Building across the Caribbean, Africa & the US
The Proving Ground: Why We Build for a Country of 100,000 First
Kenroy George · 2026-08-20 · 7 min read
TL;DR: Conventional startup wisdom says to build where the market is biggest. Cari does the opposite: we build for small nations first, because a country of 100,000 people is the most honest test a product can face, because scarcity forces better engineering, and because trust earned there compounds in a way no ad budget can replicate. The Caribbean is our proving ground, not our boundary.
There is a piece of advice every founder hears early: go where the market is. Chase the biggest total addressable market you can find, because a small slice of an enormous pie beats a large slice of a small one.
We heard it too. We inverted it on purpose.
Cari builds for small nations first. Countries where the entire population would fit inside a single borough of a major American city. Here is why we think that is not a limitation but the strategy itself.
The honest test
In a large market, a product can be mediocre for years and survive. There is always another customer segment to pivot toward, another growth channel to paper over the churn, another press cycle to ride. Vaporware can live a long and comfortable life in a big country.
A small nation offers no such shelter. When you deploy software for a country of 100,000 people, one of two things happens: it serves real citizens, or it visibly fails. There is no anonymous middle. The system either works at the counter, in the clinic, at the ministry, or everyone knows it does not.
The feedback loops are correspondingly short. In a small country, you can sit with the permanent secretary who owns the policy on Monday and the shop owner who lives with its consequences on Thursday. The distance between the decision and the ground truth is one week and a short drive, not a quarterly report filtered through three layers of account management.
That compression changes how you build. You stop designing for a persona and start designing for a person you have met. Assumptions die fast, which is exactly what you want, because assumptions that die fast are cheap and assumptions that die slow are fatal.
Constraints are the best designers
Small markets come with hard constraints: connectivity that comes and goes, government budgets with no room for waste, teams of three doing the work that a large ministry would assign to thirty.
Most companies treat those constraints as reasons to stay away. We treat them as the best design partners we have ever had.
Intermittent connectivity forces offline-first architecture. If the software must keep working when the connection drops, you design for local state, graceful sync, and honesty about what the network can and cannot promise. That discipline does not become useless in a well connected country. It becomes resilience. Software that survives a rural Caribbean outage shrugs off a subway tunnel in New York.
Thin budgets force efficiency. When there is no money for oversized infrastructure, you learn to do more with less compute, less bandwidth, and less operational overhead. Those savings do not disappear at scale. They multiply.
Small teams force simplicity. When the person maintaining a system is one civil servant with many other jobs, the system cannot depend on a standing army of specialists. It has to be understandable, operable, and boring in the best sense. Simplicity chosen under pressure is the kind that lasts.
Products hardened this way are not worse versions of rich-market software. They are better versions. The constraints of a small nation are a preview of the failure modes every market eventually has.
Trust compounds differently
In a big country, reputation is something you buy: campaigns, sponsorships, growth loops. In a small country, reputation is something you are.
Word of mouth in a nation of 100,000 people is not a marketing channel. It is national coverage. If the software fails a shop owner on Tuesday, the ministry hears about it before the weekend. If it genuinely helps people, that travels just as fast, and it travels with names attached.
This is a harsher environment, and that is the point. Trust earned under those conditions is real, and real trust exports. Small nations watch each other closely. Governments in one country notice what worked for their neighbor, because their neighbor faces the same constraints, the same budgets, the same skepticism of tools built elsewhere for someone else. One honest deployment is worth more than a hundred pitch decks, and in a region of small states, one honest deployment is visible to all of them.
History agrees
This is not a novel bet. Transformative infrastructure has a habit of debuting at small scale before it reaches world scale.
Mobile money did not begin in the world's largest economies. It grew out of East Africa, where the constraint was real, the need was daily, and the existing system had left people out. It proved itself there first, then spread outward, and the rest of the world spent a decade catching up to what a supposedly small market had already made ordinary.
Digital-first government did not emerge from a superpower either. It famously took shape in a small Baltic state that treated its size as an advantage: few enough people to move fast, high enough stakes to be serious, small enough institutions to actually change. The playbook written there is now studied by governments many times its size.
The pattern is consistent. Small places are where the future gets debugged. Large places are where it gets copied.
Cari's arc
Our arc follows that pattern deliberately.
The Caribbean is where we prove things. It is where our AI, our government work, and our identity and payments infrastructure meet real citizens and real institutions with no room for theater. Every hard property of our stack, offline resilience, efficiency, operational simplicity, is being earned there, not claimed.
Africa is where the network already extends. Cari Market connects businesses across markets that share the same constraints the Caribbean taught us to respect, and the same appetite for tools that were actually built for them rather than adapted as an afterthought.
The United States is where the financial rails live and where the companion lives. It is a huge market, and we build for it, but we do not treat it as the origin of the product. We treat it as a destination the product arrives at already hardened, already trusted somewhere, already proven against conditions most American software never has to survive.
Common questions
Isn't a small market a ceiling on growth? Only if you stop there. The market defines where you prove the product, not where you sell it. Software that works in Antigua does not need to be rebuilt for Atlanta. Distribution scales. Proof does not need to.
Why would a US customer care that you deployed in the Caribbean first? Because the properties that survive a small-nation deployment, offline resilience, efficiency, simplicity, are exactly the properties large customers say they want and rarely get. Where a product was hardened tells you more than where it was marketed.
Does building for governments slow you down? It slows down the parts that should be slow: trust, correctness, accountability. It speeds up everything else, because a government deployment in a small country gives you direct access to the people your decisions affect, and that access is the fastest feedback loop in software.
The lesson we keep relearning is simple. Scale is a copy operation. Trust is not. The Caribbean is the proving ground, not the boundary, and if you want to see what we are proving there, you can read more about us at cari.global/about.