Cloudflare - Inflection Points Across All Acts, Expanding AI Moat, and Raising Long-Term Target to $370
Post Q2 2025 Reflection
Cloudflare delivered a very strong quarter (q2 2025), with multiple operating and financial metrics showing signs of inflection across its three core “Acts.” In addition, the company has taken a high-profile step toward Act Four—its ambition to create a new business model for content creators in the AI-driven web. This initiative seeks to establish a third monetization path, beyond subscriptions and advertising, by enabling direct compensation from AI platforms to publishers, representing a potentially transformative long-term growth opportunity.
Inflection Points Across the Business
After three years of decelerating top-line growth, revenue momentum turned upward, with growth modestly accelerating from 27% to 28% year-over-year, signaling improving demand trends. Dollar-based net retention (DBNR) also improved by 3 points sequentially—the first expansion in three years—reflecting both stronger upsell activity and reduced churn. The developer ecosystem continues to scale rapidly, with active Workers developers surpassing 3 million by the end of 2024, up ~50% year-over-year, highlighting growing traction for Workers AI. On the enterprise side, go-to-market initiatives under CRO Mark Anderson are delivering, as evidenced by the company’s first-ever $100M deal in Q4 2024 and a rising number of $1M+ contracts.
Financial Discipline and Strategic Investment
Non-GAAP operating margin remained strong at 14%, even as the company accelerated investments in sales, marketing, and R&D—partly via higher stock-based compensation (SBC in the mid-20% range of revenue). Given the stock’s sharp appreciation over the past year, dilution from SBC is projected to remain under 1% at current run rates. Capex has increased meaningfully to a ~$300M annualized run rate (16% of revenue) to support GPU and AI infrastructure build-out, yet free cash flow margin remains healthy at ~10%. Management continues to target 20%+ non-GAAP operating margin and 25%+ FCF margin longer term. The balance sheet is fortified by $4B in cash following the issuance of $2B in 0% convertible notes, providing ample flexibility for potential strategic acquisitions.
Core Competitive Advantage: Innovation Velocity and Structural Cost Advantage
CEO Matthew Prince frequently emphasizes two enduring pillars of Cloudflare’s business model: rapid product innovation and a structural cost advantage derived from its network architecture. The innovation pace is evident in the company’s rapid expansion from CDN and DDoS protection (Act One) into Zero Trust/Security (Act Two), developer services (Act Three), and now AI-driven monetization models (Act Four). The cost advantage, often underappreciated by the market, is rooted in two factors: (1) network elasticity and ingress/egress dynamics, and (2) high CPU/GPU utilization.
Network Structural Advantage
Cloudflare’s network architecture creates a structural cost advantage that is difficult for competitors to replicate. Unlike most peers, Cloudflare operates a highly efficient, globally distributed network where a large portion of traffic is served from cache, meaning outbound traffic volumes significantly exceed inbound. Since bandwidth costs are determined by the greater of ingress or egress, this design allows Cloudflare to absorb even large-scale DDoS attacks without materially increasing its cost base. The company’s ability to mitigate attacks without passing costs on to customers reflects both its scale and its architectural efficiency, reinforcing customer trust and competitive pricing power. This cost resilience—rooted in Act One (CDN/DDoS)—extends into Act Two (Zero Trust/Security), where its massive, scalable, and programmable network—claimed to have 100x the capacity of certain competitors like Zscaler—enables cost-effective Zero Trust deployment. By leveraging excess capacity, Cloudflare can efficiently handle both outbound traffic (via CDN) and inbound traffic (via DDoS protection and Zero Trust solutions), increasing overall infrastructure utilization and lowering the cost per unit of traffic. In Act Three (Applications), developer services such as Workers and Containers can be executed where capacity is cheapest and latency lowest, further enhancing performance and economics.
AI and Utilization Advantage
On the utilization front in the era of AI and agentic workloads, Cloudflare’s architecture offers a distinct advantage. Typical AI inference deployments suffer from sub-10% GPU utilization because query patterns are unpredictable, leading to costly idle capacity. Hyperscalers rent out machines, but optimization is left to the customer—driving inefficiency. In contrast, Cloudflare’s global network can aggregate and dynamically route AI workloads to where capacity is available, achieving peak CPU utilization rates around 70% today, with GPU averages in a similar range. This explains why Cloudflare’s capex requirements are far lower in dollar terms than those of hyperscalers. While Cloudflare is not competing in large language model training, its infrastructure is well-suited to the growing inference market. One prominent foundation model vendor has even publicly called out Cloudflare as a third-party subprocessor, underscoring its strategic relevance in competing for AI workloads—particularly in inference—against much larger hyperscale cloud providers.
