Building technology in Uganda, for Ugandans

Most technology advice I read was written for a world I don’t build in. It assumes fast internet, a customer with a credit card, a warehouse that runs on address lookups, and a user who will happily download a 60MB app and sign up with an email. Almost none of that is true for the person I am usually building for — a trader in Kikuubo, a boda rider, a mother running a shop in Kawempe, a woman-owned MSME in Gulu.

So over the years I stopped trying to copy those playbooks. I started building for Ugandans as they actually are. That single decision — ⓘ build for the constraint, not around it — has shaped almost everything I’ve shipped since.

My clearest lesson came from Minute5, the on-demand grocery and retail business I co-founded and ran as CEO and CTO from 2019 to 2024. On paper it was a delivery startup. In practice it was a five-year education in Ugandan reality. We grew it from zero to over UGX 500M MRR, past 2,000 customers, across four markets, and raised only $30K to do it. But the numbers are not the lesson. The lesson is in how we had to build to get them.

Take payments. The Western startup default is card-on-file. In Uganda, most of my customers didn’t have a card — they had a phone and mobile money. So we built around MTN and Airtel money as the primary rails, not an afterthought bolted on at checkout. ⓘ If you make people leave their money where it already lives, they don’t convert. You go to the money. The money doesn’t come to you.

Then there was the warehouse. Formal addresses barely exist in much of Kampala — you navigate by landmarks, not street numbers. We ran a dark store model and built our logistics around how the city actually works, not around a map that assumes plot numbers. With an AI order-prediction model we pre-positioned stock and cut delivery time by around 60%. And because margins on groceries are painfully thin, we couldn’t waste a single conversion — a suggestive-cart feature that recommended items at checkout lifted our average order value by about 20%. In a thin-margin business, that 20% is the difference between a company and a hobby.

I’ll be honest about the hard part too. Self-hosting to keep costs down, squeezing hosting bills, keeping payroll under control — that discipline isn’t glamorous and nobody writes blog posts about it. But building in Uganda means your unit economics have to work at Ugandan prices, not at prices subsidised by a Silicon Valley round you’re never going to raise. ⓘ If the maths only works after a $10M raise, it doesn’t work.

The same principles keep showing up in everything I’ve built since.

At Crested AI, we’re putting a civic information assistant on WhatsApp — not a slick standalone app — because WhatsApp is where Ugandans already are. You don’t teach a country a new interface; you meet it inside the one it already opens every morning. At Averor, we’re building an ESG scoring platform for African agribusinesses, because the compliance tools designed for European firms simply don’t fit a coffee cooperative in eastern Uganda.

If I had to compress all of this into one law, it would be this: build for the constraint, and the constraint becomes your moat. Offline-first, chat-first, mobile-money-first, low-bandwidth by default — the things that look like limitations are exactly what makes a product fit here, and fit is something a foreign competitor can’t copy from a distance. They’re optimising for a market that isn’t ours. We’re building for the one we actually live in.

I don’t think Uganda needs to wait for the world’s infrastructure to catch up before it builds serious technology. I think we build serious technology because of our conditions, not in spite of them — and some of what we’re forced to figure out here (mobile money years before the West took it seriously, WhatsApp-native commerce, doing a lot with very little) the rest of the world eventually copies. We’re often early, not behind.

This is the work I intend to keep doing: building things that actually work inside African constraints, and running real businesses around them — impact with a business model, not charity. If you’re building for this market too, or you have feedback, ideas, or something you want to build, reach me at paul@katumbapaul.com. I’d genuinely like to hear from you.

To everyone building in this part of the world with thin resources and thick ambition — we’re not behind, we’re building the playbook. Let’s keep going. 🇺🇬💪