How to pick your first international market when everyone has an opinion

Everyone has an opinion about your first international market. Your investor wants the US because that is where the rest of the portfolio went. Your engineering lead grew up near Frankfurt and knows the buyers there. A founder friend closed one deal in the Gulf and now swears by it. The opinions are confident, they contradict each other, and they are all built the same way: from someone else's anecdote.

That is the actual problem. Not that people disagree, but that every position in the room is a compressed story about a different company, at a different stage, selling a different product. An anecdote cannot be argued with, only matched with a louder one. A decision this expensive needs criteria, scored against your company, your ICP and your runway. This post gives you the framework to do that yourself. And if you later bring in outside help, the same framework doubles as your interview script: anyone serious about international market entry for SaaS companies should be able to walk you through a version of it unprompted.

One more reason to take the decision seriously before you take it at all. The most expensive outbound mistake is not bad copy. It is a year of good execution aimed at the wrong market. The failure mode is familiar: four markets on a whiteboard, no agreed way to choose, so expansion becomes a side project owned by whoever has the least on their plate, and two quarters later there is a spreadsheet of companies and no meetings.

The six criteria that predict traction

Score every candidate market against the same list. Not against vibes, not against where you already have one logo, and not against market size reports, because a huge market you cannot reach is smaller than a modest one you can.

1. Accounts that genuinely fit your ICP. Not the TAM slide. The count of companies in that market that match your written ICP closely enough that you could name them, list them and defend each one. A market with a small number of perfect-fit accounts can beat a market with a huge number of vague ones, but you have to know which situation you are in before you commit.

2. Existing spend on the problem. Are buyers there already paying for something that solves this, even badly? Existing spend means the budget line exists and the pain is acknowledged. No spend means your outbound has to create the category before it can sell into it, which is a different and much slower motion.

3. Local category crowding. Who is already established there, including the incumbent your buyer will compare you to whether or not you consider them a competitor? A crowded category is not automatically a no. It changes what your messaging has to do and how sharp your wedge needs to be.

4. Sales cycle length. Procurement culture, compliance review, the number of stakeholders who can slow a deal. A market where deals close inside your runway is worth more to you right now than a richer market where they do not.

5. Cost to reach a buyer. Data quality for list building, channel norms, whether buyers there answer cold email, LinkedIn or the phone, what events cost, whether you need local language. Reach cost varies far more between markets than most founders expect, and it compounds across every account you touch.

6. Whether you can support it today. Timezone overlap with your team, the compliance requirements you can honestly meet right now, language coverage for calls. Today is the operative word. A market you could support after three hires is a different market, and it should be scored as one.

You do not need precision on any of these. You need enough evidence to rank. A rough count with a source attached beats a polished estimate with none, and the discipline of writing the source down is what keeps anecdote from sneaking back in.

One market properly beats three at once badly

The pull towards launching in several markets at once is strong, because it feels like hedging. It is not hedging. It is dividing.

Every market needs its own list, its own messaging, its own sequences and its own reply handling, because the buyer is different in each one. Split a small team across three markets and you do not get three experiments, you get three underpowered programmes, none of which sends enough or learns enough to tell you anything. Then the readout says outbound does not work, when what actually happened is that nothing was tried properly anywhere.

Concentration also compounds. Reply data from one market rewrites your messaging for that market week by week. Spread thin, the signal from each market is too weak to act on. Pick one, run it properly, and let the second market inherit a tested playbook instead of a guess.

The buyer is not the same person in every market

A title that owns budget in San Francisco often does not in Frankfurt or Riyadh. The same product can be bought by a different function, signed by a different level and blocked by a different stakeholder depending on the market, and the objections that come up in one market may not exist anywhere else.

So for your shortlisted markets, map the buying committee before you finalise the ranking: who evaluates, who signs, who blocks, and what each of them is measured on. This feeds directly back into the scoring. A market can look strong on account count and existing spend, then drop in the ranking because its committee structure doubles the sales cycle, or because the person your messaging targets turns out not to hold the budget there at all.

This is also why imported messaging quietly breaks. What converted at home was written for one committee's incentives. Do not judge a new market by how your old messaging performs in it.

The output is an ordered list with reasoning attached

Not a heatmap. A heatmap of green and amber cells looks rigorous and settles nothing, because it hides the argument inside the colouring and everyone keeps their prior opinion.

The output that actually ends the debate is an ordered list. Each market gets its rank, the reasoning behind the rank, the evidence that is thin, and the finding that would change the position. Now disagreement has to attach itself to a criterion and a fact rather than to a feeling, which is the whole point. Most of the time the exercise also makes the decision feel smaller than it did on the whiteboard, because two of the four candidates fall away on support or cycle length before the interesting comparison even starts.

Then give the winning market a 90 day plan: channel mix, target account universe, the volumes and tools required, what to measure in weeks two, six and twelve, and the specific signals that should make you double down or pull out. Write the pull-out signals before you launch, while nobody is defending sunk cost.

Decide fast, and make the decision reversible

This scoring does not deserve a quarter of deliberation. The data you need is gatherable in weeks, and the ranking it produces is a bet, not a verdict. What makes the bet safe is not certainty. It is reversibility.

You make it reversible by how you enter. An in-market sales hire is a long bet on one person at a loaded cost you feel immediately, made before you know whether the market wants what you built. An outbound motion you can redirect is a different shape of commitment: if the signals you wrote down say the market is wrong, you change direction in weeks, with a tested playbook and a scored second choice already waiting.

That is the standard to hold anyone to, including us. If you are evaluating a go to market agency for AI startups, ask them to show you how they would rank your candidate markets and what evidence would make them tell you to pull out. An answer with criteria in it is worth hearing. An answer with anecdotes in it is just one more opinion.

If you would rather run this exercise with people who do it for a living, book a call and bring your ICP and the markets on your whiteboard: you will leave with a view on which one is first and why.