Visa's Real Moat Isn't Scale. It's That Nobody Can Spin the Wheel Twice.

In 1958, in Fresno, California, Bank of America mailed 65,000 working credit cards to its own customers - unsolicited, already activated, each carrying a $300 line of credit nobody had asked for.³ It was reckless. It was also the only way to solve an impossible problem: no merchant would accept a card no one carried, and no one would carry a card no merchant accepted. You cannot grow a network one careful step at a time when each side is waiting on the other. So Bank of America did the only thing that breaks the deadlock. It forced one side into existence overnight, and let the other side chase it.

The official story is that Visa is a brilliantly run financial company that scaled its way to dominance. It is closer to say Visa is a flywheel that got pushed hard, once, in the right place - and has been coasting on its own momentum ever since. The company itself is emphatic that it isn't even a lender: it issues no cards, extends no credit, sets no consumer rates.² What it owns is the wheel. And the whole point of the wheel is that you only get to spin it the first time.

Each side of the market builds the other

A payment card is a textbook two-sided market - cardholders on one side, merchants on the other, and a platform whose entire value is connecting them.⁸ The flywheel turns like this: every new cardholder makes the card slightly more worth accepting, which pulls in another merchant; every new merchant location makes the card slightly more worth carrying, which pulls in another cardholder. Neither side is buying the platform. They are buying access to the other side. Round and round, and after enough turns the thing acquires a property that money can't manufacture: ubiquity. By fiscal 2024 the wheel had spun to roughly 4.6 billion payment credentials and more than 150 million merchant locations, carrying some $16 trillion of payments and cash volume across 234 billion transactions.¹ That isn't a market share. It's a habit the world has, and habits don't switch on a better fee.

The pricing inside the wheel is stranger than it looks, and it's where most strategy writing goes wrong. To keep both sides spinning, a two-sided platform deliberately under-charges the price-sensitive side and over-charges the other - cardholders often pay below cost, even negative prices, in the form of rewards.⁸ That asymmetry is funded by interchange, and here is the part everyone gets backwards: Visa sets interchange but does not keep it. The interchange flows from the merchant's acquiring bank to the cardholder's issuing bank, which uses it to bankroll the rewards that keep cardholders loyal.⁷ So when a merchant curses the 'swipe fee,' the money it resents is mostly the banks' - not Visa's. Visa charges its own, separate network and processing fees, regardless of where interchange lands.⁷ The wheel pays the banks to keep feeding it cardholders. Visa just takes a sliver for owning the axle.

The cooperative was the real invention

Building a two-sided network is hard. Building one that won't be torn apart by antitrust regulators or strangled by the bank that owns it is harder, and that is the part the founder myth obscures. Popular tellings credit Dee Hock as the genius who dreamed up the card in 1958. He didn't - the program was the brainchild of Joseph P. Williams at Bank of America, and Hock didn't enter the story until 1968, brought in to manage a regional rollout in the Pacific Northwest.³⁵ What Hock actually did was more interesting than founding. Bank of America had started licensing BankAmericard to rival banks in 1966; by 1970 the program was spun out of a single bank into a cooperative owned jointly by its member banks.⁵ The card had stopped belonging to one bank and started belonging to a network of competitors who all needed it to work. That structure diffused the control - and the legal exposure - across thousands of institutions, and turned a proprietary product into shared infrastructure. The rebrand to Visa followed in 1976, the name chosen for how it reads in every language.⁵ The brand was the easy part. The cooperative was the moat.

If it's just a flywheel, why hasn't someone spun a new one?

The fair objection is that flywheels are not invincible - networks do get displaced, and a 66% margin is exactly the kind of fat that attracts attackers. True. But notice what a challenger actually has to do. It cannot win cardholders without merchants, or merchants without cardholders, so it must bootstrap both sides simultaneously, from zero, against an incumbent that already has 4.6 billion credentials in wallets and 150 million places that take them.¹ Bank of America solved that by mailing live cards into a single city; you cannot mail your way past a network the size of the planet. The honest counter is that the wheel is being tested at the edges - regulators worldwide keep poking at interchange, and the very two-sided economics that built the moat are now the subject of academic and antitrust scrutiny.⁷⁸ When Visa raised $17.9 billion in its 2008 IPO - the largest in U.S. history at the time - it was monetizing a wheel already two decades into its coast.⁶ The position isn't eternal. But its durability has never depended on being unbeatable on price. It depends on the one thing a rival can't buy: the irreversibility of a network already spun.

Visa's margin looks like pricing power. It isn't. It's the signature of a machine built once and run forever - near-zero cost to carry the next transaction, because the costly thing, the network itself, was paid for decades ago in Fresno and never has to be paid for again.¹ The real genius was never the card, the brand, or the fee. It was understanding that a two-sided wheel is brutal to start and impossible to restart - and then starting it, under a structure that made sure no one could ever take it back. Everyone else has been trying to push their own wheel uphill ever since. Visa is just coasting, and collecting a sliver every time the world goes round.