Monday evening. I’m back in the same hotel in Mikocheni, Dar.
Spent most of today with Trade Wars Are Class Wars going in my ears on Audible.
Thanks for reading! Subscribe for free to receive new posts and support my work.
Second time in this hotel.
First time was maybe two months ago. Back then I was running on the belief that a moving man will surely meet his luck, so my cofounder and I booked flights to Tanzania to poke around for two weeks. We stayed seventy days.
We were building financial infrastructure for Africa, for the stablecoin economy. My conviction was that financial institutions would eventually drop the traditional, fragmented rails and move onto stablecoin rails. Inevitably.
Import and export businesses inside Africa, I figured, stood to gain the most from that. Rift came out of that conviction.
On the first visit I never bothered changing my Kenyan shillings into Tanzanian ones. Didn’t pick up dollar notes either, even though dollars are wanted almost everywhere. I was a neobanker. Most of my money sat in USDC. In Kenya I could spend that USDC through Rift and never open Binance P2P, which is how most people there turn digital dollars into shillings.
Kenya runs on mobile. Almost every business has an M-Pesa till.
So with Rift I could pay any person or business in Kenya straight from USDC into their M-Pesa. Instant. No P2P.
Landing in Tanzania, I assumed Rift would carry us the same way. Pay people, pay businesses, exactly like home.
I’d integrated the APIs with our payment service provider and our liquidity provider before we flew. Ran a one dollar test transaction and everything. It went through. We were set.
Long story short, Rift didn’t work.
Tried my virtual stablecoin card instead and the POS machine kept rejecting it.
The Rift problem wasn’t a bug in the backend. Not the smart contracts, not the frontend, not a tech glitch at all. Dollar liquidity. The provider we were leaning on didn’t have enough dollars sitting there to process the volume and finish the settlement.
That one night ended up setting the agenda for the next seventy days. We spent them pulling money in Africa apart, slowly, to see how it actually moves. Cross-border payments. PAPSS. Nostro and vostro accounts. The dollar detour. All the macro stuff quietly sitting under systems we touch every single day.
One simple question, why didn’t Rift work in Tanzania, opened into something much bigger.
And to answer it we first had to work out how money crosses this continent at all.
Here’s where we landed. Payments inside Africa are slow and expensive in large part because they take a detour through the dollar.
Money going from Kenya to Nigeria will often pass through correspondent banks somewhere abroad. Extra spreads. Extra compliance checks. More intermediaries, each taking a cut on the way through. And for any dollar transaction to settle, dollar liquidity has to be sitting somewhere along that chain.
African economies can’t print dollars, though. They have to earn them or attract them. Exports. Remittances. Foreign investment, development finance, debt.
In a healthy, productive economy a good share of those dollars comes from selling goods and services to the rest of the world. For a lot of African countries the problem is that we don’t export nearly enough relative to the foreign currency we need. Trace that back and you usually find limited industrial capacity, plus heavy reliance on imported goods, machinery, energy, basic inputs.
Which gives you dollar scarcity.
Scarce dollars means banks struggle to get enough foreign currency to fund their offshore correspondent accounts and settle international trades. Payments get delayed. Spreads widen. Moving money over a border gets expensive.
Scarcity drags on the local currency too. Let the rate float and it depreciates, because businesses are competing over a limited supply. Defend an official rate instead and the shortage just resurfaces somewhere else. Getting dollars at the official rate becomes nearly impossible and a parallel market shows up where they trade at a markup.
So the thing that looks like a payments problem on the surface is usually a production problem underneath. Not the only factor. But it’s the one that doesn’t go away just because you shipped smoother software.
Anyone running cross-border payments in Africa is really in the business of sourcing dollars cheaply and efficiently.
That’s the takeaway that stuck. We learned far more in those seventy days than I can fit in here, but I keep circling back to that one.
And it pushed me into a bigger question. What actually matters for Africa? What kinds of companies should we be building?
Africa isn’t one country, obviously, and my experience is mostly Kenya and Tanzania. Still, I think we share enough of the same structural constraints and opportunities that the question holds.
So, another rabbit hole.
I wanted to see how other regions did it. Asia especially. Read Joe Studwell, then kept going into other books and papers on economic development.
Kept landing in the same place.
We need more companies that build up our productive capacity.
My framework for what to build here has gotten embarrassingly simple. Does this help the country or market I’m building for produce more?
I should say what I mean by produce more, because it’s easy to hear that as a vague slogan.
I mean things we can sell outside our borders. Exports. That’s the direct line back to the dollar problem I started with. Every export is a dollar earned, and dollars earned are what make the payment rails work at all.
Start with agriculture, since that’s what most of our economies already do. Then move up. Right now a lot of what leaves here leaves raw. Cashews go out unprocessed and come back to us as packaged snacks. Tea leaves the country in bulk, gets blended and branded somewhere else, and the margin lands there instead of here. The processing step is where the money is, and we mostly hand it away.
So value-added agriculture. Processing, packaging, branding, cold chain. Same crop, several times the price.
Then manufactured goods. Harder, slower, more capital, but that’s how every economy that got rich did it.
And talent. Skills are exportable too, whether people leave to earn abroad or stay and sell services out. But talent is also the input for everything above it. You can’t run a processing plant or a factory without people who know how.
That’s the ladder. Raw goods, then processed goods, then manufactured goods, with people capable of running all of it.
That’s the test. That’s all of it.
Agriculture, manufacturing, services, whatever. Somewhere down the line, does the thing you’re building help people produce more, or make them more productive?
A few days ago I got to spend time with a brilliant mechanical engineer who studied here in Dar and now works in energy. I knew close to nothing about power systems before that conversation. She taught me a lot.
At some point I started thinking of energy as the operating system of the world. Nearly everything runs on top of it.
One thing she said has stayed with me. A lot of energy projects aim at household access, lighting homes, that kind of thing. Which matters. But there’s a massive opening in energy for agriculture, business, heavy industry.
Power a factory and it produces more output. Power irrigation and a farmer grows more crops. Supply energy for fertilizer production. Build cold storage so food doesn’t rot before it reaches a market.
I started seeing all of it as one mechanism wearing different clothes. It expands what we’re able to produce.
Fintech has a version of this too. Credit so a farmer can buy equipment or inputs. Financing for a small factory. Software that makes these real-economy businesses bankable, so capital can actually reach them.
Education slots in as well. Teach young Africans practical skills, plug them into a productive value chain, and you’ve raised what the whole economy can generate.
That’s the thesis. Build companies that increase productive capacity. And whatever you pick, ask whether it helps the market produce more.
Now, the elephant. Financing.
The reason so many founders end up in consumer fintech or light software isn’t that they lack a vision for hard infrastructure. It’s that cold storage, irrigation networks and small manufacturing need heavy upfront capital, involve physical assets, and run on seven year payback periods. Almost none of the venture capital available to African founders today is shaped for that.
Which means a thesis on what to build demands a thesis on who funds it and how that capital is structured. Bigger topic. Another post.
Jahazi came out of all these questions.
Jahazi is an attempt to gather people who think seriously about Africa and where it’s going. Founders, engineers, operators, critical thinkers, all chewing on versions of the same questions.
What should we be building? What does it really take for our economies to produce more? And what would Africa look like if more of us spent our time building toward that?
The idea is simple. Put those people in one room, give them time together, see what conversations and ideas and companies fall out of it.
We’re early and still working through details. But if any of this resonates, I’d love to have you join.
Thanks for reading! Subscribe for free to receive new posts and support my work.