How to validate a wellness product in a new country before launching

Why "it works at home" proves nothing abroad

Most wellness and preventive health products get their first users from people who know the founder: friends, former colleagues, a partner clinic. That is a fine way to reach a first version. It is a terrible way to predict a new country, because every one of those users had a reason to be kind, and none of them lives where you are going.

The failure pattern is well documented. In the Startup Genome study of 3,200 technology startups (2011), 70% scaled before they were ready, and none of those passed 100,000 users. CB Insights' review of startup post-mortems (2021) found that 35% of failed companies had built something the market didn't need, and 38% ran out of cash, usually a few months after the first problem showed up. An international launch combines both risks: a market you don't know, and a burn rate that doubles the day you hire locally.

So the question is not whether to validate before launching. It is which method answers the question you actually have.

Start with the question, not the method

Founders tend to pick a validation method by habit or by budget. It is more useful to pick it by the decision it has to support. Three decisions come up again and again:

• Go or no-go on one country. "Is Germany ready for this product as it is?"
• Which country first. "We could go to the UK, Sweden or the Netherlands. Where will it land best?"
• What to change before launching. "Which parts of the product travel, which need adapting, which should we drop for this market?"

Each decision needs different evidence, and no single method produces all three. Here are the four methods founders use, with what they are good at.

Method 1: market research

Desk research, industry reports, competitor mapping, regulatory review, sometimes a survey. It tells you how big the market is, who already serves it, what rules apply, and what people say they want.

What it cannot tell you: whether anyone will use your product. Market research describes the market; it says nothing about product fit. It is necessary before any expansion and sufficient for none.

Best for: ruling out countries early (regulation, market size, a dominant incumbent) and preparing the questions a later test should answer.

Method 2: user research panels

Recruited participants, in the target country, in one-hour sessions: interviews, usability tests, concept reactions. Platforms such as User Interviews, Respondent or Prolific make this fast and relatively cheap.

What it cannot tell you: whether people come back. A panel captures what people think when a researcher is watching, on a day they were paid to pay attention. Wellness products live or die on week three, when nobody is watching. Panels are also prone to politeness: participants tend to say what they expect you want to hear, in every culture, in different ways.

Best for: catching obvious localisation problems (language, onboarding, cultural misreads) before a longer test, and understanding why people react the way they do.

Method 3: a soft launch

Put the product live in the new country, run some ads, see what happens. It is the most "real" option, and the most expensive to read: you are measuring your acquisition channel as much as your product, and the two are confounded. A weak result could mean the market doesn't want the product, or that your ads didn't reach the right people, or that your store listing was wrong. You won't know which.

What it cannot tell you: why. And it costs the launch to find out. If the soft launch disappoints, you have already spent on localisation, listing, support and ads.

Best for: products with a proven channel playbook that transfers between countries, usually later-stage companies with a local team already in place.

Method 4: an in-market usage test

Recruit real users in the target country who match your profile, have them use the product for several weeks, measure what they do, and ask them what they think at the end. The point is time: eight weeks of usage shows habit, not curiosity. It separates "interesting" from "I still use it".

What it cannot tell you: your acquisition cost. The testers were recruited, not acquired. It tells you whether the product holds once people have it, not how expensive it will be to get it into their hands.

Best for: the go / no-go decision on a country, and for ranking several countries on the same evidence. This is what Eldwin does, and the reason it exists: we found no other way to answer "will people here keep using it" without launching.

Side by side

In practice the methods stack. Market research to shortlist countries; a panel or a usage test to pick and prepare; a launch once the evidence says go. The expensive mistake is skipping the middle step.

Six questions to ask before you commit

Whatever method you use, the launch decision should be able to answer these. If it can't, you are launching on hope.

1. Have real users in the target country used the product, for more than a session?
2. Did those users come from outside your personal network?
3. Has anyone in that market agreed to pay, or set aside budget?
4. Do you know which consumer and data rules apply there, especially for health data?
5. Does the product fit local habits: language, daily routines, the channels people trust?
6. Did you set the go / no-go threshold before spending on the launch?

A note on health data

If your product touches sleep, mood, pain, activity, weight or anything that reveals a person's state of health, usage data is "data concerning health" under GDPR and most equivalent laws (UK GDPR, Korea's PIPA, Japan's APPI, Washington's My Health My Data Act). That raises the bar for any validation that involves real users: explicit consent per test, pseudonymisation from day one, a data protection impact assessment, and a processing agreement between whoever recruits and whoever owns the product. Any provider who can't explain this in two minutes should not be running your test.

What a good validation plan looks like

One page. It names the country, the user profile, the question the test must answer, the evidence that would count as a go, and the date by which you decide. It is signed before anything starts, so that the result cannot be argued with afterwards. Most of the value of validating is in writing that page; the test only fills in the numbers.