For six years I had one of the best vantage points in the world on a single, expensive question: why do education-technology companies — which raise enormous sums and promise to transform learning across whole continents — so rarely actually scale?
As managing director for a global EdTech company across the Middle East, Africa, and Central and Eastern Europe, I sat across the table from universities, ministries, and national innovation programs in thirty-seven countries. I watched companies arrive — several of them far better funded than anything I had ever built — carrying the same confident promise, and leave two or three years later quietly diminished. After you have seen enough of these, you stop noticing the individual failures and start seeing the pattern underneath them. That pattern taught me something about scale that I have built my entire current company around.
Thirty-seven countries is reach, not scale
Here is the first thing six years taught me, and it is fatal to the companies that miss it: almost everyone in this industry confuses reach with scale.
Reach is being present in thirty-seven countries. Scale is the thing working the same way in country thirty-seven as it did in country one, without being rebuilt by hand each time. On a map and a fundraising deck they look identical. Underneath, they are entirely different businesses. A great deal of what calls itself "global EdTech" is, in truth, thirty-seven local operations wearing a single logo — each one rebuilt, restaffed, and re-sold market by market. That is not scale. That is reach, paid for at full price, thirty-seven times over.
And companies die of it in two specific, recognisable ways. I came to think of them as two diseases.
The first disease: the localization tax
EdTech sells the dream of deploying everywhere. The part nobody puts in the pitch is that every new country demands the work be redone — the content rewritten, the examples re-grounded, the language translated, the regulatory fit re-established, the local partnerships rebuilt from nothing. That cost does not shrink as you grow. It grows with every market you enter. I watched company after company spend its growth capital not on growth but on re-localisation — rebuilding the same product for the next territory, and the next — and never reach the margin that scale was supposed to deliver. They were running up a tax that compounded faster than their revenue.
The lesson is brutal and simple. You cannot translate your way to scale. If adding a language or a country means rebuilding the thing by hand, you do not have a scalable company. You have a growing liability with excellent marketing.
The second disease: the services trap
At small scale, a high-touch, human-delivered service feels like a premium product. Clients love it; outcomes are strong. But at scale it becomes a treadmill: every new customer requires its own proportional dose of human delivery, so headcount tracks revenue almost one to one, and the margins that were meant to compound never do. The sector is littered with companies that called themselves platforms and were, underneath, consulting shops with a login screen. You genuinely cannot tell the two apart at ten customers. You find out, brutally, at ten thousand — when the model that felt premium reveals itself as a business that must hire a person for every increment of growth.
Both diseases have the same root
It took me longer than I would like to admit to see that both diseases are really the same disease. The localization tax and the services trap share one root: work that has to be redone, by humans, for every new unit of scale. A new country, a new cohort, a new language, a new thousand learners — if each of those requires the core work to be performed again by hand, the company cannot scale, no matter how much it has raised or how many flags sit on its website. It can only grow more expensive.
Why we made the architectural call on day one
When I started Simsy AI, that lesson was not an afterthought. It was the first architectural decision, made before almost anything else. The entire premise of an operating system — as opposed to a course, a service, or a program — is that the repeatable work is done by the system once, and then runs for one founder or ten thousand, in one language or any language, without being rebuilt. Adding a country is a configuration, not a reconstruction. Adding a language is a setting, not a six-month project. Serving the ten-thousandth founder costs almost nothing more than serving the first. That is not a feature we bolted on later. It is the reason the company is shaped the way it is — a direct, deliberate answer to the two diseases I spent six years watching kill better-funded companies than mine.
The track record that actually matters
People sometimes assume the value of a second-time founder is simply that they have scaled something before. I think that is the wrong lesson to draw, and I will be honest about why. Plenty of EdTech founders had scaled something before, and still walked their next company straight into one of these two diseases. Having done it is not the same as understanding why it breaks. The track record that actually matters is not the size of what you built last time — it is whether you can name, precisely, the ways the thing kills you, and whether you built the next one to survive them. I can name two. Simsy is built against both.
The startups of the future aren't destined to fail — and the companies built to help them aren't destined to stall at thirty-seven flags on a map. But only if we stop confusing reach with scale, and build the system that makes the difference between them.
— Satish Kumar N · Founder & CEO, Simsy AI · Dubai · Bengaluru · New York
Build. Launch. Scale.