Hungry Enough to Build

It's 1 a.m. You have class at 8, but there's a bug in your code and your brain won't let it go. Then it clicks, you fix it, and only then can you sleep.

That stubborn feeling is exactly where startups begin. The only difference is that the bug isn't in your code. It's in the world: the portal that crashes every registration week, the shop still tracking debts in a notebook, the transfer that takes three days to arrive.

Most people complain about bugs like these. Founders can't stop thinking about them. And if you already know how to write code, you're closer to fixing one than you think.

Irritation or starvation

I believe the need to be a founder starts with one of two things.

Irritation is when something exists but it's broken, slow or too expensive, and it bothers you more than it seems to bother anyone else.

Starvation is when the thing you need doesn't exist at all where you live. There's no worse option or cheaper option. There's no option. So you have to build it yourself.

A good example from Addis is Gebeta Maps. Its founders were trying to launch a food delivery startup. A delivery app needs maps, and the obvious choice was Google Maps. But Google charges in dollars, and for a small Ethiopian startup earning in birr, that cost was painful and complicated. That frustration led them to start Gebeta Maps in 2021, a local mapping and location platform. They set out to build food delivery. The problem they hit along the way turned out to be the bigger business.

Notice what happened. They didn't sit in a room asking "what startup should we build?" They were already building, they hit a wall, and the wall became the idea.

That's why "I want to start a startup, I just need an idea" is usually the wrong starting point. Ideas invented from nothing tend to be solutions looking for a problem. Ideas that come from a real frustration come with their first user already attached: you.

So before asking what should I build, ask: what keeps annoying me, and who else does it annoy?

Marketing matters as much as code

Here's the uncomfortable part. As software engineers, we're trained to believe good code wins. Write it clean, make it fast, ship it, and users will come.

They won't. Not on their own.

Many technical founders spend six months perfecting features, then one afternoon "doing marketing" by posting a link. When nobody signs up, they decide the idea was bad. Often the idea was fine. Nobody knew it existed, nobody understood why it mattered, and nobody trusted it enough to try.

That's what marketing is: helping the right people discover your product, understand it, and trust it. For an engineer, building is the comfortable part. Talking to strangers, explaining your product in one sentence and handling rejection is the part that feels unnatural. It's also the part that decides whether you have a company or a side project.

Look at Stripe, the payments company started by brothers Patrick and John Collison while they were students. Their product was built for developers, by developers. You'd expect them to just put up great docs and wait. Instead, when someone agreed to try Stripe, they asked for that person's laptop and set it up for them on the spot, rather than sending a link and hoping. At Y Combinator, the move became so well known they named it the "Collison installation."

For a student founder, marketing is mostly free:

• Talk to users before you code. Ten honest conversations can save you months of building the wrong thing.
• Explain it in one sentence. If you can't, your users can't explain it to their friends either.
• Build in public. Share progress on Telegram, LinkedIn, X or TikTok. Your first fans are people who watched you build.
• Go where your users already are. Campus groups, Telegram channels, the merchants on your street.
• Treat your first 100 users like gold. Answer every message and fix their complaints fast.

A decent product that people know about will beat a brilliant product nobody has heard of.

You don't have to wait until graduation

A lot of students think entrepreneurship is for "later," after a degree and a few years of experience. The people above didn't wait, and neither did one of Ethiopia's best-known young tech founders.

Betelhem Dessie grew up in Harar and taught herself computers in her father's electronics shop. She was writing HTML by age 10. At 18, she was already a project manager at iCog Labs while still a student at Addis Ababa University.

Her frustration was simple: Ethiopian kids had talent but nowhere to learn coding. So she founded Anyone Can Code, which teamed up with iCog Labs in 2017 to open a coding academy for young learners. She has since been credited with teaching around 20,000 children to code.

She didn't have a big office, a degree or investors when she started. She had a skill, a problem she cared about, and the willingness to begin before she felt ready. As a student, that's all you need too.

