The question "which international markets should we enter?" gets answered in the worst possible way in most companies. Someone comes back from a conference in Berlin with thirty business cards. A competitor announces a London office. A partner sends over a warm intro from a Singapore distributor. An investor says the Nordic market is "underpenetrated." These are treated as market selection signals.

They are not signals. They are noise.

The difference between signal and noise is whether the data point tells you something about the actual demand for your specific product, in that specific market, from buyers with the specific problem your product solves, at a price your business can actually capture. Everything else is vibes.

The Framework: What Real Demand Looks Like

A real demand signal has specific components. Build your market assessment around these:

Addressable TAM, not TAM. Total addressable market figures are almost always inflated in international contexts. The relevant question is not "how big is the market?" but "how big is the part of the market we can actually reach and serve?" A €5B market where your product requires 12-month enterprise sales cycles, local support infrastructure you don't have, and regulatory certification you haven't obtained is not a €5B addressable market for you.

Buying motion complexity. Some markets buy software in a specific, well-defined way that maps to your sales motion. Others have entirely different procurement norms — long-standing vendor relationships that are hard to displace, government procurement processes that require local entity presence, cultural norms that make cold outreach ineffective regardless of product quality. The question is not just whether buyers exist, but whether they buy the way you sell.

Competitive saturation and differentiation. If the market has three entrenched incumbents with 80% combined share and switching costs built on years of integration, "we have better technology" is not a go-to-market strategy. The relevant question is whether your specific differentiation maps to a real buying trigger in that market — not whether it's technically superior in the abstract.

Regulatory accessibility. Some markets require specific certifications, local data residency, local entity presence, or regulatory approvals before you can sell. These are not deal-breakers — they are cost and time inputs that belong in the model, not surprises that appear mid-launch.

The resource question. Can you win in that market with the resources you actually have, or only in theory? Every market selection decision should be stress-tested against realistic resource constraints. A market that's theoretically addressable but requires a local sales team of eight to compete is not a market you can enter with two missionaries and a partner agreement.

What Good Market Selection Looks Like

A company that does market selection well starts with a clear model of their ideal customer profile — not just company firmographics, but the buying motion, the competitive context, and the proof points that close deals in their category. Then they look for markets where that ICP exists in sufficient concentration, with a buying motion they can execute against, without requiring resources they don't have.

This produces a short list — often two or three markets, not twelve. The temptation is to think this is too narrow. It isn't. Entering three markets where you can actually win is better than entering eight markets where your presence is too thin to matter and your product isn't ready to support anyway.

The goal is not to be present everywhere. The goal is to win where you're present.