A sharp Cloudflare disruption on Friday evening left large parts of the online economy stuttering, with users reporting they were locked out of household-name services including Bet365, Sky Bet and Uber Eats. The outage rippled across the UK and the US, where thousands of people reported failed logins, frozen apps, error screens and timeouts — the kind of sudden, widespread breakage that happens when a core layer of internet plumbing falters.
Users began flagging problems shortly after 6:30pm GMT on February 20, with complaint volumes accelerating fast as the issue spread across platforms that share the same delivery and security backbone. In the UK, reports tied to Bet365 alone peaked at 5,247 around 6:57pm. Cloudflare itself drew a surge of incident reports, with additional spikes recorded across services that rely on high availability and low-latency access, from betting to gaming to workplace tools.
What Cloudflare does and why outages spread so quickly
Cloudflare sits in the critical middle layer between users and websites. It helps route traffic, speed up page loads, block malicious attacks and keep services stable during heavy demand. In normal conditions, that means faster performance and better protection. During a disruption, it can mean entire clusters of websites suddenly become unreachable or behave inconsistently — loading for some people, failing for others, or breaking only on specific app functions like payments, authentication or live content updates.
Cloudflare says it supports a vast slice of the internet ecosystem, and the “single point of failure” anxiety returns every time a large provider has a wobble. When services depend on the same infrastructure, a single incident can look like dozens of unrelated brands going down at once. The experience for users is simple: pages don’t load, apps spin, and customer support queues fill instantly.
Which services were affected
As the incident unfolded, users listed a wide range of platforms showing symptoms of disruption. In the UK, Downdetector reports clustered around betting and consumer apps including Bet365, Sky Bet, Paddy Power and Betfair, alongside mainstream services such as Wikipedia, Uber and Uber Eats. Gaming communities also reported issues, with complaints spanning titles and services including Minecraft, Steam, Counter-Strike, Call of Duty, Overwatch 2 and more.
Work and productivity tools were not immune. Users flagged issues involving Workday and Microsoft 365, while broader chatter touched on cloud and infrastructure dependencies across the wider web. Not every report indicates a direct technical dependency — outages can cascade through login providers, embedded services, payment layers and routing changes — but the overall pattern pointed to a concentrated infrastructure shock rather than a single-company failure.
Bet365 confirms issues as peak-time traffic collides with disruption
Bet365 acknowledged problems publicly, telling customers it was aware of an issue affecting the website and app and that its technical team was working to restore normal service. The timing mattered. Friday evening is a heavy-use period for sports betting and live in-play markets, where seconds matter and demand ramps quickly. A disruption during a busy window can amplify user frustration, increase error rates and make recovery feel slower even after the underlying incident starts to stabilise.
For betting platforms, reliability is not just a nice-to-have — it is essential. Markets move, odds refresh constantly, and customers expect withdrawals, deposits and verification checks to work seamlessly. Any interruption can trigger a flood of failed transactions, delayed confirmations and customer service requests, even if the outage is short-lived.
What users typically see during a Cloudflare-style outage
When a major edge network struggles, the failure often looks messy and uneven. One user might load a site homepage but fail at checkout. Another may be stuck at login. Others might see a generic error page or repeated timeouts. Because traffic is distributed across regions and data centres, impact can vary by location, internet provider, device type and even whether a person is using a mobile network or home broadband.
That variability is why “the internet is down” can trend even when some people can still access the same service. It also explains why companies often urge customers to wait rather than repeatedly refreshing or reinstalling apps — repeated retries can add load at the worst possible time.
Infrastructure risk moves back into focus
The outage also revived a bigger conversation: modern online life is highly concentrated around a small number of infrastructure gatekeepers. Cloudflare is one of several companies that power huge volumes of global web traffic, alongside major cloud hosting providers and large-scale DNS services. When any one of those layers hits turbulence, the effects can spill far beyond a single brand.
Cybersecurity experts have long warned that internet routing and legacy design choices can create bottlenecks. The challenge for businesses is balancing performance and protection with resilience — spreading risk across providers, building fallback options and ensuring customers have clear updates when problems occur.
How to check status and what to watch next
During fast-moving incidents, the most reliable place to track updates is the provider’s own incident reporting. Cloudflare’s status page typically posts real-time notes as engineers identify affected systems and restore service. For consumers, the practical reality is that many impacted services begin to recover in waves — partial access first, then stabilisation of logins and payments, then normal performance.
For markets, widespread outages can briefly shift sentiment toward “digital infrastructure risk,” especially when the disruption hits large consumer brands at peak demand. If you’re tracking broader risk mood into the next data-heavy session, you may also like our coverage of market volatility and macro catalysts here: Nasdaq 100 futures swing near 24,850 as traders brace for key inflation data.
Cloudflare incidents tend to be short but loud, because the companies affected are the ones people use every day. Friday’s disruption was a reminder that even in 2026, the web can still hinge on a few critical pipes — and when one of them narrows, everyone feels it at once.
