Nigeria spent a decade solving the wrong constraint

Nigeria's cashless policy dates back to 2011. For more than a decade after that, the country was told — patiently, repeatedly, expensively — that digital payments were faster, safer and more convenient than cash.

There were banks. There were mobile money operators. There was USSD. There was Paga. There were campaigns.

And cash dependence barely moved.

That distinction matters. The infrastructure did expand — POS terminals, agent banking, electronic transfers, mobile money licences. Digital payments grew. What stayed stubborn was the underlying behaviour: what people actually reached for when it was time to pay.

Then, in the space of about three years, Nigeria recorded one of the largest declines in cash usage anywhere in the world.

I find that more interesting than any of the numbers that followed it. Because the numbers are the result. The interesting question is what changed — and why it wasn't the thing everyone had been spending money on.

The numbers, briefly

OPay's gross transaction value reached $358 billion in 2025, up from $166.2 billion the year before — a 115% increase. It reports more than 45 million users and over a million merchants. It has hired Citigroup, Deutsche Bank and JPMorgan for a US listing targeting a $4 billion valuation. Targeting, not achieved — the IPO hasn't happened.

One caveat, because it stopped me when I first read it: gross transaction value is not revenue and it is not GDP. It counts every transaction, including transfers that recirculate the same money many times. It measures activity on the platform, not the size of the economy. The figure is remarkable; it just isn't the figure people sometimes assume it is.

Understanding was never the binding constraint

I've been developing a framework I call the Five Constraints of Market Formation: Understanding, Access, Trust, Momentum, Coordination.

The point of it isn't the list. Most markets face all five in some form. The point is that at any given moment, usually only one or two are actually binding — and effort spent on the others is waste.

For a decade, Nigeria spent against Understanding.

Nigerians did not need to be told that digital payments existed. They knew. They had been told, at length, by banks, by operators, by the central bank. Comprehension was not the problem, which is why a decade of communicating it changed so little.

Cash has an extraordinary user experience. If I hand you ₦10,000, you have it. No pending transaction. No “try again later.” No SMS that may or may not arrive. No customer service ticket. No twenty minutes wondering where the money went.

That is a very high bar, and it isn't a comprehension bar. It's a reliability bar.

So the binding constraints were Trust and Access — whether the rails existed for you at all: an account you could open, an agent who could help, a merchant who could receive, a transaction that would actually settle.

And Trust here is worth splitting in two, because the distinction turns out to matter enormously later.

Transactional trust: will this payment actually work?
Safety trust: will I be protected from fraud, misuse and abuse?

They are not the same thing, they are not solved the same way, and solving one hard can cost you the other.

Everyone was advertising at Understanding. The market was stuck on Trust and Access.

Transactional trust is not the same as security

A system can be technically secure and still not be trusted. And a system can be trusted by people who could not begin to describe its security architecture — because it reliably does what they expect.

Transactional trust is the kind no campaign can manufacture. It is earned one successful payment at a time, and it compounds. If it works every time, you stop thinking about it. That is what trust looks like in payments: not confidence in the institution, but the absence of anxiety at the moment of paying.

OPay's positioning has never really been digital payments are the future. It has been closer to this works. That is a much less impressive sentence and a much more valuable one.

Access, and the thing you can't see from outside

Access isn't whether an app exists. It's whether the rails reach you. Can you open an account? Get money in? Send it out? Can a merchant receive it? Can an agent help someone who isn't comfortable doing this alone? Does it settle?

Nigeria's financial inclusion strategy has long recognised agent networks and simplified onboarding as central to this. OPay built heavily around that distribution layer, and it's a large part of how it reached people the branch network never did.

But there's an earlier chapter that I think is the most instructive part of the whole story.

OPay didn't arrive as a payments company. It launched in 2018, after Opera took a controlling stake in PayCom, pursuing a broad super-app strategy. There was ORide, motorcycle ride-hailing. There was OFood. There were other verticals.

ORide didn't become the business they hoped for. But consider what running it required. You cannot operate motorcycle logistics at scale in a market where payments are slow, cash-based and untrackable. The driver has to know they've been paid. The platform has to know the transaction happened. The business needs a digital record of both.

ORide didn't just need reliable digital payments. It broke without them.

Sometimes the binding constraint isn't visible from outside, and building something is how you find it.

