Your neobank didn't choose Visa. Your country did.

September 29, 2026 · 5 min read · ← all posts · the dataset

Across the whole dataset, the card networks look evenly matched. 155 of the neobanks we track issue on Visa, 119 on Mastercard, 24 on both. A near-tie, the kind of split that suggests a genuine competitive market where each company weighed its options and roughly half went each way.

Then you group by region and the tie disappears completely.

Fourteen African neobanks issue on Visa; two on Mastercard. Twelve British ones issue on Mastercard; two on Visa. Those are the same product category, the same decade, the same investor base in several cases — and opposite answers, with almost no dissent inside either market.

A genuine product choice made independently 381 times does not produce that pattern. It produces something closer to the European column, which is 28–28.

What is actually being chosen

Almost none of these companies are network customers in the way the branding implies. A neobank without its own banking licence — 82 of the ones here run the partner-bank model — doesn't hold a network membership at all. Cards are issued through a licensed BIN sponsor: the institution that owns the card range, carries the scheme membership, and takes the regulatory responsibility for what gets issued on it.

BIN sponsors are not globally interchangeable. They are licensed in specific jurisdictions, and each one carries whichever scheme it is set up for. So the real sequence looks like this:

1. You decide which country you're launching in.
2. That narrows you to the handful of BIN sponsors licensed there and willing to take a startup programme.
3. Whichever you can actually sign with determines your network.

The network is the output of step three, not a decision in its own right. We map that layer separately in the infrastructure directory, because it is the part of a neobank that customers never see and that determines a surprising amount of what the product can do.

Why this matters to someone choosing an account

Mostly it doesn't, and that is worth saying plainly rather than manufacturing stakes. Both networks work nearly everywhere. If you are picking between two neobanks, the network on the front of the card is one of the least informative differences between them — far behind who holds your money, which licence sits underneath it, and what happens if the company fails.

It matters in three narrow cases:

• Regional acceptance edges. Some countries still have merchant bases that lean hard one way, and a traveller carrying two cards on the same network has less redundancy than they think.
• Scheme-level benefits — rental insurance, dispute handling, purchase protection — differ by network and tier, and they come from the scheme, not from the app.
• Sponsor concentration. If two of your accounts are issued through the same sponsor, an interruption at that sponsor takes out both at once. The network on the card is a weak but real signal of where that risk clusters.

That last one is the only reason we record the field at all. The card network is a visible fingerprint of an invisible dependency, and in an industry where three of our six dead neobanks died because the institution underneath them withdrew, the invisible dependencies are the ones worth being able to see.

The 66 with no card at all

One more number the regional table leaves out: 66 of the 381 neobanks here issue no card whatsoever. Almost all are self-custodial products where a card would require exactly the licensed intermediary the design exists to avoid. They are a reminder that the card is a distribution decision, not a definition — and that the fastest-growing part of this industry, by founding rate, is the part opting out of it.