In 2019, before crypto, wallets or video, I co-founded a marketplace that held the buyer's money until the item arrived. Most of what I still build started there.
Dennismary Chidiebere25 October 20264 min read
Operator Notes
Get the next essay in your inbox.
Field notes from building, deciding and carrying the cost of clarity.
I came across an old link this week. A Techpoint feature from May 2019, about a new entrant in Nigeria's online classifieds market. The co-founder and CEO quoted in it is me.
I had half forgotten it existed. Reading it again was strange, the way finding an old notebook is strange. The handwriting is younger, the confidence is louder, and the ideas are, uncomfortably, the same ones I am still working on seven years later.
In 2019, online classifieds in Nigeria all ran the same model. A platform connected buyers and sellers, and stepped back. Payments, delivery and disputes were somebody else's problem. Mostly they were the buyer's problem.
That gap had a name everyone knew: trust. You could find almost anything online. What you could not easily find was a reason to believe the seller would deliver, that the item would be what the picture said, or that anyone would help if it went wrong. So most commerce defaulted to pay on delivery, and every transaction carried a little fear on both sides.
We called MaketAam a social marketplace, and we set out to make trust the product rather than an afterthought.
What we built into it
Reading the feature now, I can see four decisions that still look right to me.
Verify the seller, properly. Before launch we had signed up more than a thousand merchants and service providers, and every one of them was verified, including their bank details. Selling on MaketAam meant being known. We also chose to list only new items, and services, rather than anything anyone wanted to offload.
Hold the money until the buyer confirms. This is the one I am proudest of. A buyer could prepay, and the seller was not credited until the buyer confirmed the item had arrived. Service providers were paid only after the client was satisfied. If a buyer received an item and never confirmed, we could step in and release the payment, so sellers were protected too. A buyer could still pay a seller directly, but then our protection did not apply. The rule was explicit, and both sides knew it.
Make the conversation evidence. Every marketplace has chat. Ours did not allow phone numbers to be shared, and each session was archived, visible only to authorised staff, for one purpose: resolving disputes. The chat was not just a feature. It was the record that let us be fair.
Treat the people doing the work like people. We planned for service providers to be treated more like staff than listings, with health insurance, transport partnerships to get them to jobs, and insurance protecting clients against damage.
The business model was simple to the point of stubbornness: no paid promotions, a flat ₦200 on every successful sale regardless of value, and a commission on services. We wanted to earn when a transaction actually worked, not when someone paid to be seen.
The thread I did not see at the time
At the time, I thought I was building a marketplace. Looking back, I was learning to separate two things that most systems blur together: moving money, and earning the right to it.
Holding a buyer's payment until delivery is confirmed is exactly that separation. The payment is one event. The seller's right to the money is another, and a clear rule decides when one becomes the other. I have spent the years since building that same bridge in different forms.
At ComiBlock, it was investors' money and the plans it was meant to serve. At Yuki, it was wallet infrastructure for fintechs, where custody and settlement are the whole job. At SFX, it is tuition crossing borders, where a payment only matters once the school has it. At Circo, it is the licence that turns a creator's paid video into something a viewer can actually watch, a rule I later wrote about as a transaction is not the same thing as access.
Even the chat design came back. The way MaketAam treated a conversation as a record for resolving disputes, with no phone numbers and controlled access, is the same instinct behind how we think about messaging and moderation at Circo today.
I did not have the vocabulary for any of this in 2019. I had a market, a trust problem and a conviction that the platform should carry some of the risk instead of leaving it all with the buyer.
What a first startup is for
A first company teaches you things no job can, mostly by putting all of them on your desk at once. Merchant onboarding and verification. Payments and the rules around them. Logistics partners, insurance, pricing, a launch date. Every one of those became a discipline I went deeper into later.
What it really gave me, though, was a question I have never stopped asking: what would it take for both sides of a transaction to trust it? Every company I have built since has been another answer.
The headline in that old feature says MaketAam was working to provide value for both buyers and sellers. Seven years on, I would not change a word. I would just recognise it for what it was. The beginning of everything else.