Loading page…
Two founders leased one computer in a Nairobi cyber cafe for 4,500 shillings a month, then beat IBM and the banks to a 2.7 billion shilling tender to collect the city's parking fees, permits and rents — with no venture money at any point.
over KSh 1.3 billion collected for the county in the preceding three months · seasonal parking grew from 9,000 PSVs billed before the system to about 14,000 · single business permits issued online in under 15 minutes against about 4 months before · operating in 4 countries, self-funded with zero VC (all figures Kenyan shillings; the transcript gives no exchange rate, so none is applied)
"A business card costs about six shillings, and that's what it takes to be a business." The actual pitch was never software: a Nairobi single business permit used to take about four months, and on their system you generate it online in under fifteen minutes.
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
In early 2009 Danson Muchemi and a friend wanted to start a company and had no money for an office. So they leased one computer inside a Nairobi cyber cafe for 4,500 shillings a month and stayed there six months. "It's from that cyber that we've grown into the organisation we are now, in four countries," he says. The stated ambition from day one was a pan-African payments company operating in more than fourteen African countries. The business is JamboPay, an electronic payments provider. The moment that changed its size was a public tender: Nairobi County wanted to automate its revenue collection — parking, land rates, single business permits, rents, advertising and more — and JamboPay competed against global firms including IBM, local firms and banks, and won. The contract is valued at 2.7 billion shillings. In the three months before the interview the system had collected over 1.3 billion shillings for the county. He is precise about why a small local company beat much larger bidders: "understanding of the local market, a very agile and customised solution, a business model acceptable to this market, and being brave." The lightness was not a virtue they chose — "we deployed a solution that is very light because we did not have funds," he says, which forced research into something cheap to run. The outcomes he points to are civic rather than financial. Nairobi has no paper parking tickets left; you pay by dialling a USSD short code, at midnight, for tomorrow or for the month. Seasonal parking grew from 9,000 public service vehicles billed before the system to roughly 14,000. A single business permit that took about four months is issued online in under fifteen minutes. JamboPay has never taken venture capital: "it is all self-funded from our activities."
The first customers were not won at a desk; they were won by competing in a public procurement against IBM, international firms, local firms and banks, and being the bidder who understood how revenue actually moves in this particular city.
The quick-reference card — acquisition channels, replication playbook, risks & traps, and our verdict — is members-only.
Join the Unicorn Club — $5/moCancel anytime, in one clickPaste into Claude Code / Codex and get a working version of this product end-to-end.
Two production-grade prompts per idea — a full build spec and an SEO growth plan — plus 10 tailored build specs a week. Members only.
Join the Unicorn Club — $5/moCancel anytime, in one clickEvery claim traces back to a listed source — open any of them and check the number yourself.
Data credibility: Kenyan television interview, Citizen TV's Business Center, aired January 2015, with the founder answering on the record. Graded founder-reported because Danson Muchemi states the numbers himself about his own company and no filing, audit or third-party dataset appears anywhere. Stated by him: a computer leased in a cyber cafe for KSh 4,500 a month in early 2009 for six months, operations in four countries, the Nairobi County contract valued at KSh 2.7 billion, over KSh 1.3 billion collected for the county in the preceding three months, seasonal parking growing from 9,000 PSVs billed to about 14,000, single business permits moving from about four months to under fifteen minutes, and no venture capital at any point. Put to him by the interviewer rather than restated by him: roughly KSh 80,000 and later KSh 1 million raised early, and the closing description of him as a multi-millionaire — that last one is the anchor's characterisation, not a figure the founder gave. Never disclosed: JamboPay's own revenue, its commission or fee structure, its headcount, and which four countries it operates in. The KSh 2.7 billion is the contract's value and should not be read as company revenue. All figures are Kenyan shillings with no exchange rate given, so no USD conversion is made. The auto-caption garbles the founder's surname and the name of the technology firm they took a stake in; the surname is corrected here from the video title. Finally, this is a decade-old snapshot and nothing in the transcript speaks to what happened afterwards.
The first money in this business was other people's money moving, with a cut attached — an electronic payments provider earns when a transaction clears, so the product has to be attached to a payment somebody was already obliged to make.
One contract with a visible outcome became the reference that opened the rest. After Nairobi County, the company was working for the Treasury and had a large project with the National Hospital Insurance Fund — he calls that a vote of confidence from government in young people.
Six months in a cyber cafe with one leased computer is the stretch, and he describes it as formative rather than painful: "we know what it is to save, we know what it is to cut costs."
