#51 1958 · Bank of America (the 'Fresno Drop') · Banking / consumer finance
No bank knew if ordinary people would actually use unsecured credit — so instead of running a pilot with volunteers, Bank of America just mailed 60,000 live credit cards to an entire city and watched what happened.
the problem
you need proof that a genuinely new consumer behavior will happen at scale, but a small opt-in pilot can't tell you what an entire population will do, and no one wants to be first to find out
background
By the late 1950s, no financial institution had proven that ordinary consumers would use unsecured revolving credit at meaningful scale — charge cards like Diners Club existed, but those required full monthly repayment and didn't extend actual credit. A true revolving-credit card was an unproven behavioral and financial bet: would people accumulate a balance and pay interest on ordinary purchases, and would enough merchants accept an unfamiliar new payment method to make the card useful at all? No bank wanted to be the first to risk finding out, and a small volunteer pilot risked only attracting the most credit-hungry, unrepresentative applicants rather than revealing how typical consumers would actually behave.
Bank of America's Customer Services Research Department, led by Joe Williams, decided the only way to get an honest answer was to test the product on an entire real population at once rather than a self-selected group of eager volunteers.
what everyone would do
The standard way to test a genuinely new financial product was a small opt-in pilot, recruiting a limited group of volunteer applicants to try the new revolving-credit card and observing how they used it before committing to a wider rollout.
what they saw
Bank of America's research team saw that a volunteer pilot would only attract the most credit-hungry, unrepresentative applicants, people already motivated enough to seek out an unproven product, which meant it couldn't actually answer the real question: would ordinary, typical consumers who hadn't self-selected for credit-seeking behavior actually adopt and use revolving credit at scale. Rather than accepting a pilot's necessarily biased sample, they mailed fully active, unrequested cards to an entire real population at once, an entire city, so the resulting usage data would reflect how unselected, ordinary people actually behaved, not just how the most eager volunteers would.
the move
On September 18, 1958, Bank of America mailed 60,000 fully active BankAmericard credit cards, unrequested, directly to residents of Fresno, California, each pre-loaded with $300 to $500 of instant credit and backed by more than 300 local merchants who had already agreed in advance to accept the card, with recipients receiving no advance notice before the cards simply arrived ready to use.
why it works
Sending 60,000 live, pre-loaded credit cards directly to Fresno residents who hadn't applied or volunteered meant the resulting adoption and usage patterns reflected the behavior of an entire, ordinary population rather than a self-selected group of unusually credit-hungry applicants, giving Bank of America the only kind of evidence that could actually answer whether the product would work at true societal scale. Because the test included the real infrastructure needed for the product to function, more than 300 local merchants already agreed in advance to accept the card, the experiment tested the entire ecosystem simultaneously, consumer adoption and merchant acceptance together, rather than validating consumer behavior in isolation and hoping merchant participation would follow later. This is why the Fresno Drop produced decisive, unambiguous proof that ordinary consumers would use revolving credit, evidence strong enough to scale BankAmericard into a national program that eventually became Visa, a result a smaller, biased pilot could never have delivered with the same confidence.
the payoff
The Fresno Drop proved decisively that ordinary consumers would use revolving credit, generating real usage data no small pilot could have produced, and BankAmericard scaled from the experiment into a national credit card program that eventually spun off into the independent card network Visa — one of the largest global payment systems in existence. The mass-unsolicited-mailing method itself was later made illegal in the United States precisely because of the risks it demonstrated (unauthorized use, theft in transit, unmanaged debt exposure).
where it breaks
The mechanism depends on being willing to accept the real financial and reputational exposure of deploying an unproven product to an entire uninformed population at once, since the same design that produced decisive data also created the exact risks, unauthorized use, cards lost or stolen in transit, unmanaged debt exposure, that eventually made mass unsolicited card mailings illegal in the United States precisely because of what the Fresno Drop itself demonstrated. It also depends on genuinely needing population-level proof rather than early signal, since a company still iterating on product design would waste enormous resources and risk deploying at full population scale before the product itself was ready, when a smaller test could have caught design flaws more cheaply first. And the approach requires enough capital and risk tolerance to absorb losses if the population-scale test reveals the product doesn't work as hoped, a bet only an organization with Bank of America's resources could responsibly place, since a smaller company betting its entire viability on one unselected mass rollout has no fallback if the results come back negative.
what came after
The Fresno Drop is the foundational origin story of the modern consumer credit card industry and Visa specifically, and it remains a widely cited, if now legally unrepeatable, example in behavioral finance and product-launch strategy of testing a genuinely novel behavior on a real, unselected population rather than a self-selected pilot group — the underlying instinct, that volunteer test groups can mislead about true population-level adoption, still informs how fintech and consumer products design large-scale launch experiments today, minus the mass unsolicited mailing.
references
- [1]The Fresno Drop99% Invisible, 201899percentinvisible.org
- [2]The Fresno Free-for-All Behind the Original Credit CardAndreessen Horowitz, 2021a16z.com