Neobanks have moved from experiment to business. Revolut’s 2025 annual report shows revenue up 46% to £4.5 billion, profit before tax of £1.7 billion and 68.3 million retail customers, and at the end of 2025 it operated as a licensed bank in 30 countries (Revolut Annual Report 2025). Monzo’s year to March 2026 brought £1.7 billion of revenue, £172.6 million of adjusted profit before tax and 15.2 million customers (Monzo, Year in review 2026).
For a founder, those numbers raise a practical question: what is the model made of, and which piece of it can you build first? This guide compares neobanks with traditional banks, shows where the money comes from, walks through the licence step, and gives an example run of our estimate for a payments app. It is general information, not legal or investment advice.
| Neobank | Traditional bank | |
|---|---|---|
| Where customers bank | App and web only | Branches, phone, app and web |
| Licence | Often starts on a partner’s licence or an e-money licence; the largest later get banking licences | Full banking licence |
| Revenue | Many small lines on one app: card payments, subscriptions, foreign exchange, interest on balances, wealth products | Lending, deposits and fees across many channels |
| Sign-up | In the app, with identity checks done on the phone | In a branch or online |
| Main fixed cost | Software, compliance and support | Branches and staff, plus software |
The licence row is narrowing: the largest neobanks now hold banking licences, as Revolut’s 30 countries show. For a founder, the difference that matters is the starting point. A neobank begins as an app on top of someone else’s licence, then adds products and licences one at a time.
Both reports split revenue into several small lines rather than one big one. Revolut’s 2025 revenue mix lists card payments, interest income, subscriptions, wealth, foreign exchange and other income. Monzo names five lines: current account balances, borrowing, payments, wealth (savings and investments), and subscriptions and other fees. Monzo also reports £183 of average revenue per active customer for the year and more than 1.6 million customers on paid subscriptions.
That is what “universal banking” means in practice: one app, many revenue lines, added in order. For your own product, decide the first line, the customer it serves and the line you add next.
Every product that holds or moves money runs under someone’s licence. In the UK, the Bank of England’s New Bank Start-up Unit describes “authorisation with restrictions”, also called mobilisation, aimed mainly at UK start-ups. During mobilisation a new bank may hold deposits of no more than £50,000 in total, and the period typically lasts no longer than 12 months. For a complete application, the PRA targets a decision in four months (Bank of England, New Bank Start-up Unit, updated 21 July 2026). Even Revolut went through it: its annual report says the PRA lifted the restrictions on its UK bank’s licence in March 2026, after the mobilisation period.
In the US, the FDIC reminds consumers that nonbank companies are never FDIC-insured, and that deposit insurance does not protect against the insolvency or bankruptcy of a nonbank company (FDIC, banking with third-party apps). For pass-through coverage at a partner bank, the FDIC says records must identify who owns the money and how much each person owns, so plan that ledger from the start.
Our guide to building a neobank app compares the three licence routes and lists the parts a first release needs. A licensing lawyer in each market has the final word.
You don’t need to start with a bank. Start with one money job on a licensed partner. On Koncierge, a payment platform for Japan, the payment provider holds the licences. We modelled the data and the API first, then built a matching engine that assigns each incoming bank deposit to the right customer, with each party seeing only its own data. On IPOPBOOKS, a bookkeeping service for US small businesses built on bank feeds, the product went live 10 weeks after kickoff.
A wallet-style first version usually needs onboarding with identity checks, a ledger, balances and transfers, limits and an audit log, a back-office panel, and the payment provider and identity check integrations. Our fintech app development page lists the questions to ask any team that builds this.
Our project estimate turns your brief into user stories. You confirm what is in scope, and every hour in the budget traces back to a story you kept. We check the hours the model suggests against what our team actually books on finished projects, then multiply by our published rate of $39-55 an hour. The unit economics then show how many customers pay for the build and in which month it is paid back. You can change any starting number and the math updates.
Here is an example run of our estimate’s math with sample inputs. The product: a wallet-style mobile app on a licensed payment provider, with onboarding and identity check flows, balance and transfers, an accounts and ledger API, security, audit log and limits, a back-office panel, and the payment provider and identity check vendor integrations. Every input is an assumption. Only the arithmetic is ours.
| Line | Type | Value |
|---|---|---|
| Budget range for the build | Output | $8,700 to $18,000 |
| Timeline for the build | Output | 4 weeks |
| Revenue a month per active user | Assumption | $6 |
| Margin | Assumption | 60% |
| Months a customer pays | Assumption | 18 |
| Cost to win a customer | Assumption | $25 |
| New customers a month | Assumption | 400 |
| Profit a month per customer | Output | $4 |
| Lifetime value of a customer | Output | $72 |
| Left per customer after the cost to win them | Output | $47 |
| Customers who pay for the build ($13,350, the budget midpoint) | Output | 285 |
| First month that earns more than it spends | Output | Month 7 |
| Month the build is paid back | Output | Month 13 |
The 4 weeks cover the build only. The licence or partner onboarding runs on the regulator’s and the partner’s clocks, so start them first. For comparison, across 95 agencies the median published MVP starting price is $13,234 and the median promised timeline is 6 to 12 weeks (Rocketech MVP Price Index 2026, H2 edition). Those are published prices, not what agencies charge after scoping.
To get the same view for your idea, run your idea through our project estimate. If you ask for the market view too, it suggests customer groups and gives each a TAM, SAM and SOM with a note on what the number covers. Treat them as hypotheses to test, not verified demand.
A neobank serves customers only through an app and the web, with no branches. Many start on a partner’s licence or an e-money licence and add their own banking licences later. A traditional bank has a full banking licence and serves customers through branches as well as digital channels.
From many small revenue lines. Revolut’s 2025 annual report lists card payments, interest income, subscriptions, wealth, foreign exchange and other income. Monzo names current account balances, borrowing, payments, wealth, and subscriptions and other fees.
Some do and some don’t. The largest now hold banking licences: Revolut operated as a licensed bank in 30 countries at the end of 2025. Many smaller neobanks run on a partner bank’s or an e-money institution’s licence. In the US, the FDIC notes that nonbank companies are never FDIC-insured themselves.
It depends on the scope you confirm. In an example run of our estimate’s math with sample inputs, a wallet-style payments app on a licensed provider came to $8,700 to $18,000 and 4 weeks for the build, before licence and partner work. Across 95 agencies, the median published MVP starting price is $13,234 (Rocketech MVP Price Index 2026, H2).
Carefully curated content for resourceful Devs, CTOs, and PMs. No spam.