By Monzo activated its backup bank this week after its main platform failed. Customers could not use the app normally. The bank switched to Monzo Stand-in and kept core payments running.
People across the UK reported issues on Tuesday afternoon. They could not check balances or view transactions. Monzo posted on X that the app was not fully functional. Engineers investigated the problem.
A spokesperson said: “After identifying some issues, we instantly activated Monzo Stand-in – our fully independent backup bank. As a result customers can still make payments with their card, withdraw cash, freeze their card, send and receive bank transfers and more.”
This was the biggest live test of Stand-in so far. Unlike traditional disaster recovery, Monzo built Stand-in as a separate bank from scratch.
Why Monzo built a separate bank
Monzo runs its main platform on Amazon Web Services. Stand-in runs on Google Cloud Platform. The team planned this split on purpose.
Cloud providers now handle most hardware failures. Software bugs cause most long outages today. If you run the same code in two clouds, the same bug can take both down.
Monzo built two stacks that do similar jobs but share almost no code. Each stack has its own Kubernetes clusters, databases and queues. Each has its own card processor. Each connects to payment networks through its own data centres.
You can read Monzo’s explanation of its Stand-in architecture. The bank calls Stand-in a last resort. It provides an extra layer of defence.
Cost also matters. Monzo says Stand-in costs about 1% of the primary platform to run. Copying the whole bank would double costs. A minimal backup keeps bills low and covers what customers need most.
How the backup bank works
Stand-in does not copy the whole bank. It stores only what it needs for critical tasks. That includes balances, recent transactions, card details and payees.
A service called the Data Syncer listens to events from the primary platform. It pushes updates to Stand-in in real time. The data is eventually consistent. Monzo tracks any lag and alerts if it grows too large.
Tokenized data, like card PANs, follows a different path. Each platform encrypts the data with its own keys and exchanges it securely. This way, neither side holds the same encrypted material.
When Stand-in turns on, it makes its own decisions. If it approves a card payment, it logs the decision locally. It also writes an Advice to a durable queue. Later, the primary platform replays that Advice verbatim. The primary ledger remains the source of truth.
Sometimes Stand-in works from a slightly stale balance. In that rare case, a customer could slip into an unarranged overdraft. Monzo says it built controls for this and the risk is very low.
To avoid double-counting, Monzo uses a correlation ID. It merges the Stand-in view with the synced primary view. You can read more in Monzo’s breakdown of payment processing in Stand-in.
Switching happens via a config service that lives in both platforms. The Monzo app checks this config in the background. If Stand-in is active, the app shows a simplified UI. For payments, Monzo first proxies traffic through the primary platform to Stand-in. This gives fine-grained control over who moves. If the primary is fully down, Stand-in can connect straight to card networks and Faster Payments. Monzo tests both routes in production.
What still worked during the outage
During the outage, customers could still:
- Pay with their Monzo card online and in-store
- Withdraw cash from ATMs
- Freeze and unfreeze cards
- Send and receive bank transfers via Faster Payments
- Check a basic balance and recent transactions
The full app did not work. Pots, search and spending insights were offline. The limited API is intentional. It needs less data and reduces risk.
Monzo first turned on all of Stand-in for all customers in August 2024, when an outage lasted about an hour. Since then, Monzo has kept small groups on Stand-in for testing. That practice helped this week. One live report noted that Monzo told customers it had activated its backup bank to keep payments running.
Why backup systems matter now
Monzo now serves over 12 million customers. It has no branches. Even a short outage can stop people paying for food, travel or bills.
Regulators have tightened rules. In the UK, FCA guidance on operational resilience tells firms to define important services, set tolerance for disruption, and prove they can stay within it.
In Europe, the Digital Operational Resilience Act (DORA) now applies to banks and key tech suppliers. It pushes for better ICT risk management and testing. Monzo cited DORA when it described Stand-in. The bank argues that classic disaster recovery – running the same software in another zone – no longer works.
Other UK banks have struggled. A Treasury Committee inquiry found nine large UK banks logged at least 803 hours of outages in two years. That equals over 33 days. Barclays said it expects to pay £5m to £7.5m after a three-day outage where over half of online payments failed.
Monzo has spoken openly about its approach. At LeadDev LDX3 in 2025, engineers Andrew Lawson and Daniel Chatfield explained how Monzo built Stand-in to avoid costly outages. Their message was simple: design for failure.
What this outage shows
Stand-in does not prevent outages. The main platform still failed this week. It changes how an outage feels.
Card payments and transfers kept working. The full app did not. For most customers, that is the difference between inconvenience and real harm.
Monzo fixed the primary platform within hours and confirmed the app was back to normal. The team will run a full post-mortem. For now, the key point is clear: fast communication plus a working backup for essential payments matters more than promising zero downtime.