This morning, people across Europe tried to pay for coffee, withdraw cash, and log into their bank accounts — and couldn’t. Not because their bank failed, but because a single Google Cloud region in London had a bad morning.
On October 20, 2026, Google’s service health dashboard reported “elevated errors and intermittent network failures” in its europe-west2 region (London), starting around 7:08 AM UK time. Initial reporting pointed to a power event as the trigger. At roughly the same time, British digital bank Revolut reported a widespread service outage — card payments, transfers, ATM withdrawals, and even customer sign-ins showed as partially unavailable on its status page. More than 1,500 Revolut users reported problems on Downdetector on Tuesday morning alone. A person familiar with the matter told Reuters the two events were directly connected.
It wasn’t just Revolut. Outage reports also spiked for Lloyds Bank and HSBC customers, and services including Royal Mail, DPD, Rightmove, and the budgeting app Emma were affected. Revolut told customers on X that “all our customers’ money and data are safe” and that its engineers had been working on the issue since that morning — but no full restoration time was confirmed in the initial reporting.
Google Trends tells the rest of the story: “Revolut” surged past 100,000 searches in the UK and Germany today, as millions of people searched to find out whether their money was safe.
This is the strongest tech-infrastructure story of the day — and if you run a business, it’s not someone else’s problem. Here’s why it matters to you, and the 7 lessons worth acting on this week.
1. What actually happened this morning
To keep it simple: Google Cloud runs its services from data centers grouped into “regions.” The London region (europe-west2) had infrastructure problems starting Monday evening US time, extending into Tuesday morning UK time. Services including BigQuery, App Engine, Cloud SQL, and Firestore were affected.
Revolut — an app-based bank with no physical branches — runs on Google Cloud. When the London region failed, its card payments, transfers, ATM withdrawals, and logins went partially offline across Europe. Google later said the issue was mitigated for most cloud customers, but a full Revolut service restoration had not been confirmed in the reporting available at the time of writing.
The key fact for business owners: nobody hacked anyone. No one did anything wrong. One region of one cloud provider had a power event, and the blast radius covered banks, delivery companies, and property sites across a continent.
2. Why one region broke so many businesses
Cloud computing feels abstract until it isn’t. Every website, payment system, booking tool, and customer portal you use lives on physical servers in physical buildings. “The cloud” is just someone else’s data center.
Many businesses — including very large ones — run their entire operation in a single cloud region, sometimes without realizing it. When that region has a problem, everything stops at once: the website, the checkout, the internal tools, the customer app. There is no fallback because there was never a plan for one.
That’s exactly what Europe watched happen today. The businesses that stayed up were the ones that had planned for a region failure before it happened — traffic shifting automatically, databases replicated elsewhere, critical services running in more than one place.
3. The real cost of downtime for a business
When Revolut’s cards stopped working, customers didn’t just wait politely. They complained on social media, checked Downdetector, flooded support, and questioned whether their money was safe. Revolut had to issue public reassurances within hours.
For your business, the math is simpler and harsher: if your checkout, booking form, or lead-generation site goes down during business hours, every minute is revenue you don’t get back. Customers who can’t pay you don’t bookmark your site for later — they buy from whoever works right now. And the trust damage lasts longer than the outage itself, because “their site was down when I tried to pay” is the story people tell.
A few hours of downtime during a busy season can quietly erase months of marketing spend. That’s the part nobody prices into their hosting bill — until it happens.
4. Lesson 1: Know where your business actually lives online
Ask yourself a simple question: if your hosting company’s main data center went dark right now, what would your customers see?
Most small and mid-sized businesses genuinely don’t know. Your website might sit on a single shared server in one city. Your payment processing might route through one provider’s one region. Your CRM, your email automation, your booking system — each is a dependency, and most owners have never mapped them.
This week’s homework is a one-page map: list every system your business depends on to take money and serve customers, note where each one is hosted, and identify which ones have no backup. That single page is the foundation of everything below.
5. Lesson 2: Single-provider, single-region hosting is the real risk
You don’t need to become a cloud architect. But you should know that modern hosting makes resilience affordable in ways it wasn’t five years ago:
- Content delivery networks (CDNs) can keep your site’s pages serving from cached copies even when your origin server struggles.
- Managed platforms can replicate your database across regions automatically.
- Multi-region failover — sending visitors to a healthy region when one fails — is a standard feature, not an enterprise luxury.
The question isn’t “should we spend a fortune on infrastructure.” It’s “does our current setup survive one bad morning in one data center?” For many businesses, the honest answer is no — and fixing it costs far less than one serious outage.
If your site or store runs on a single server with no redundancy, this is the week to change that conversation with your developer or agency.
6. Lesson 3: Your communication plan matters as much as your infrastructure
Notice what Revolut did right today: within hours, they posted publicly that customer money and data were safe, said engineers were working on it since that morning, and apologized. Their status page showed which services were affected. That didn’t fix anything technically — but it stopped panic.
Now imagine the same outage hitting your business with no status page, no pre-written customer message, no one assigned to post updates. Your customers find out from an error page, or from Twitter, or from a competitor.
Every business needs three things ready before an outage, not during one: a status page or a pinned update channel, a pre-approved holding message for customers, and one person responsible for publishing updates. Write them this week. You’ll thank yourself later.
7. Lesson 4: Automate your monitoring and failover before you need it
The businesses that recovered fastest today weren’t the ones with the biggest IT teams — they were the ones where machines did the first responding. Automated monitoring catches a failing region in seconds. Automated failover reroutes traffic without a human waking up. Automated alerts tell the right person immediately, with the right context.
This is where modern AI automation and systems integration earn their keep: connecting your hosting, your monitoring, your alerting, and your customer communication into one workflow that fires before your team has finished their coffee. A chatbot that tells customers “we’re aware of an issue and working on it” beats a silent error page. An integration that switches your checkout to a backup processor beats a dead payment form.
Resilience is no longer just infrastructure — it’s automation. The companies that treat it that way sleep better.
8. The 5 questions to ask your web team this week
Don’t file this story under “interesting” and move on. Send these five questions to whoever builds or hosts your website — your developer, your agency, or your IT person — this week:
- Which cloud provider and region does our website run in — and what happens if that region goes down?
- Do we have automated backups, and have we ever tested restoring from one?
- Is our checkout/payment flow dependent on a single provider with no fallback?
- Do we have uptime monitoring with instant alerts, or do we find out from customers?
- Do we have a customer communication plan (status page, holding message) ready for an outage?
If the answers are vague, that’s your answer. A short resilience audit — mapping your dependencies, testing your backups, and setting up monitoring and failover — is one of the highest-ROI projects a business can do, because it protects everything else you’ve built.
At AI Dev Expert, we build websites, integrations, and automation systems designed to keep working when things go wrong — multi-region hosting setups, automated monitoring and alerting, backup payment flows, and AI-powered customer communication that keeps your customers informed even during an incident. If today’s outage made you wonder how your business would hold up, let’s find out together before the next one.
Book a free consultation and we’ll audit your current setup’s resilience — no jargon, no pressure, just a clear picture of where you stand and what to fix first.