The origin of Cloudflare is a masterclass in accidental innovation, born not from a desire to build a global internet infrastructure company, but from a desperate attempt to solve the seemingly intractable problem of email spam. In 2006, Matthew Prince and Michelle Zatlyn were students at Harvard Business School, working on a class project that eventually evolved into a startup called Project Honeytrap. The premise was simple but ambitious: create a distributed network of honeypot email addresses to track, map, and ultimately block the sources of internet spam. They recruited thousands of users to register fake email addresses on their websites; when spam arrived at these addresses, the system would trace the origin of the message and map the botnets responsible. The technology worked brilliantly, and the team secured funding to build a commercial email filtering service. But they quickly hit a fatal wall: the email filtering market was a race to the bottom, dominated by massive incumbents with virtually zero margins, and enterprise customers had absolutely no desire to switch to a new, unproven startup for their email security. The company was burning through cash, the product was failing to gain traction, and the board was preparing to pull the plug. The pivot that saved the company and birthed Cloudflare happened entirely by accident. While analyzing the traffic data from their honeypot network, Prince and Zatlyn noticed something strange: the same botnets that were sending spam were also probing the web servers of their users, looking for vulnerabilities to exploit and launching distributed denial-of-service attacks to take websites offline. They realized that the underlying network they had built to track email spam was actually a highly sophisticated, globally distributed proxy network that could intercept, inspect, and filter all web traffic, not just email. In a desperate Hail Mary, they scrapped the email filtering product and rebranded as Cloudflare, pivoting entirely to web performance and security. They offered website owners a radical proposition: change your DNS settings to route your traffic through Cloudflare’s network, and we will make your website load faster by caching it on our global servers, and we will stop DDoS attacks for free. The offer was so compelling that the product went viral almost overnight. When they launched at the TechCrunch Disrupt conference in 2010, the demand was so overwhelming that their servers literally caught fire—metaphorically and, in one case, literally—under the weight of the traffic. The early days were a chaotic scramble to keep the network online; the open-source software they were using was never designed to handle the scale of the global internet, and the servers were constantly crashing. It was this existential crisis that forced Lee Holloway, the company’s first engineer, to write the custom Linux kernel module known as L4Drop, the foundational technology that allows Cloudflare to process millions of packets per second without crashing. From that near-death experience, Cloudflare was born: a company that accidentally built the infrastructure of the modern internet while trying to stop spam.