The debit card is a commodity. It has been for three years. When Revolut launched, it felt genuinely disruptive. Real-time FX at interbank rates, a slick app, no hidden fees. Then Wise copied
The debit card is a commodity. It has been for three years.
When Revolut launched, it felt genuinely disruptive. Real-time FX at interbank rates, a slick app, no hidden fees. Then Wise copied the best parts. Then Monzo. Then N26. Then somewhere north of 250 others. The result is a market where differentiated features have a six-month shelf life before they become table stakes, and where the cost of acquiring a depositor has quietly climbed past the unit economics that justified the business in the first place.
The obvious historical analogy is the early 2000s internet services market. AOL, Earthlink, NetZero, and a dozen regional ISPs were all selling the same product: a pipe to the internet. The differentiation strategies they deployed, branded software, loyalty points, customer service, all collapsed eventually because the underlying service was undifferentiated. What ended that war wasn't a better ISP. It was the shift from access as the product to content and infrastructure as the product. Google didn't sell a better dial-up connection. It made the connection irrelevant by owning what happened inside it.
Neobanks are living through that exact dynamic right now. And the exit path looks the same.
The companies that survive this consolidation wave aren't the ones with the prettiest card design or the most viral referral mechanic. They're the ones that find an economic layer below the consumer product and become structurally indispensable to it.
Three places make that genuinely possible.
The first is corridor monopoly. Cross-border payment infrastructure is still deeply regional. Moving value from Lagos to Accra, or Nairobi to Kampala, costs an absurd amount relative to the actual settlement cost on modern rails. The neobank that doesn't just offer stablecoin payments but owns the liquidity relationship at both ends of a specific corridor builds a moat no app redesign can touch. Bitso didn't win the US-to-Mexico remittance corridor by being a better bank. It won by becoming the market maker. By 2024 it was handling $6.5 billion annually, roughly ten percent of that entire corridor's volume. That's infrastructure, not product. It's the same logic behind a settlement engine like CAPP: the value sits in owning what clears at both ends of a corridor, not in the wallet interface sitting on top of it.
The second is the settlement layer for non-human commerce. The agentic economy is the most underappreciated structural shift in financial infrastructure since the credit card. AI agents already spend money autonomously on API access, compute, data, and inference. The median transaction size sits around $0.31. Legacy payment rails charge $0.30 plus a percentage, which means traditional processors would consume nearly the entire value of most agentic transactions. The neobank that builds identity primitives and payment rails for machine-to-machine commerce captures a volume explosion no consumer product can touch. The x402 protocol, designed for sub-cent programmable micropayments, is an early signal of where this goes. First-movers here don't need depositors. They need API calls, and a settlement layer built for agents, not humans, from day one.
The third is the identity layer. Know Your Customer frameworks built around passports and utility bills are architecturally obsolete for both the global unbanked and the agentic economy. Zero-knowledge proofs, behavioral reputation systems, and Trusted Execution Environment attestations are emerging as the replacement primitives. "The real opportunity is not to digitize KYC as we know it, but to make trust portable, verifiable, and privacy-preserving," says Francis Berwa, founding member of zkPass.org. "Zero-knowledge technology changes the equation because a person can prove the fact that matters without surrendering the underlying data. For neobanks, that means identity can become infrastructure rather than another onboarding hurdle and potentially the foundation for entirely new financial markets." The neobank that owns identity and trust verification for a specific market, whether that's informal traders in West Africa or AI wallets executing micro-settlements across the internet, owns something no competitor can replicate by cloning a feature.
There's a version of this article that focuses on developed market neobanks, Bunq versus Revolut versus Monzo, and that's the wrong fight to analyze. Those markets are oversaturated, heavily regulated, and contested by incumbents with deep pockets and existing customer relationships. The regulatory moat alone in a single US state, a $2,500 license application, mandatory 1:1 reserve ratios, state-by-state fragmentation, commoditizes the business before it can differentiate.
Emerging markets present the opposite dynamic. More than 1.3 billion people globally lack bank accounts. The mobile money agent network across Africa already handles enormous cash-to-digital volume through 800,000-plus physical kiosks. The traditional banking system in these markets isn't a competitor. It's a vacancy. The neobank that builds settlement infrastructure into that existing agent network, converting physical cash into programmable stablecoins at scale, doesn't need to displace anything. It fills a gap that generates real transaction volume immediately. A gold settlement pilot moving USDT and XAUT between institutional buyers and a licensed partner in Ghana is a small, concrete version of exactly this: infrastructure slotting into a corridor that already has demand, not a new app asking people to change behavior.
That transaction volume is the metric that matters. Not deposits, not card activations, not monthly active users in a country where ten other apps offer the same product. Gross settlement volume through a specific corridor tells you whether you've built infrastructure or just another interface.
Stripe paid $1.1 billion for Bridge. Mastercard agreed to pay up to $1.8 billion for BVNK. These aren't acquisitions of consumer products. They're acquisitions of orchestration infrastructure, the routing layers that sit beneath the consumer experience and handle the actual settlement. The legacy payment networks aren't buying neobanks because they're afraid of the card design. They're buying the plumbing.
That tells you exactly where the value in this market is accumulating and exactly what the exit strategy looks like for any serious player. You're not building a bank. You're building a piece of the settlement stack that a Visa or Mastercard will eventually need to own, or that generates enough corridor-level transaction volume to be self-sustaining.
The neobanks that misread this and keep competing on consumer product features will spend the next three years acquiring customers at unsustainable cost, watching margins compress the whole way. The ones that pivot toward owning a specific economic primitive, a corridor, an identity layer, an agentic payment rail, will find the competitive dynamics look completely different. You can't clone infrastructure the same way you clone a feature.
The debit card era of neobanking is functionally over. What comes next is about who owns the rails underneath, and the corridors get claimed by whoever settles them first, not whoever has the nicer app on top.
