Why Emerging Markets Are Not a Smaller Copy of the Same Opportunity
cr4fts.com/n/ucb3c
The standard description of an emerging market is a list of absences. Fewer banked adults. Lower internet penetration. Less formal retail. Thinner capital markets. Every item is factually correct and the composite is almost entirely wrong, because a list of absences describes a market as a deficient version of somewhere else rather than as a different place with different physics.
This framing is not merely inaccurate. It is expensive, because it determines what gets built. If a market is a smaller copy, the correct strategy is to take a proven product and degrade it: fewer features, lower price, simpler onboarding. That strategy has been attempted continuously for twenty years and it has an almost unbroken record of failure.
The markets that matter here are not underserved versions of developed markets. They are places where the defaults have not been set, and the companies that understand the difference are building things that have no equivalent anywhere.
Everyone cites mobile money and almost everyone draws the wrong lesson from it.
The figures are not marginal. The GSMA's industry reporting puts global mobile money transaction value above two trillion dollars in 2025, with roughly 1.4 trillion of that in sub-Saharan Africa. The region accounts for about two thirds of global transaction value and more than half of all registered accounts worldwide, over 1.1 billion of roughly two billion. Global annual transaction value took twenty years to reach one trillion dollars and approximately four more to double.
The standard reading is that Africa leapfrogged banking. This is wrong in a way that matters enormously for anyone deciding what to build.
Mobile money is not a cheaper bank account. It runs on different primitives entirely. Value is held against a phone number rather than a verified identity document. Cash enters and exits through a network of human agents rather than branches or machines. The fee structure is built around many tiny transfers rather than few large ones. Each of those is a different architectural decision, not a reduced version of a banking feature.
And each was a direct response to a constraint. Identity documentation is patchy, so the system does not require it. Branch networks are uneconomic below a certain population density, so the system uses shopkeepers who are there anyway. Most transactions are small, so the economics are built for small transactions. A firm arriving with a simplified retail bank would have failed, and several extremely well-capitalised firms did exactly that.
Because this case carries so much of the argument, it is worth examining more closely than the headline number allows, and the details are where the transferable lessons sit.
The agent network was the product. Everyone fixates on the wallet, which is the trivial part: a balance against a phone number is a database row. The difficult part was assembling tens of thousands of small retailers willing to hold float, handle cash and be trusted with both, then constructing the liquidity management, commission structure and fraud controls that keep such a network solvent. That took years, it looks like operations rather than technology, and it is the reason the position was never successfully attacked.
The regulatory position was negotiated rather than granted. Permitting a telecommunications company to hold customer value looked, to most central banks at the time, like an obvious prudential hazard. The arrangement that emerged, a trust account holding customer funds separated from the operator's balance sheet, was argued for over years, and the operators who moved early in permissive jurisdictions built leads that later entrants never closed.
The third detail is the uncomfortable one. Adoption was not driven by the unbanked discovering financial services. It was driven by remittances: people sending money to family in another part of the country, replacing a bus driver carrying an envelope. That is a specific, frequent, expensive task with obvious willingness to pay. Financial inclusion followed adoption. It did not cause it.
That is the lesson worth taking. The product that changed a continent's financial system did not begin by addressing the structural problem. It began by addressing a narrow, urgent, expensive task that people were already paying to solve badly. Companies that start from the structural problem build things that are impressive and unnecessary.
In a mature market almost every process has a default. There is a way payroll is run, a way a clinic keeps records, a way a wholesaler extends credit. These defaults are usually mediocre and almost immovable, not because they are good but because every adjacent process has been arranged around them. Displacing one means paying the switching cost for everything connected to it, and that bill is what actually protects most incumbents in developed markets.
Where a process has never been formalised, that bill does not exist. Nobody migrates off the previous system because there is no previous system, only a person, a notebook and a set of habits. This is a fundamentally different competitive situation and it is the single most underrated feature of these markets.
It is also harder than it sounds, and the difficulty is not the one people expect. A market with no default frequently has no budget line either. In a mature market the customer is already spending money on the problem and you compete for a known sum. Where nothing is formalised, you are not taking share. You are creating a category, and the first question is not whether you are better than the alternative but whether anyone will pay for something they have always done for free with a notebook.
That is the real risk in these markets, and it is why the fastest available test is whether a business can charge money in month four. Everything else is opinion.
Connectivity is intermittent. Devices are shared and replaced often. Cash remains dominant in daily commerce. Trust is local and personal rather than institutional.
Read as deficiencies, these are reasons to wait for a market to develop. Read as a specification, they produce products that work where the imported version does not, and that is a defensible position rather than a compromise. The difference is visible in the details.
