Visa slashed its reported stablecoin volume after a major analytics overhaul, but transaction counts barely changed. The update raises new doubts about what on-chain numbers really say about crypto payments.
Visa just overhauled its stablecoin analytics. The company's reported adjusted stablecoin volume dropped sharply. But the number of counted transactions barely moved. This change shows how much on-chain data can shift when the rules change-and how little that might say about real payment activity.
Classification shakeup
In September 2026, Visa reported a $20 billion annualized run rate for stablecoin settlements, reflecting a sharp increase in corporate adoption despite ongoing uncertainty in retail usage.
Visa's new method tries to cut out transfers involving exchanges, contracts, bots, bridges, infrastructure, and minting or burning. With more addresses now flagged as non-organic, more transfers get filtered out of the adjusted volume. The company also set up new rules for short-term routing and tweaked how it separates payment activity from other blockchain moves. Still, Visa's public docs mention "over 3 million" labeled addresses-a number that no longer matches the updated 600 million. This leaves some confusion about which baseline is current.
Volume versus count
The gap between volume and transaction count shows a key problem with blockchain analytics. Cutting out a few high-value, non-payment transfers can shrink reported volume a lot, but barely touch the number of transactions. Visa gave an example: an automated program on Solana cycled stablecoins through thousands of throwaway wallets, creating big volumes in just a few transactions. These patterns are now excluded from the adjusted volume, but Visa did not say how much this changed the numbers for Solana or other networks.
Visa's September 2026 disclosures show over 160 stablecoin-linked card programs in operation, with payment volumes up nearly 200% year-over-year and support for stablecoin settlements across nine blockchains. This technical expansion highlights Visa's push to scale stablecoin infrastructure, even as 56% of surveyed U.S. adults report never having heard of stablecoins.
Payments or just transfers?
Not every transfer left in Visa's adjusted dataset is a payment. The company's method tries to separate payments from DeFi, exchange flows, investment, trading, store of value, minting, burning, and short-term routing. Even "retail sized" transfers under $250 are not always purchases or merchant settlements. A separate study by the Bank for International Settlements found that nearly 60% of Ethereum stablecoin transfers in 2025 happened inside complex transactions. That means a single blockchain transaction can include several token moves and financial steps. Counting every transfer as a payment risks overstating real economic activity.
Market size is also a moving target. On September 26, CryptoSlate listed USDT's market cap at about $183.79 billion. But that number shows outstanding tokens at one moment, not value moved over time. CryptoSlate's 24-hour trading volume is a different measure, and neither fills the gap left by Visa's lack of before-and-after adjusted volume data.
What the numbers really show
Right now, the only clear takeaway is that Visa's new rules have filtered out more high-value, non-organic activity from its adjusted stablecoin volume metric, while the adjusted transaction count barely changed. There is no direct proof that real-world payment use has crashed or jumped, or that one blockchain has pulled ahead in payment activity. To know if payment adoption is rising or falling, we would need consistent, dated, payment-specific data using the same definitions-something Visa has not released.
Visa's approach to stablecoin analytics is different from other payment giants testing blockchain settlement. For example, reported earlier, Mastercard and SoFi Bank have started settling card payments in stablecoins, but the real volume of actual transactions is still unclear. Without standard, open reporting across providers, users, merchants, and regulators cannot easily judge the real impact of stablecoins on payments.
Visa's data reset shows how tricky it is to read blockchain activity. The company's new method may cut out more noise, but it also shows the limits of on-chain analytics as a way to see real-world economic behavior. Until providers publish consistent, comparable, and payment-specific data, claims about stablecoin adoption in payments will stay hard to check. For now, the numbers tell us more about how companies define activity than about how people actually use stablecoins.
Visa's September 18 update to its Onchain Analytics platform expanded its labeled address set from about 15 million to roughly 600 million, according to the company's changelog. The adjusted transaction count fell by less than 2% after the refresh, while the adjusted stablecoin volume metric dropped more sharply. Visa has not published side-by-side, pre- and post-update totals for adjusted stablecoin volume, nor has it provided matched-window comparisons for individual blockchains such as Ethereum, Tron, or Solana. As of September 26, CryptoSlate listed USDT's market capitalization at about $183.79 billion, but this figure represents outstanding tokens, not value transferred.
Stablecoin analytics depend heavily on how transfers are classified and filtered. On-chain data can be sliced to show raw transfer counts, total value moved, or only activity deemed "organic" or "payment-related." Each approach has trade-offs. Cutting out exchange, contract, or bot activity may give a cleaner view of end-user payments, but risks missing real use cases or mislabeling complex transactions. Including all transfers can inflate activity with non-payment flows. For users, merchants, and regulators, knowing these differences is key to understanding what blockchain data really says about stablecoin adoption and payment trends.
