If you want to sell a phone in Kigali today, here is roughly what happens. You post it in two or three WhatsApp groups. A friend reposts it. Someone you have never met replies "still available?" and then goes quiet. You agree to meet at a petrol station because it feels safer than a house. You have no idea what the phone is actually worth, because the only comparison you have is whatever your cousin sold his for.
It works. Millions of shillings change hands this way every day. But it works the way a footpath works before anyone builds a road — through effort that nobody counts.
What is actually broken
Three things, and none of them is "people do not have an app".
Discovery is trapped inside groups. Your listing reaches the two hundred people in that WhatsApp group and stops. The person across town who wants exactly what you are selling never sees it. The market is not small; it is fragmented into hundreds of small ones that cannot see each other.
Price is guesswork. Without visible comparable sales, both sides negotiate blind. Sellers underprice out of impatience and overprice out of hope. Buyers have no way to know which is happening.
Trust does not travel. Inside a group, reputation is real — people know who you are. Step outside it and you are a stranger with a phone number. Every transaction restarts the trust question from zero, and that friction is why so many trades simply do not happen.
What we built instead
RAY Markets is our attempt at the road.
It is a marketplace for Rwanda: fifteen categories covering phones, electronics, cars, bikes, rentals, commercial space, furniture, fashion, jobs and services. Listing is free. It runs on the web and as an Android app, because plenty of people will never install anything and plenty of others live in one.
The design decisions that mattered most were not the visible ones.
Mobile-first, genuinely. Not a responsive desktop layout. The primary experience is a phone on a connection that comes and goes, and the listing flow is built to survive that.
Location is a first-class field. Not because it looks good in an interface, but because "where" is the second question anyone asks and the first thing that decides whether a trade is practical.
Free to post, permanently. A listing fee is the fastest way to guarantee an empty marketplace. Liquidity comes first; a marketplace nobody lists on is not a business model, it is an empty room.
What we got wrong
We assumed people would browse categories. Mostly they search, or they scroll recent listings and stop when something catches them. Category browsing turned out to be how you explore when you do not know what you want, which is not the common case.
We also assumed the app would matter more than the web. It is closer to even than we expected — a link someone can open immediately beats an install, even when the install is better once you have it.
The part that is not solved
Trust. We show who is listing and where, and the marketplace makes reputation visible in a way a WhatsApp group cannot. That is better than nothing. It is not the same as knowing someone.
We think this gets solved by accumulation rather than by a feature: enough completed trades, enough visible history, enough people who have used it without regretting it. That takes time and cannot be shortcut with a badge.
Where it goes
RAY Markets is live and growing, which is the only interesting thing about it right now. Every listing added is a signal about what people actually want to trade, and every one that stalls is a signal about what still does not work.
The larger idea is that a marketplace is infrastructure. Once buying and selling has somewhere reliable to happen, other things become possible on top of it — delivery, payment, credit built on a real transaction history. Those are not features we are planning next week. They are what a road is for.
For now, we are building the road.