Why Blockchain’s Speed Is Finally Catching Up to Credit Card Networks

Millions of people simultaneously send money to friends, punch in orders on food delivery apps, and tap their cards at checkout counters during a busy Friday afternoon. Visa takes care of that. At a rate of 65,000 transactions per second, the infrastructure operates silently and covertly, humming along like a highway that no one bothers to look at because traffic is moving.

For years, it seemed more idealistic than practical to think that blockchain could ever replace or even significantly challenge that level of throughput. Bitcoin’s maximum transaction rate was seven transactions per second. It’s not a network for payments. That hardly qualifies as a suggestion.

However, there has been a change. Initially slowly, then more quickly than the majority of conventional finance observers were prepared to acknowledge. The blockchain industry, which has long been ridiculed for its network congestion and confirmation delays, has been stealthily working its way out of a humiliating bottleneck. And in ways that seem less theoretical now than they did even three years ago, the gap—once a canyon—is getting smaller.

Decentralization and ideology were never the main issues. It was a mechanical problem. Transactions are stored in blocks on a blockchain, such as Bitcoin, which fill up, close, and then need confirmation from later blocks before a transaction is fully settled. It takes about an hour on average for a Bitcoin transaction to be deemed final—six blocks, each lasting ten minutes.

In contrast, when you tap your Visa card at a pharmacy, the approval flashes green almost instantly. No retailer will allow you to spend sixty minutes at the register. It was not a philosophical gap. It was harmful and pragmatic.

Ethereum made advancements, such as switching to proof-of-stake, which significantly decreased energy consumption and slightly boosted throughput, but 20 to 30 transactions per second is still far less than what Visa processes before the majority of people finish their morning coffee.

Blockchain proponents, who accurately note that Visa’s infrastructure was constructed over six decades across four data centers connected by cables that, when laid end to end, could circle the Earth roughly 400 times, always felt that the comparison was a little unfair. That is not something you can duplicate with a few years of whitepaper engineering.

However, it’s possible that the competition has always been framed somewhat incorrectly. It was never clear if a single blockchain could outperform Visa in terms of raw TPS. Layer-two solutions, innovative consensus techniques, and off-chain processing tools that stack on top of base blockchains to accomplish something the original architecture was never able to do—manage scale without compromising security—are the more intriguing developments.

As this develops, it gets more difficult to write off these advancements as theoretical. Under actual circumstances, networks such as Solana have shown throughput in the thousands of transactions per second. Built on top of Bitcoin, Lightning Network completes transactions in milliseconds. These are no longer experiments.

The way engineers view blockchain congestion contains a helpful analogy. Consider a well-known bus route that takes an hour to get to its destination, runs every ten minutes, and can accommodate a set number of passengers. The bus doesn’t slow down when the route becomes congested. The travelers simply have to wait longer.

That’s precisely how blockchain throughput operates: during congestion, confirmation times increase rather than transaction capacity, making it unpleasant for anyone attempting to pay for lunch. Finding shortcuts that don’t jeopardize the integrity of the route, building faster buses, or operating more buses are the engineering challenges.

Visa, on the other hand, has hardly faltered. Since its 2008 IPO, the company’s stock has increased by more than 800%, and it has consistently reported double-digit revenue growth. Mastercard has performed even better, growing by more than 1,200% over the last ten years. By 2026, the volume of online payments could surpass $50 trillion, resulting in annual fees of over $200 billion. According to one Morgan Stanley analyst, Visa and Mastercard are essentially just running data centers where each new user makes the cost base more efficient.

These are remarkable numbers for a company that is fundamentally a data center business. It’s difficult to ignore the fact that social media platforms weren’t the most subtly powerful tech firms of the previous generation. They were plastic-covered payment rails.

However, a growing number of engineers, fintech investors, and even some dubious traditional finance observers believe that blockchain’s speed issue is temporary. There was a time when the scalability trilemma—which maintains that a blockchain can only maximize two of the three attributes—speed, security, and decentralization—was taken for granted. It is still important.

However, it appears to be more of an engineering constraint awaiting the ideal set of solutions than a fixed law. We are testing new consensus mechanisms. Both off-chain and on-chain scaling tools are in use. Finality windows are getting smaller.

It’s genuinely unclear if any of this will pose a serious threat to Visa’s well-established infrastructure in the next five to ten years. The card networks are also evolving, and their size—3.2 billion cards, millions of merchants, and decades of trust—makes it difficult for any rival to undercut them. There is just no way to undervalue Visa and Mastercard based solely on economics, as one analyst put it bluntly.

However, disruption isn’t always motivated by economics. Occasionally, minor technical advancements add up to a tipping point that no one anticipated. Blockchain is catching up in speed. It’s still unclear if it will catch up quickly enough and whether it matters when it does.