning an international move out of Central Europe, the first question is not which market. It is who in that market will say your name out loud, and what it takes to get them.
