Why Fish

July 2026

The honest beginning is that I wanted to start a company. There was no childhood pond, no family tradition of fishing, and no moment in which I looked at a tilapia and understood my purpose. I wanted a startup before I knew what it should make.

This is not an unusual way to begin. The desire to build can arrive before the object. You want the small team, the independence, the difficult decisions, and the possibility that something which did not exist yesterday might be useful tomorrow. Then you have to decide what the thing is.

Wanting a company

At first, almost any sufficiently ambitious idea could hold my attention. I considered making electric bicycles. Kenya has traffic, expensive fuel, a large transport economy, and many short journeys. An electric bicycle could be explained with numbers and placed inside a story about cleaner cities. It looked like the kind of thing a technology company might make.

I also thought about financial technology. M-Pesa is the unavoidable proof that a Kenyan product can change ordinary life at enormous scale. Once you have seen that happen, it is tempting to search for another transaction that could receive a touch of M-Pesa: make it digital, make it immediate, put a clean interface over it, and call the result a company.

Neither thought was dishonest. Electric transport matters. Financial systems matter. The problem was that I could exchange one idea for the other without losing anything. I was choosing the appearance of a company before finding the reason for one.

A pitch can conceal this problem for a long time. Give the idea a name, estimate the market, describe the technology, and place several arrows between the present and a much larger future. It begins to feel inevitable. But nothing has happened to a user. The company exists only as a description of itself.

The obvious place to look for an idea was the startup world itself. It offered an endless menu. There were companies delivering groceries in ten minutes, companies lending against tomorrow's salary, companies selling software to other companies selling software, and companies applying artificial intelligence to every noun available.

Y Combinator was particularly good at making this world visible. I could scroll through batches and see hundreds of answers to the question I was trying to answer. Each company came with a compact explanation: this problem exists, this market is large, and this team has found the wedge. It was education, but it was also a trap. The more examples I saw, the easier it became to imagine choosing one and rebuilding it here.

A copied product is rarely an exact copy. The currency changes. The payment method changes. The photographs show people who look local. The company says it is doing for Africa what another company did for America, Europe, India, or China. These changes can be necessary, but they do not guarantee that the product belongs where it has been placed.

The deeper assumptions often remain untouched. The product expects the same habits, the same margins, the same infrastructure, the same trust in institutions, and the same willingness to pay for convenience. It changes its address without changing its understanding of the person at that address.

Made in Kenya

Looking at Kenyan startups, I began to think that many of them failed before money ran out. They failed when they chose a problem that was not important enough to the people expected to use the solution. Capital, regulation, hiring, and distribution may have delivered the final blow, but the distance between the company and the user was already there.

We often say the market was not ready. It is a convenient sentence because it preserves the intelligence of the idea. The company was correct; reality arrived late. Sometimes that is true. There are useful technologies that must wait for infrastructure, cost, or regulation to catch up. But the sentence can also mean that people understood the offer and did not care enough to change what they were doing.

A company is not Kenyan because it was incorporated in Nairobi, raised money for an African market, or placed an M-Pesa button at checkout. Its product has to contain some knowledge of life here. It should understand what people already do, what they already trust, where money is lost, how goods actually move, and which inconvenience is serious enough that someone will pay to remove it.

This does not mean every Kenyan company must be small, traditional, or concerned only with Kenya. It means the first user cannot be an abstraction. A product built for everyone usually begins with an imaginary person who lives nowhere.

M-Pesa did not succeed because Kenyans needed to be taught to appreciate financial technology. People already needed to send money. Families were already spread between towns and rural homes. Cash was already moving through buses, friends, and informal arrangements. The product entered an existing river. It did not first have to convince people that water should flow.

Building the walls

I stopped looking for ideas and started building walls. If I was going to make something, it had to survive inside them.

The product had to be useful without a pitch. A grandmother in Nyeri and a bodaboda rider should be able to understand what was being offered. This was not because either person is unsophisticated. It was because a founder should not use sophistication to hide a weak exchange. If the user has to understand my technology, my market category, and my theory of the future before the product makes sense, I have given them my work to do.

