Neobank vs Challenger Bank vs Digital Bank: A License Guide for Fintech Founders

What is the difference between a neobank, a challenger bank, and a digital bank?

The features people usually compare, like app-only access and low fees, no longer separate anyone. In the American Bankers Association's 2025 survey, 54% of US bank customers named their bank's mobile app as the way they bank most often. A traditional bank with a good app looks a lot like a neobank from the customer's side of the screen. The difference sits underneath, in who holds the license.

Open five tabs for these terms and you get five definitions. One says a challenger bank is licensed and a neobank is not. Another says a digital bank is the online arm of an existing bank. A third says challenger banks keep branches. Each is partly right, because each describes one market.

Here is where the terms come from:

• "Challenger bank" became popular in the UK. It described newer banks competing with the Big Four (Barclays, HSBC, Lloyds, and NatWest) after the 2008 financial crisis. It covers app-only banks like Monzo and Starling, but also Metro Bank, which launched in 2010 with high-street branches.

• "Neobank" is the more common term in the US, where most app-first account providers work through a partner bank instead of holding a charter of their own.

• "Digital bank" is a real license in some markets. Singapore awarded its first digital banking licenses in December 2020, paving the way for GXS Bank, MariBank, ANEXT Bank, and Green Link Digital Bank. Malaysia announced five successful digital bank license applicants in April 2022. In 2024, Hong Kong announced that its "virtual banks" would be renamed "licensed digital banks."

Using the word "bank" without a license has consequences. In 2021, Chime agreed to stop calling itself a bank after California's Department of Financial Protection and Innovation objected. Today its site says Chime "is a financial technology company, not a bank," with banking services provided by The Bancorp Bank and Stride Bank.

One note on vocabulary: in the US, a banking license is called a charter. This article uses "license" throughout, and in the US sections the two words mean the same thing.

Revolut shows both sides of the line.

• In the EU, the Bank of Lithuania granted Revolut a specialized bank license in December 2018 and a full banking license in December 2021.

• In the UK, Revolut operated as an EMI for years. It received a banking license with restrictions in July 2024 and entered mobilization, the phase in which a new bank builds out its operations under limits on deposits. It left mobilization with a full UK license on March 11, 2026, about 20 months later.

Same app, same brand, different category depending on the market and the year. A UK customer in 2023 was using an e-money account. A Lithuanian customer in 2022 was banking with a licensed bank. For both of them, the practical difference was the protection row in the table above.

So when someone on your board raises challenger bank vs neobank, the fastest honest answer is a question back: whose license?

Two forms of digital bank show up most often:

• An incumbent's digital brand. JPMorgan launched Chase UK in September 2021 as an app-only bank run by its UK banking entity. By the end of 2025, it had reached 2.8 million customers and over $35 billion in balances.

• A new digital-only bank with a digital license. GXS Bank in Singapore operates under a digital full bank license from MAS. It has no branches and was built for mobile from the start.

For product teams, the difference is inheritance. An incumbent's digital bank inherits a core banking system, a compliance function, and a deposit base. It also inherits the constraints that come with them: release cycles tied to the core, approval chains built for a branch network, and a brand that has to stay consistent across dozens of products. A neobank starts with a clean stack and speed, and it rents the license.

We saw the incumbent side of this in the Digital PUMB redesign for FUIB, an established Ukrainian bank. The visible job was the business internet banking site, including an online account-opening flow for entrepreneurs. The harder problem was the design system behind it, which had to work for this product and carry over to FUIB's other services. It did: the new corporate style now runs across other FUIB products.

That is typical of digital banking work inside an incumbent. The new channel has to launch fast and still fit everything the bank already runs. For most founders, the digital bank vs neobank question only becomes real when a bank is the one building.

Simon-Kucher's 2026 study counts 1.4 billion neobank accounts worldwide. More customers are satisfied with their primary neobank (52%) than with their large or national bank (40%). Yet neobanks earn about 5% of global retail banking revenue, and in the US the average neobank customer brings in $70 to $80 a year.

Two findings explain the gap. First, 72% of consumers bank with two or more providers, so a neobank account is often one of several a customer keeps. Second, most banking revenue comes from lending, and lending from deposits needs a balance sheet and a license.

That second point helps explain why some neobanks eventually seek their own banking licenses.

The distinction has a practical side. A neobank is usually a customer of other fintechs. A typical US neobank stack combines a partner bank, a card processor, a KYC provider, and often a BaaS platform and an open banking connector. The neobank is the layer the customer sees.

Every one of those layers is a place where something can break, which is what the next section is about.

This section is general information, not legal or financial advice.

Compliance: shared on paper, yours in practice

On a partner bank model, the bank holds the license and answers to regulators. In practice, a large share of the compliance work lands on you.

In 2024, US regulators issued enforcement orders against several banks involved in banking as a service, including Blue Ridge, Lineage, Sutton, Piermont, Mode Eleven, and Thread. The orders cited problems ranging from BSA/AML compliance to third-party oversight and risk management.

Expect a partner bank to push those requirements down to you by contract: KYC at onboarding, transaction monitoring, and audit trails a bank examiner can follow. In the business banking and payment platform for SMBs we built, identity verification runs through Persona, wired into onboarding so every account opens with a KYC record attached.

SoFi's CEO, Anthony Noto, gave the reason plainly when the charter was approved: it lets SoFi "lend at even more competitive interest rates." Deposits can fund loans more cheaply than wholesale borrowing, improving lending margins. That helps explain the revenue gap in the Simon-Kucher numbers above.

The path keeps moving. On September 26, 2026, Bloomberg and Sky News reported that Nubank was in early talks to acquire Monzo at a valuation of £8 billion to £10 billion. No deal had been announced in those reports. Whatever the outcome, the reported discussions illustrate the industry's consolidation potential: a Brazilian neobank pursuing a US banking charter looking at a UK bank that started as an app.

Our view: design for the possibility of changing banking partners from day one. If your roadmap includes a license, or even a second partner bank, a few decisions made at the start set the cost of that move:

• Maintain reliable customer-level ledger records. Keep an accurate record of balances and transactions, reconciled against the bank's records. That gives you the data needed for a migration, though balances and accounts still need to be transferred through the appropriate banking processes.

• Maintain access to customer data and KYC records. Store verification results in a form you can securely share with a new bank or regulator where permitted. A new partner may still require additional checks.

• Put partners behind your own API layer. Your app talks to your services, and your services talk to the bank, card processor, and KYC vendor. That can limit the changes needed when switching providers, though integrations and compliance workflows may still need updating.

Retrofitting these after launch is close to a rebuild. It is the part of fintech software development where early architecture choices cost the most to reverse.