Why can one company's bad day take down so much of the web?

Because many sites sit behind the same providers, so one provider failing can look like the whole internet failing. The network itself is still decentralised: DNS, for example, was designed as a distributed and fault-tolerant service, made to avoid a single large central database. But a website's owner chooses services to rely on, and many choose the same ones.

Two events show the pattern. On 21 October 2016 three consecutive distributed denial-of-service attacks hit Dyn, a DNS provider, and major platforms and services became unavailable to large swathes of users in Europe and North America. The traffic came from tens of millions of IP addresses, believed to be a botnet of ordinary devices such as printers, IP cameras and baby monitors infected with Mirai malware. Then on 18 November 2025 Cloudflare, a company that sits between visitors and website hosts, suffered a major global outage. Twitter, Spotify, Uber, DoorDash, Indeed, Canva, Grindr and IKEA were among the affected services. Cloudflare blamed a change in a database configuration that caused an invalid file to be sent to all servers on its network.

The reach explains it. According to W3Techs, Cloudflare is used by approximately 21.3% of all websites as of January 2026. When a service that large fails, many unrelated sites fail together.