Revenue grew 34% to $2.17 billion with 74% gross margins, and the edge network spanning 300+ cities creates a competitive advantage that hyperscalers cannot easily replicate.
Cloudflare sits at a peculiar intersection: too expensive for value investors, too small for mega-cap growth funds, and too infrastructure-focused for thematic AI portfolios. That positioning gap is the opportunity. At $66.9 billion market capitalisation and 31x revenue, Cloudflare trades at a premium that looks extreme until you understand what the company is actually building.
Revenue grew 34% to $2.17 billion in fiscal 2025. Gross margins of 74.4% are among the highest in infrastructure software. Free cash flow has turned positive and is accelerating, reaching $320 million in the most recent fiscal year. These are the early-stage economics of a platform company approaching an inflection point.
The thesis: Cloudflare's global edge network, spanning over 300 cities in more than 100 countries, positions it as the default infrastructure provider for AI inference workloads that need to run close to end users. Training happens in centralised data centres. Inference happens everywhere. Cloudflare is everywhere.
Cloudflare's competitive moat is its network topology. The company operates compute nodes within 50 milliseconds of 95% of the world's internet-connected population. Replicating this network would require billions of dollars in capital expenditure and years of peering agreements with internet service providers globally. Akamai, the nearest comparable in terms of network reach, took 25 years to build its footprint. Fastly, the closest pure-play competitor, operates in fewer than 90 cities.
The network was originally built for content delivery and DDoS protection. Those use cases remain the foundation, generating predictable recurring revenue from enterprises that cannot afford downtime. But the strategic significance of the network has expanded dramatically with the rise of AI inference.
Modern AI applications increasingly require real-time inference at the edge. A chatbot needs to respond in under 200 milliseconds. A fraud detection model needs to score transactions before they clear. An autonomous content moderation system needs to evaluate content before it reaches users. All of these workloads benefit from compute that is physically proximate to the end user, and Cloudflare's network provides exactly that.
Amazon's CloudFront, Google's Cloud CDN, and Microsoft's Azure CDN all offer edge capabilities, but they are attached to centralised cloud platforms with fundamentally different architectures. Cloudflare's edge-native design gives it a performance advantage for latency-sensitive workloads that the hyperscalers cannot match without rebuilding their network architectures from scratch.
TickerXray Report
Get the complete Cloudflare report with all 12 quantitative models, AI-generated investment thesis, and real-time data.
Cloudflare's product strategy follows a deliberate pattern: establish presence with a free or low-cost security product, then upsell compute, storage, and AI services to the installed base. The company calls this the "connectivity cloud" vision, and the execution has been methodical.
Workers, Cloudflare's serverless compute platform, allows developers to deploy code that runs at any of Cloudflare's edge locations. R2, the object storage service, competes with Amazon S3 on price by eliminating egress fees. D1, the edge database, enables applications to query data locally rather than round-tripping to a central cloud region. Together, these products create a complete development platform at the edge.
The AI inference products are the newest layer. Cloudflare Workers AI allows developers to run AI models at edge locations with usage-based pricing. The early traction is promising: AI-related product revenue is growing faster than the company average. The developer community, a leading indicator for Cloudflare, has embraced the edge AI tools because they solve a genuine problem: running inference cheaply and quickly without managing GPU infrastructure.
The playbook resembles AWS in its early years. AWS started with basic compute and storage (EC2 and S3), then expanded into dozens of services that created an ecosystem customers could not easily leave. Cloudflare is following the same expansion pattern but anchored at the edge rather than in centralised regions. AWS took eight years to reach $2 billion in revenue; Cloudflare reached $2.17 billion in roughly the same timeframe.
At 31x revenue and 154x forward earnings, Cloudflare is priced for perfection. There is no disputing that. The question is whether the growth trajectory and competitive position justify the premium.
The revenue growth rate of 34% places Cloudflare in the top decile of publicly traded software companies by growth. Companies growing at 30%+ with 74% gross margins have historically sustained premium valuations. CrowdStrike, Datadog, and Snowflake all traded at 25-40x revenue during their comparable growth phases. Cloudflare's multiple is elevated but within the historical range for companies at this growth and margin profile.
The path to profitability is becoming visible. Operating losses have narrowed from $250 million in 2022 to $200 million in fiscal 2025, despite revenue doubling over that period. The operating leverage is real: each incremental dollar of revenue requires less incremental spending because the network infrastructure is already deployed. As revenue scales toward $3-4 billion, operating margins should inflect toward 15-20%, producing $450-800 million in operating income.
