Cloudflare deep dive: from $2M Series A in 2009 to $42B public company competing with AWS

Summary

Cloudflare built a $42 billion public company by doing the unglamorous infrastructure work that AWS ignores. CEO Matthew Prince is now publicly needling Vercel over it.

The company started in 2009 with a $2.1 million Series A to solve a concrete problem: websites were slow because traffic had to traverse the internet to origin servers, and DDoS attacks could knock anyone offline. Cloudflare positioned itself as a global content delivery network sitting between users and origin servers, caching static assets and filtering malicious traffic at the edge. The freemium model was aggressive—cheap or free at small scale, then tiered pricing as customers grew—and underpriced against AWS CloudFront and competitors.

Within two years, the company hit billions of page views and millions of domains, signaling early product-market fit. It raised a $20 million Series B in 2011 and a $50 million Series C from Fred Wilson at Union Square Ventures in 2012. By 2012, Cloudflare was valued near $1 billion. It had deployed 62 points of presence worldwide by 2015, a critical advantage for edge computing where latency matters and physics constrains speed.

Cloudflare expanded beyond pure CDN work. In 2013, it gained credibility by defending against the Spamhaus DDoS attack, then the largest in history. In 2014, it acquired StopTheHacker and CryptoSeal to deepen threat detection and encryption. It launched Project Galileo to protect vulnerable public-interest websites and made HTTPS free for all customers via Universal SSL, a move that undercut competitors while positioning Cloudflare as infrastructure with a public purpose.

The 2015 Series D raised $110 million from Capital G (Google, Microsoft, and Baidu together) and marked a shift toward broader infrastructure. Cloudflare expanded to over 30 data centers, introduced a web application firewall, and launched "I'm under attack mode" to counter layer-7 DDoS attacks. The 2016 acquisition of Eager, a platform company, seeded Cloudflare Workers—serverless code execution at the network edge. That move took Cloudflare beyond content delivery into application hosting, the terrain it now competes on with Vercel and Replit.

Cloudflare went public on September 13, 2019, at $15 per share, raising over $525 million. It had already launched Workers (serverless execution), Warp (a VPN mobility service), and begun acquiring developer-facing tools including S2 Systems (browser isolation) and Linktap (which became Cloudflare Pages). Revenue hit $431 million in 2020, $656 million in 2021, and surpassed $1 billion in 2022. The company launched D1 (a serverless SQL database) and Workers AI (on-edge AI workloads with Nvidia GPU partnerships), positioning itself as a full platform to run applications, not just serve them from origin.

Operational challenges emerged. Cloudbleed in 2016 was a security vulnerability that leaked data, though actual customer impact was limited. Outages and deployment errors occurred. Cloudflare navigated thorny content-moderation decisions, dropping the Daily Stormer and 8chan after public pressure and forcing itself to articulate a content policy after initially taking a purist infrastructure-provider stance.

Today Cloudflare is a $42 billion public company targeting $5 billion in annual revenue with profitability ambitions. It owns its own hardware and operates a distributed network rather than relying on AWS infrastructure the way Vercel does. That gives Cloudflare pricing power and the ability to compete with AWS itself on egress fees and bandwidth costs. It launched R2 (S3-compatible object storage without egress fees) explicitly as a migration path off AWS.

The public tension with Vercel centers on infrastructure. Prince argues that Cloudflare has the real infrastructure (62 data centers, owned hardware) and full-stack capabilities to compete on application hosting, while Vercel relies on AWS and has no independent compute footprint. Rauch said Cloudflare's Workers platform didn't work for an internal project. Prince is needling him publicly, though notably, Cloudflare's website does not include a "Cloudflare vs. Vercel" comparison page, suggesting either ongoing internal debate or strategic restraint.

Cloudflare did what most cloud companies avoid—it built infrastructure instead of outsourcing to AWS. From 2009 to now, that edge-network advantage compounded into a platform that runs full applications, not just delivering content. Vercel is faster to market and has better developer ergonomics, but Cloudflare can undercut it on price because it owns the hardware and poles. The market is large enough for both, but Prince's comments suggest he sees Vercel as overreaching into his territory without the infrastructure to back it.