JamboPay is described as an electronic payment solutions provider, and what it deployed for Nairobi County is a revenue-collection platform sitting across every stream the county bills: parking, land rates, single business permits, rents, advertising and others. The citizen-facing surface is deliberately low-tech. Parking is paid by dialling a USSD short code from any phone — no app, no card — and can be paid after midnight for the following day or for a whole month. Permits move to a web portal at the county's own domain. The measurable change is the removal of paper: he says there is not a single paper or manual parking ticket left in Nairobi. Our read: this is the part that generalises best. The product's value is not the payment rail, which was already there in Kenya's mobile money ecosystem; it is the replacement of a cash-and-paper workflow that leaked at every hand-off. A manual ticket is an opportunity for the money not to arrive; a USSD payment against a vehicle record is not. Every design decision described — USSD instead of an app, a light deployment, a portal instead of a counter — points at the same thing, meeting the user on the device and the habit they already have.
The company's own economics are never disclosed. What is stated is the contract's value, 2.7 billion shillings, and the collections it produced: over 1.3 billion shillings for the county in three months. Our read: treat the 2.7 billion as the size of the mandate, not as JamboPay's revenue, because the interview never says which it is, and the two are very different numbers. The normal structure for this kind of deal is a commission on collections, sometimes with an implementation fee, which aligns the vendor with the only outcome the county cares about. That structure is also the reason such contracts attract political attention later — a percentage of a city's cash is a visible line item in a way a software licence never is. The funding side is unusually clean and is stated plainly: "JamboPay has not attracted one VC firm to date, it is all self-funded from our activities." The only capital mentioned is what the interviewer describes as roughly 80,000 shillings and later 1 million shillings raised early on. He extends that into advice — you cannot wait for money, and a business card costs about six shillings.
No stack detail is given beyond one constraint that shaped everything: "we deployed a solution that is very light because we do not have funds — we had to research extensively to come up with a solution that does not require a lot of funds." The citizen interface being USSD rather than an app is consistent with that, since USSD runs on every handset and needs no distribution. Cost discipline is presented as a competitive asset rather than a phase to grow out of. Six months of shared cyber-cafe desk space at 4,500 shillings a month is the entire infrastructure budget of the first half-year, and he returns to frugality repeatedly when comparing his path to incubated startups. Our read: margins are unknowable from this interview and we will not guess at them. What can be said is where the money goes in this category — implementation and integration labour against legacy county systems, field support during rollout, and the working capital needed to run a deployment while a public client pays on public-sector timelines. That last one is the quiet killer, and it is the reason he built capacity through a university partnership and a stake in another local firm rather than by hiring a large bench.
Copy the wedge. Find a public body whose revenue is collected on paper, pick the single leakiest stream, and sell the two numbers that matter: money collected and citizen minutes saved. The permit going from four months to fifteen minutes is a better pitch than any feature list, and the parking figure — 9,000 vehicles billed before, about 14,000 after — is the kind of evidence that wins the next tender for you. Copy the interface discipline too. USSD was the right choice because it works on every phone already in the market. The equivalent today is whatever your citizens already have open, not whatever is fashionable to build. Change the concentration. A single client that is a city government, on a contract worth 2.7 billion shillings, is both the making of this company and its largest exposure. Political cycles change, procurement is contested, and a percentage of public money invites scrutiny. A version of this built now should aim for many small municipalities on a standard product rather than one large county on a bespoke one — less revenue per logo, far less that can end overnight, and a product that can be sold across borders instead of rebuilt for each buyer.
Single-client concentration is the obvious one. The company's visible scale rests on one county contract; losing it, or losing the political sponsor behind it, removes most of the story at once. Procurement risk sits alongside it. Winning against IBM and the banks once does not mean winning the renewal, and public contracts of this size are routinely challenged, audited and re-tendered. Any business whose revenue is a share of public money should expect that scrutiny as a permanent operating condition rather than an event. Regulatory and platform exposure follows: payments are licensed activity, and the rails underneath — mobile money operators, banks, telecom short codes — are a small number of parties who can reprice or withdraw. Our read: the largest caveat for a reader is the date. This interview aired in January 2015 and describes the position at that moment. We make no claim about what happened to the company or the contract afterwards, because nothing in the transcript speaks to it, and anyone treating this as a current case should check the last decade of public record before drawing conclusions. Read it as a well-documented snapshot of how a two-person company with a rented desk won a city-scale mandate, not as a statement about where that company stands today.
The four founder playbooks — first customers, first dollar, the flywheel, the silent stretch — are in the Unicorn Club, on this idea and on all 1,435 others. The story, the numbers, the sources and every later section stay open to everyone.
Join the Unicorn Club — $5/moCancel anytime, in one click