There is a number making the rounds of stablecoin decks right now: 56%.
It comes from a Visa-commissioned study on US consumer sentiment toward dollar-denominated tokens. The baseline in the same study is 36%. The twenty-point gap between those figures is, as far as I can determine, the entire reason the study exists โ and it depends on a condition that does not currently exist anywhere: hypothetical "bank-level consumer protections." Remove the hypothetical and you are left with a market where roughly two in three American consumers say they have no intention of touching a stablecoin at all. In a bear market where narrative is the only asset still generating yield, that asymmetry deserves more scrutiny than the headline received.
Stablecoins are, functionally, a settlement primitive. A token that holds a dollar peg lets two counterparties clear value without a correspondent bank, a card network, or a business-day window. That is the product; everything else โ the treasury yield, the chain choice, the governance theatre โ is packaging.
That is precisely why Visa's interest should be mapped rather than reported. The company sits on both sides of the table. It has spent several years wiring USDC settlement into its own rails, and it operates the network most exposed if consumers begin moving dollars on-chain without a card in the loop. A study commissioned by an entity with that exposure is not neutral data. It is a strategic artifact, and it should be read the way a skeptical analyst reads a sell-side note: for the model, not the conclusion.
History rhymes, but the code doesn't. Payment-adoption surveys have a long and humbling record. Early mobile-wallet studies in the mid-2000s routinely produced double-digit willingness-to-adopt among demographics that never installed a wallet. P2P lending intent surveys between 2015 and 2017 overcalled origination volume by wide margins. The pattern holds across categories: stated preference inflates, revealed preference corrects. Consumer-finance intent surveys have historically converted into actual behavior somewhere in the 20% to 40% band of the stated figure, depending on friction and trust. I have not found any reason to assume stablecoins are exempt from that base rate.
The study discloses no sample size, no sampling frame, no confidence interval, and no question wording. Each omission matters; the last one matters most. "If your stablecoin balance carried the same protections as a bank deposit, would you use one?" is a hypothetical built to elicit agreement. It is the survey equivalent of asking a driver whether they would buy a car with free fuel. The affirmative answer is not a demand signal; it is an acknowledgment of a price.
The twenty-point delta is therefore the price tag on a consumer-protection regime that has not been legislated, funded, or staffed. Stripping the hypothetical leaves the 36% baseline โ a number that says roughly two-thirds of the largest consumer market on earth remains unmoved by a product the industry has been shipping for a decade.
There is a deeper tension the study steps around, because the moment you specify what "bank-level protection" means at the code level, the stablecoin stops being the thing its advocates describe. Deposit insurance, redemption guarantees, and reversible transfers require an entity with the authority to reverse, freeze, or claw back. On a public ledger, that authority is implemented as an administrative key: blacklist functions, freeze capabilities, upgradeable proxy contracts with a controlling address. USDC already ships a blacklist. Tether already freezes addresses on request. That stack is not an accident of engineering; it is the enforcement mechanism. It is also the exact stack that makes a token censorable.
When I spent most of the 2022 drawdown buried in the proof systems of optimistic and validity rollups, the thing that kept resurfacing was never throughput. It was the admin key โ the escape hatch every trustless architecture retained for emergencies. Stablecoins have the same structure, only louder, because the emergency authority is the product feature regulators will demand. You can have a permissionless dollar or a bank-protected dollar. The market has been quietly choosing the second, and this study simply prices what that choice costs in adoption terms.
Now the regulatory read. The study's real audience is not consumers. It is the people drafting US stablecoin legislation. By putting a number โ plus twenty points โ on "consumer protection is the adoption bottleneck," Visa hands lawmakers an argument for a specific shape of regulation: reserve requirements, audit mandates, and some form of guarantee. That framework structurally favors issuers that already operate audited, compliant, freeze-capable reserves over offshore ones. It also hands the same argument to banks that want a license to issue.
And then there is the part nobody in the ecosystem wants to say out loud. The demand the study cannot see lives elsewhere. In Argentina, Nigeria, Turkey, and much of Southeast Asia, stablecoin usage is not a consumer-protection question at all; it is an inflation and remittance question. No survey framed around FDIC-style guarantees will capture it, because the respondents are not asking for insurance. They are asking for a dollar. Scoping the study to the United States almost guarantees an undercount of where adoption is already real.
The consensus reading of this study is bullish for stablecoins. I think the more defensible reading is bullish for the compliance stack, and materially less bullish for anyone who bought the censorship-resistance narrative.
Ask who wins if bank-level protections become the price of entry. Incumbent issuers with audited reserves and live freeze functions win; they get a moat built out of regulatory friction that offshore competitors cannot cheaply replicate. Banks win, because the natural provider of a "bank-level" guarantee is a bank, and banks that gain issuance rights move directly into the payment chain. Card networks win in the medium term, because the settlement layer they already operate becomes the compliant rail for a tokenized dollar, and disintermediation gets deferred. Visa commissioning this research is not a company celebrating its replacement. It is a company pricing the option value of its own disruption and buying time to become the custody layer rather than the casualty.
The contrarian case is uncomfortable for a different reason. If protections are what unlock adoption, then the asset capturing that adoption is no longer the asset that attracted the original users. It is a permissioned database with a familiar ticker, a bank charter behind it, and a legal department with a kill switch. That is a fine product. It is not the one the pitch decks sold.
Watch two things, and ignore the survey in between. First, the legislation: whether a US framework actually specifies who provides the protection โ the issuer, an insurer, or the state โ determines which part of the stack captures the value. Second, the chain: tokenized-dollar supply, payment-destination volumes, and address activity are the only adoption metrics immune to framing bias. A twenty-point jump in stated intent is a data point about question design. On-chain settlement is a data point about behavior. Only one of them pays.
Visa will publish more of these. The next one will carry the same twenty-point structure, and the same missing footnote.