Building on its structural cost advantage and industry-leading utilization rates, Cloudflare is well positioned to capitalize on AI in four distinct ways:
Internal Efficiency – Like many enterprises, Cloudflare is deploying AI internally to automate and optimize its own business processes. By running these functions on its own infrastructure, rather than relying on external providers, the company improves efficiency while showcasing the capability and cost-effectiveness of its platform.
Smarter Products – AI enhances the intelligence of Cloudflare’s performance and security offerings. Its machine learning–driven systems proactively identify and mitigate emerging threats that human analysts might miss, improving customer protection and strengthening product stickiness.
AI at the Edge – Workers’ serverless architecture, combined with GPUs deployed across Cloudflare’s global network, enables AI inference and agentic workflows to run as close to the end user as possible—reducing latency, meeting emerging regulatory requirements for data locality, and improving cost efficiency. Customers benefit from a paid-per-inference model that eliminates the costly idle capacity typical of hyperscalers, with some realizing >10x price-performance gains through AI Gateway and caching optimizations. This operational efficiency ties directly to Cloudflare’s broader utilization advantage, allowing the company to deliver competitive pricing while maintaining strong margins and lower capex.
Post-Search Economics – “Act Four” – Cloudflare has named this opportunity as Act Four, reflecting its potential to reshape the business model of the AI-driven web. For nearly three decades, the web’s economics were built on Google’s content-for-traffic deal, where publishers allowed indexing in exchange for user visits that could be monetized. That model is now collapsing as Google’s AI Overviews and answer boxes, along with tools like ChatGPT and Claude, increasingly deliver answers without sending traffic back to content creators—undermining their monetization potential. Cloudflare, which sits in front of more than 20% of the web and serves ~80% of the leading AI companies, is uniquely positioned to mediate this new reality. On “Content Independence Day” (July 1), Cloudflare and partners began blocking AI crawlers by default unless they compensate publishers, laying the groundwork for a new economic model where content owners are paid for the data that fuels AI. More strategically, the same infrastructure enabling payments from AI companies to publishers could facilitate transactions between AI agents themselves, paving the way for an Agentic Web. With over half of its dynamic traffic already flowing between APIs, Cloudflare is structurally positioned to provide the rails for this emerging economy, bridging content creators, AI platforms, and automated agent-to-agent commerce.
Finally, Cloudflare has launched a major new capability on the Workers platform in Jun 2025 —Container-enabled Workers—which represents the long-missing piece in its developer ecosystem. This feature bridges the gap between Cloudflare’s ultra-lightweight V8 isolate environment and the need to run heavier, language-agnostic workloads at the edge. Developers can now deploy custom container images—capable of running Python, Java, FFmpeg, or AI inference—seamlessly alongside Workers, eliminating the need to maintain separate infrastructure for resource-intensive components. The result is a unified, globally distributed, serverless full-stack environment that combines the low-latency scalability of Workers with the flexibility of traditional compute, all managed via the same Wrangler toolchain. This frictionless development experience removes a major adoption barrier, attracts new workloads and developer segments, and accelerates platform growth—building on a Workers developer base that already surpassed 3 million last year.
Taken together, these AI and developer platform initiatives leverage Cloudflare’s cost-efficient network, high utilization rates, and market position to capture growth from AI and the evolving internet stack at multiple layers—from internal productivity gains to low-latency inference, to redefining the economics of the web, to enabling full-stack edge applications.
As I have said before in my previous posts, Cloudflare is one of the best companies to invest in during the age of AI. With all the positive momentum across its business and strategic positioning, I am raising my target price from $200 to $370 per share by 2030, implying a $150 billion market cap. This target assumes a more optimistic scenario where Cloudflare accelerates to a 35% (CAGR) through 2030, reaching approximately $10 billion in revenue and $3 billion in free cash flow. The $150 billion valuation reflects a 50x multiple on FCF, and the $370 target price corresponds to an annualized total return of about 13% over the next five years.