Bangladesh must dream bigger

The easiest conclusion from the startup funding downturn is that founders dreamed too big. It reminds us the value of prudence. However, it is the wrong conclusion.

Bangladesh has allocated too little capital behind businesses that could teach us how to compete in the 21st century and help us escape the overcrowding trap. The tenth hawker shop on the same street does not create value. It takes the same footpath and the same customers, and divides them across more sellers. What looks like entrepreneurship is actually an overall increase in costs.

Similarly, the hundredth rice mill duplicates capacity. The next factory competes for the same energy supply, the same port capacity and the same buyers. At some point, more investment stops creating wealth and starts destroying it.

This capital allocation problem is the deeper root of our banking crisis. A financial system that only funds familiar businesses will eventually produce bad loans, idle assets, and disappointed depositors. Capital evaporates as excess power capacity, inflated land and empty buildings. The amount of non-performing loans changes with rescheduling policy, but the direction is clear: enormous capital is trapped in assets that did not produce value.

Hence, the recent moral panic about startups is misplaced. We believe startups are uniquely risky, but capital has been wasted across Bangladesh in conventional places. Machines rusting, plants sitting idle and buildings decaying. Government projects overrun budgets and increase the country’s debt burden. Bad loans sit inside banks for years, weakening the financial system.

Against that, the startup sector produced much more than its critics admit. With paltry local financial support, it has trained engineers, digitized suppliers, attracted foreign capital and produced companies that changed how Bangladeshis live. Even where performance fell short, much of the loss was borne by foreign venture capital — money that would not have entered the country otherwise. By contrast, domestic capital flowed to conventional sectors and cratered our savings.

Software behaves differently from physical assets. It can fail too, but when software works, it compounds. A system built once can serve millions at near-zero marginal cost. Data gets better with use. Networks strengthen as they grow. A habit, once formed, can reshape an economy. That’s why bKash matters. It didn’t just process payments. It changed how Bangladeshis think about money. It built rails other businesses run on. bKash’s value isn’t just its profit. It upgraded the country’s operating system.

This points to something Bangladesh hasn’t internalized: the power law. In venture-backed innovation, a few companies create most of the value. Startups are experiments. Most don’t become large and some fail. But one success can redefine an industry, create new infrastructure, and return enough to justify many failures. This is not a flaw in the system. It is the system. Try to eliminate failure, and you eliminate extraordinary success.

The power law isn’t an excuse to fund everything blindly. It’s an argument for building a portfolio of serious attempts across stages. Seed capital to start. Early-stage capital to find product-market fit. Growth capital to scale, hire senior talent, improve governance, survive shocks. Exit markets to recycle capital into the next generation. Bangladesh has small pieces of this chain, but it hasn’t learned to visualize the full chain.

An ecosystem in much more than pitch contests, events and seed grants. It needs growth capital, debt products, public-market access, acquisition paths, and institutions that understand a company can be valuable because of customers, software, and recurring behavior — even if it owns no land.

Innovation doesn’t come from spreading money thin so everyone gets a little. It comes from funding many experiments, then making disproportionate bets on the few with real traction. Power-law outcomes require power-law allocation. If promising company receives only enough survival capital but not enough to scale, the country will produce many wounded ventures and no category-defining institutions.

The countries that built companies like Amazon, ASML, Apple, Nvidia, Alibaba, and Mercado Libre — among the most valuable companies in the world — did not do so by avoiding risk. They financed experiments, many of which failed, and few exceptional ones transformed their economy. The modern world isn’t built by the safest borrowers. It’s built by people trying to make something that didn’t exist before.

Bangladesh’s startup investment — ~0.03% of GDP — is shameful. Per capita, we’re off by orders of magnitude relative to our peers. A country that spends this little on experimentation isn’t careful. It’s risking its future. The real question isn’t “how many startups failed?” The question is “where else should the capital have gone?” More land speculation? More factories competing for constrained infrastructure? More public projects where costs rise but productivity impact remains unclear? We already know how those stories end.

Our success in garments deserves respect. However, it cannot be the ceiling of ambition. A country that exports low-margin labor and imports high-margin systems stays structurally behind. The arithmetic is brutal. A wide-body aircraft costs hundreds of millions of dollars. How many man-years of $120/month factory labor pays for one plane? This isn’t an argument against low-cost manufacturing. It’s an argument for moving up the value chain — from labor to sophisticated systems.