You can diagnose a great deal from research and evidence — it's most of what I do. But some constraints only announce themselves when you try to operate on top of them and they give way. That is an uncomfortable thing for someone who sells diagnosis to admit. It's also true.

The 2023 trigger, and the part people miss

In 2023 the naira redesign produced a severe cash shortage. Currency in circulation fell 29.2%, to ₦982.1 billion in February 2023 — the lowest level since 2008.

The question in millions of heads stopped being why should I try digital payments and became how do I pay for this.

Bank apps struggled under the load. OPay and other digital-first players became increasingly attractive precisely because their payment channels were perceived as more reliable.

But the trigger is the least interesting part of this, and it's where most accounts stop. The interesting fact is what happened after cash came back. Participation didn't collapse. Nigeria went on to record one of the world's largest drops in cash usage, years after the shortage ended.

That is why I've stopped calling this constraint Trigger and started calling it Momentum — because it has two halves. Ignition converts intention into action today. The loop is what brings people back tomorrow without a new prompt. A trigger without a loop produces a spike and nothing else.

The cash crunch was ignition. It was free, external, and OPay didn't create it. What OPay had done was be the thing that worked when everyone was forced to try — so that the behaviour survived the emergency that started it.

The policy created the shortage. The shortage created the trigger. Neither created the readiness.

What it cost

Any honest account of this has to include the other side, and it's the most methodologically interesting part.

Relieving Access aggressively — near-frictionless onboarding, minimal barriers, get everyone in — is exactly what made the growth possible. It is also what made the fraud problem possible.

By late 2023, OPay was facing public scrutiny over weak KYC, including demonstrations of accounts being opened using other people's details. It introduced stricter KYC requirements from March 2024. In April 2024 the Central Bank ordered OPay, PalmPay, Kuda and Moniepoint to stop onboarding new customers entirely for around two months, over compliance concerns and accounts being used for illicit transactions. Fines followed. The EFCC was still pressing OPay on KYC and anti-fraud compliance in December 2025.

Read through the framework, that sequence is not a scandal. It is a substitution effect running in reverse — and it is where the two kinds of trust separate.

OPay solved transactional trust exceptionally well. That is what won it the market. But relieving Access at that speed loaded safety trust, which it had not solved to the same standard. The regulator then forced friction back into onboarding — which is to say it deliberately re-constrained Access in order to protect the other trust.

Market formation isn't a checklist. It's a moving bottleneck.

Relieve one constraint hard enough and you don't finish; you shift the binding constraint somewhere else, often somewhere less convenient.

Coordination is where it goes next

The final constraint matters more as a market matures. A payment method becomes more useful as more people accept it — more merchants, more agents, more platforms, more places where the same behaviour simply works.

The shift you're looking for is from use this to this is part of how the market works. That transition is the difference between a successful product and a formed market, and it's the one that's hardest to buy.

What I take from it

The constraints didn't appear in 2023. They were there for over a decade. The technology existed, the policy existed, the demand existed. What was missing was reliability, distribution and trust — and a moment that made the behaviour necessary rather than preferable.

You solve constraints before you need the market to exist. Because the trigger, when it comes, tends to arrive without warning. A currency redesign. A pandemic. A regulatory change. A fuel price. And when it does, the companies that spent the preceding years quietly relieving constraints look like overnight successes. They aren't. They were ready.

Readiness is expensive, and that's a structural problem here. OPay could afford several years of experiments that didn't work, because it had capital from Opera and later SoftBank behind it. Most companies in this market cannot afford to be wrong for three years. If discovering the binding constraint sometimes requires building things that fail, then the capacity to be patiently wrong is itself a precondition for market formation — and it is exactly what's scarce.

The caveat: this is survivorship bias, and I want to name it rather than have it pointed out. For every OPay whose failed experiments turned out to be capability-building, there are many companies whose failed experiments were simply failures. You only see the pattern in the winners.

Which is also true of this entire piece. It is a retrospective reading, and retrospective readings always flatter the framework applied to them — the constraints line up neatly precisely because we already know how it ended. That's why the version of this work that matters isn't the case study. It's the call made in advance, in writing, with a date on it and a stated condition that would prove it wrong. I'd rather be judged on those.

The next time you're looking at a market that won't move, the question probably isn't how to get more people to buy. It's what's actually preventing people from taking part — and which one of those things is binding right now.

Because solving the wrong constraint is almost as expensive as solving none at all.

Markets don't just happen. They are formed.