Think like a founder, not just a developer

Starting early is one thing. Lasting is another. The moment you decide to build a startup, you're no longer only a developer or a student. You're responsible for a business. That shift asks you to think about things no course in your curriculum covers.

People pay for outcomes, not code. Nobody cares that you used microservices or the newest framework. A shop owner cares that they get paid faster. A student cares that they get their class schedule without the portal crashing. Learn to describe your work in terms of what changes for the user. The technology is how you deliver it, not what you're selling.

Money is part of the product. Ask early: who will pay for this, how much, and how? A product with thousands of free users and no plan to earn can quietly die. Try charging something, even a small amount, sooner than feels comfortable. Whether people pay is the most honest feedback you'll ever get. Learn the basics too: what it costs you to run, how long your savings last, and the difference between revenue and profit.

Rejection is data, not a verdict. Users will ignore you. Investors will say no. Friends will tell you the idea won't work. As engineers we're used to problems with clear right answers, so "no" can feel personal. It isn't. Each no tells you something about your product, your pitch or your audience. The founders who last are the ones who learn from rejection without letting it stop them.

Your team matters more than your idea. Ideas change. Most startups end up building something different from what they started with, just like Gebeta Maps did. What stays is the people. Choose co-founders you trust and who bring skills you lack. Agree early, and in writing, on who owns what and who does what. Many promising startups fall apart over friendships that never had those conversations.

Understand the ground you build on. A startup doesn't live only in code. It lives in a country with its own rules, payment systems and customer habits. Learn how businesses are registered, how people actually pay, and what regulations apply to your field, especially if you touch money, health or personal data. In a market like Ethiopia, knowing these realities is often a bigger advantage than any feature.

Take care of yourself. Building a startup while studying is a marathon, not a sprint. Burning out in your second year helps no one. Protect your sleep, keep up with your studies, and remember that the startup is something you're building, not the whole of who you are.

None of this replaces your technical skills. It's what turns those skills into something that lasts.

What to do if you want to build a tech startup as a student

Being a student is one of the best times to start. Your living costs are low, you're surrounded by potential users and co-founders, and failing costs you little. Here's how to use that.

1. Collect problems, not ideas. Keep a running note of every time something wastes your time or money, or someone around you complains. Your campus, your family and the businesses near you are a free source of real problems.
2. Validate before you build. Pick one problem and talk to at least ten people who have it. Ask how they solve it today and what it costs them. If nobody cares much, move on early.
3. Ship small and fast. Your first version can be a Telegram bot, a simple web app or even a Google Form. If it takes more than a few weekends, it's too big.
4. Learn more than code. Your degree teaches you to build. A startup also needs product thinking, design, basic finance and, above all, selling. Read, take short courses, and practice explaining your work to non-engineers.
5. Market from day one. Talk about what you're building before it's finished. Every post, demo and conversation is a chance to find your first users.
6. Find complementary co-founders. Two engineers often build a great product nobody hears about. Look for someone who enjoys talking to customers, selling or design.
7. Get into the ecosystem. Join student tech clubs, hackathons and hubs like iceaddis, which has long focused on university students. Mentors and early investors are found there.
8. Build a portfolio either way. Even if your startup fails, you'll graduate with a real product, real users and real lessons. That's worth more in an interview than any grade.

The worst case isn't failure. The worst case is graduating having never tried.

The bug that won't let you sleep

Remember that 1 a.m. bug? The one you couldn't stop thinking about until you fixed it?

The world is full of bugs like that. Slow processes, missing services, costs that make no sense. Most people have learned to live with them. They've marked them "won't fix" and moved on.

You're a software engineer. You already know how to stay with a problem until it gives in. Entrepreneurship asks you to point that same stubbornness at a bug in the real world, fix it, and then make sure the people who have that bug actually hear about your fix.

So find the bug that keeps you up at night. Then go fix it. You don't need to wait until graduation to push the first commit.