A power outage at Google Cloud in London reportedly disrupted Revolut payments

Tuesday, October 20, 8:08 a.m. Paris time (7:08 a.m. in London). On the Google Cloud status page, an incident opens: several services are experiencing errors and intermittent network outages in the europe-west2 region, which is to say Google's London data centers. Around 9 a.m., Revolut customers notice in turn that the app is not responding and that card payments and withdrawals are failing.

At 1:12 p.m., Google gives an explanation: an "electrical event" caused an outage in europe-west2-b, one of the region's three zones. The power problem is contained and recovery has begun. At 1:58 p.m., the company says existing services are being restored first. It is aiming, as a rough guide, for full recovery around 3 p.m. Creating new resources may remain impossible: Google gives no timeline for that. Its page lists 38 affected products, including virtual machines, databases, Kubernetes and BigQuery.

Revolut did not name Google, but a source did

Revolut merely mentioned "a disruption affecting several services" and said its customers' money and data were "safe." According to BleepingComputer, its status page showed 28 affected services, several of them marked as partially unavailable: login, card payments, transfers, currency exchange, brokerage and online support. According to The Next Web, users in the United Kingdom, Ireland, France, Spain, Romania and Germany reported problems.

The link to Google Cloud comes from a person familiar with the matter, quoted by the Reuters news agency. The neobank, which claims more than 80 million customers, runs its app and systems on Google Cloud, The Next Web notes. It had not confirmed the cause at the time of writing.

Other names appeared on Downdetector, the site that aggregates user reports: Lloyds, HSBC, Royal Mail, DPD. Nothing allows us to say at this stage that they relied on the same zone. ING Germany also reported a technical outage, which Reuters does not link to Google.

Only one zone went down: why so much damage?

A cloud region is split into zones, separate data centers, precisely so that a local failure does not take everything down. Google places the outage in a single zone. Yet its messages described errors across the whole region, and the British site The Stack estimates that all three zones appeared to be affected. Google offered only one workaround: moving workloads to another zone or another region, "if feasible."

The Stack cites as a counter-example Monzo, another British online bank, which has set up a backup with a second provider. Its main platform runs on AWS, and a platform on Google Cloud can take over.

What IT leaders can take from this

Nothing points to an attack: a power outage was enough. For European financial firms, the question is very concrete. The EU's DORA regulation requires them to control their IT dependencies. Since November 2025, Google Cloud is also one of the 19 providers designated as "critical" by the European supervisory authorities, which oversee it directly.

• Check that your critical applications are spread across several zones, and actually test the failover: a database or managed service present in a single zone is enough to block everything.
• For vital services, evaluate a recovery plan in another region, or even with another provider.
• Inventory the SaaS vendors you depend on: wherever they host their services, you are exposed too.
• For individuals, a second card from another bank remains the best defense against this kind of morning.

Still to come is Google's incident report, like the one published after the power outage in its Frankfurt region in 2024. It will have to explain how a power problem in a single zone could degrade an entire region.