It had to begin with something people already used. I did not want the company's survival to depend on manufacturing a new desire or persuading people to perform a new identity. The product could improve a habit, make it cheaper, or remove difficulty from it, but the habit had to be real before we arrived.

It had to respect the user's intelligence. Too many products approach Africans as people in need of correction: the farmer who lacks data, the trader who lacks formality, the household that has not yet learned the correct way to buy. The company enters as the intelligent party and describes everyone else by what they lack.

That relationship is usually false. A farmer may know more about an animal, a season, and a local market than anyone building software for them. A trader may be operating around constraints that disappear in a spreadsheet. Technology can add capability, but it should arrive with enough humility to notice the capability already present.

Finally, the product had to leave the user satisfied. In my thoughts on product design, I argue that a good product should leave a person in a better place than it found them. In my thoughts on the artificial person, I argue that a company becomes real through the useful thing it does for someone else. These were no longer separate ideas about products and companies. They became walls around the company I was willing to start.

Fish

Inside those walls, fish remained.

People already eat it. They do not need an Aqualabs account to understand it. A customer can judge the product without reading our claims: its size, freshness, taste, price, and availability are the argument. If it is bad, no explanation can rescue it. If it is good, the plate is enough evidence.

Food is as close to personal as a product can get. It enters the body, but before that it passes through memory, family, culture, appetite, and trust. People know how they like it prepared. They know what a fair portion looks like. They know when the price has moved beyond them. A food company cannot treat taste as friction in the adoption funnel. Taste is the product meeting the person.

Fish is also repeatable. The company does not have to create a loop that makes absence uncomfortable. People return because meals return. If the product is affordable, dependable, and good, repetition follows an existing need. We do not have to confuse engagement with value.

And fish is direct. At the end of all the systems around it, somebody eats it. The chain may include hatcheries, ponds, feed, water quality, transport, cold storage, markets, restaurants, and software, but it ends in an ordinary event. A person has a meal. That end keeps the rest of the company honest.

Geography and taste

Startup advice teaches founders to search for a moat: some durable advantage that competitors cannot easily cross. Software, proprietary data, network effects, and capital can all form one. In fish, the moat is more physical. It is geography and taste.

Fish lives somewhere. It grows in particular water, under particular conditions, close to some markets and far from others. Feed arrives by a route. Harvesting happens at a time. A delay changes quality. A broken cold chain is not a small technical inconvenience; it changes whether the product can be sold and eaten.

The customer also lives somewhere. Nairobi is not Nyeri, and neither is Kisumu. The fish people know, the size they prefer, the way they cook it, what they can pay, and where they expect to buy it are not details to be standardized after the company scales. They are the business.

Geography makes the work harder, but it also makes it difficult to imitate from a distance. A competitor can copy a website. It cannot immediately copy years of knowing farmers, buyers, roads, seasons, losses, and preferences. The advantage accumulates in attention to place.

Taste creates a second discipline. A company may optimize production until the numbers look excellent and still produce something people do not want to eat. Efficiency cannot overrule the customer's mouth. The user has the final word, and the final word is often simply whether they buy the fish again.

Aqualabs

Aqualabs grew from that decision. The name contains technology, and technology is part of the work, but technology is not the product the customer owes us attention for. Sensors, software, data, and better production methods matter when they make the fish better, the supply more dependable, the farmer's work more productive, or the price more reasonable. Otherwise they are decorations around the company we imagined at the beginning.

The standard is therefore plain. The fish must be good. It must be available. It must reach people at a price that makes sense. Farmers and others doing the physical work must participate in the value created, not appear in our story only as beneficiaries of innovation.

This will not make the company easy to build. Food has thin margins, living systems fail, logistics are unforgiving, and customers do not become patient because the founder has an ambitious vision. Those difficulties are not evidence that the idea is profound. They are the work.