The 21st century runs on systems: payments, logistics, cloud, digital identity and AI. They don’t just serve customers; they reorganize sectors Calling startups “apps” misses the point. bKash is financial infrastructure. A logistics company is trade infrastructure. A health platform can be public-health backbone. If Bangladesh doesn’t build these systems, it will import them. With that, we will also import dependency.

A fair question to ask: can Bangladesh even build these systems? Building them require culture, discipline, product taste, engineering depth and patient capital. Which is why we must start now. Knowledge has never been more accessible. A motivated Bangladeshi can learn design, finance, logistics, AI from the same global base as someone in Singapore or SF. The missing ingredient isn’t intelligence. It’s focus, apprenticeship, ambition, and institutions that reward serious work.

We have millions of young people. Too many are underemployed, misemployed, or waiting for a narrow set of government or corporate jobs. Educated unemployment hurts most: we paid for education but didn’t convert it into capacity. If we don’t give these people hard problems, they’ll sit idle, or crowd into low-productivity work that creates negative wealth. But if we train them on real systems, and they become the creative talent we desperately need.

Bangladesh’s problems are big. In a dense country, coordination itself is the hardest economic challenge — something software is uniquely capable of addressing. Healthcare, food, logistics, credit, education, tax, energy, congestion: these are system-level problems. Small businesses are essential for livelihoods, but they cannot solve national coordination. Only new systems can, and building these systems require capital, execution, and belief.

It’s easy to say founders scaled too fast in 2020–2021. Easy to say they should have predicted the global funding crash, regime change, forex crisis and capital flight. Hindsight is perfect. At the time, capital was abundant, digital adoption was accelerating, and investors pushed growth. Surely some things could have been more efficient. However, the lesson should not be that ambition was misguided.

The lesson: pair ambition with better financial architecture, stronger discipline, deeper local capital. Startups were largely funded by foreign capital, which retreats when politics wobbles or global liquidity tightens. Unfortunately, Bangladesh does not have the financial imagination required to support innovation. Banks lends against land and guarantees, not customers, data, retention, cash flow. Public markets are closed to growth companies. So, startups are too modern for local finance, too frontier for global capital when cycles turn.

The deeper problem is that Bangladesh has not yet learned how to be a friend to its own builders. When the environment became tough, the country did not respond with speed, ownership, or structure. Its institutions moved through committees, approvals, and unclear mandates. But innovation cannot happen at bureaucratic speed. Companies are living organisms. If support is late, teams break, customers leave, suppliers lose trust, and knowledge scatters. When the country fails to take ownership at difficult moments, it doesn’t simply lose a few startups. It breaks the compounding by which an economy learns.

That’s not proof startups were a mistake. It proved that the ecosystem around them is incomplete. In good times, we celebrate startups. In bad times, we call them someone else’s problem. We cannot build new industries with that mindset.

If we retreat now — conclude that ambition was the problem — the consequence won’t be caution or financial discipline. It’ll be exit. Talented founders, engineers, and operators will build elsewhere: in Singapore, Dubai, London, or SF. Bangladesh will educate them, and then watch others capture the value. We lose not just companies, but the accumulated knowledge to build the next ones.

The discipline we need isn’t smaller ambition. It’s matching ambition with better execution and better institutions. More serious, not smaller. More rigorous, not more fearful. More demanding, not less imaginative.

Every serious startup leaves an imprint: trained people, changed habits, better tools, new expectations, and hard-earned knowledge about what doesn’t work. That knowledge compounds across generations and forms the ecosystem — not because every company wins, but because each attempt leaves the next one better informed.

So, the choice is simple. Keep allocating capital to the tenth shop, the hundredth factory, the next politically-connected loan, and the next idle machine. Or it can allocate more — much more — to experiments that might build the country’s future systems. The first path feels safe because it’s familiar. The second path is uncertain because it requires understanding the future. But that’s how innovation happens and countries move forward.

At 0.03% of GDP, Bangladesh isn’t overinvesting in startups. It’s barely trying. The 21st century won’t reward countries that repeat the 20th. It will reward those that build what others depend on. Bangladesh has to decide: stay permanently downstream — exporting effort, importing intelligence, calling that prudence? Or build.

Waseem Alim is a co-founder of Chaldal.