Designing for intermittent connectivity is not adding a cache. It means deciding what the product may promise when it cannot reach a server, which is a product decision rather than an engineering one. A tool that lets a shopkeeper record a sale offline and reconcile later is a different product from one that refuses the sale, and the difference is not technical sophistication. It is a decision about who absorbs the uncertainty, and the imported version almost always puts it on the customer.
Requiring a national identity document at signup excludes a large share of the market at the first screen. The alternatives, a phone number with history, a guarantor, an agent who knows the person, are weaker cryptographically and far stronger practically. They are precisely why mobile money reached people that bank accounts did not.
Where income is irregular rather than merely low, a monthly subscription fits worse than a per-transaction fee even when the annual total is higher. The customer is not optimising for total cost. They are optimising for never being asked for money in a bad week. Products that get this wrong report strong signup and terrible retention, then conclude the market did not want the product.
None of these are compromises against a better design. They are the better design for these conditions, and a competitor arriving with the imported version loses to them on their own terms.
Africa's median age is approximately nineteen against roughly forty-three in Europe. The UN Economic Commission for Africa projects the working-age population rising from about 883 million in 2024 to roughly 1.6 billion by 2050, approaching a quarter of the world's working-age people. Internet penetration sits near half the continent and continues climbing.
This is what separates the opportunity from a merely interesting market. Unset defaults in a shrinking population is a business that peaks early. Unset defaults in a population that is doubling, urbanising and coming online is a business whose addressable market grows faster than its competition arrives.
The defaults being established over the next decade will hold for a generation, exactly as they did after electrification, after the container and after the mobile phone. Whoever sets them will not be the firm that arrives once the market is proven.
Three objections are genuinely right, and any argument for these markets that does not address them directly is selling something.
Purchasing power is the first and it kills most consumer plans. Low incomes mean low willingness to pay, and a business requiring thousands of customers paying almost nothing has a distribution problem no product improvement solves. The businesses that work here overwhelmingly sell to small enterprises rather than individuals, because a business using your product to make or save money can justify a real price and an individual usually cannot.
Fragmentation is the second. Fifty-four countries with different regulators, languages and payment rails. The pan-African product working identically everywhere mostly does not exist, and the graveyard of firms that assumed otherwise is large and well-funded. Depth in one market first, with each subsequent country treated as a new market rather than a rollout, is the only pattern with a track record.
Infrastructure risk is the third. Power, logistics and regulatory stability are genuinely less reliable, and a thin-margin model with no tolerance for disruption will meet a disruption. This is not solved by cleverness. It is solved by designing with more slack than would be sensible elsewhere and accepting the margin cost of doing so.
Everything above reduces to three questions, and all three are answerable in a fortnight of fieldwork rather than a quarter of desk research. Founders who skip them do not discover the answers later. They discover them after eighteen months, in the form of a product with strong signup and no revenue.
What is the customer paying today to solve this badly? If the answer is nothing, you are creating a category and should plan for a sales cycle several times longer than your model assumes. If the answer is a real sum going to a bus driver, a broker, an agent or an accountant, you have an existing budget line to redirect and a far shorter path to money. The second case is worth an order of magnitude more than the first and almost nobody prices it that way.
What happens when the connection drops mid-transaction? Not in principle. Sit with a user and watch. The answer determines your architecture and, more importantly, reveals whether you have understood the operating environment or are designing for a version of it that exists only in a demonstration.
Who does the customer already trust, and can you reach them through that person rather than around them? Mobile money reached scale through agents who were already known in their neighbourhoods, which is why acquisition cost stayed low while volume grew. Products that route around existing trust relationships pay the difference in marketing, usually without ever recognising that this is what the line item represents.
These are unglamorous questions and they are the entire difference between the companies that work here and the ones that raise well and disappear.
None of this is optimism about a continent, which is not an investment thesis and never was.
It is an observation about a coincidence of conditions that is unusual and temporary. The cost of building has collapsed at the same moment that an enormous number of processes remain unformalised, in markets whose working-age population is about to become a quarter of the world's. All three must hold simultaneously for the opportunity to exist.
Cheap building in a market with entrenched defaults produces a hard fight against switching costs, which is the situation in every developed economy and why so much software there competes on marginal improvements. Unset defaults in an era of expensive building produces a market nobody can afford to serve, which is the situation that persisted here for twenty years. Only the overlap produces a window, and only for as long as both conditions hold.
Emerging markets are not a smaller copy of the same opportunity. They are the only place where the defaults are still unset, at the exact moment when setting them became affordable. That combination will not recur.