At a 20% operating margin on $4 billion in revenue (achievable by 2028 at current growth rates), operating income would be $800 million. At 40x operating income, a reasonable multiple for a high-growth infrastructure platform, the market capitalisation would be $32 billion, below the current level. For the current valuation to be justified, the market needs to believe Cloudflare will sustain 25%+ growth and reach $6-8 billion in revenue, with operating margins expanding to 25%+.
That is a demanding set of assumptions. But Cloudflare's competitive positioning, with a global network that is functionally irreplaceable, makes it more achievable than for most software companies at this scale.
The instinctive bear case against Cloudflare is that AWS, Azure, and Google Cloud will simply add edge capabilities and crush a smaller competitor. This argument misunderstands network architecture.
Hyperscaler edge networks are extensions of centralised cloud platforms. Traffic flows into the edge node, gets processed, and frequently must round-trip back to a cloud region for data access, authentication, or state management. Cloudflare's architecture is fundamentally different: the edge node is the compute environment. Data, code, and inference models all run at the edge location, eliminating the round-trip latency that hyperscaler architectures inherently introduce.
The peering relationships further differentiate Cloudflare. The company peers directly with over 12,000 networks globally, including virtually every major ISP on earth. This interconnection fabric means Cloudflare can route traffic more efficiently than any single cloud provider, reducing latency and improving performance for global applications.
Fastly is the most direct competitor, but it operates at roughly one-sixth of Cloudflare's revenue with a narrower product portfolio. Akamai, the legacy CDN leader, generates higher revenue but is burdened by older architecture and declining growth in its core delivery business. Neither represents a credible threat to Cloudflare's trajectory.
The first catalyst is enterprise adoption acceleration. Cloudflare's large customer count (those spending over $100,000 annually) has been growing at 30%+ year-over-year. The enterprise segment generates higher revenue per customer, lower churn, and better upsell economics. As more enterprise workloads move to the edge, Cloudflare's revenue per customer should expand significantly.
The second catalyst is the AI inference market. Gartner estimates that AI inference will represent 60-70% of total AI compute spending by 2028, up from roughly 40% today. The shift from training to inference favours edge deployment, where cost per inference query is lower and latency is better. Cloudflare's Workers AI platform is positioned to capture a meaningful share of this transition.
The third catalyst is the zero trust security market. Cloudflare's SASE (Secure Access Service Edge) products compete with Zscaler and Palo Alto Networks in a market growing at 25% annually. The integration of security with the edge network creates a unified value proposition that standalone security vendors cannot match. Gartner has positioned Cloudflare as a leader in its SASE magic quadrant, validating the competitive position.
Valuation compression is the primary risk. If revenue growth decelerates below 25%, the 31x revenue multiple would be unsustainable, and the stock could de-rate significantly. Growth deceleration from 34% to 25% would represent a loss of approximately $200 million in annual revenue relative to current expectations, which at the current multiple translates to roughly $6 billion in market cap erosion.
Execution risk on AI inference is real. The edge AI market is nascent, and Cloudflare's Workers AI platform is competing against well-funded alternatives from Nvidia (Inference Microservices), AWS (Bedrock), and specialised startups. If the AI inference opportunity fails to materialise at the edge in the timeframe the market expects, the growth premium dissipates.
The profitability timeline could extend. If Cloudflare invests heavily in AI infrastructure and sales capacity, the path from negative operating income to 20%+ margins could take longer than the market currently expects. Patient investors will be rewarded; momentum investors will exit.
Cloudflare is expensive. There is no clever analytical framework that makes 31x revenue look cheap. But the company is building infrastructure that is genuinely difficult to replicate, growing at 34% with 74% gross margins, and positioned at the centre of two secular growth trends: edge computing and AI inference.
We see Cloudflare as a core holding for growth-oriented portfolios, with a two-to-three year time horizon. The near-term price target of $230-250 reflects our expectation that revenue will exceed $3 billion in fiscal 2027 with improving operating margins. The downside is meaningful if growth decelerates (a return to 20% growth could send the stock below $150), but the probability-weighted expected return favours the long side.
For investors with a longer time horizon, Cloudflare's edge network is a strategic asset that will appreciate in value as more of the world's compute moves closer to end users. We are buyers on pullbacks to $180-190, and holders at current levels.
Full forensic analysis of Cloudflare
+ 6 more models included
150,000+ stocks covered
Global coverage across 60+ exchanges. Every report includes all 12 quantitative models and AI analysis.
View plansEvery report runs 12 quantitative models and generates an AI investment thesis. From Piotroski scores to manipulation detection -- get the full picture in seconds.
12 forensic models
Piotroski, Altman, Beneish, DuPont & more
AI investment thesis
Synthesized outlook on every stock
Manipulation detection
Spot red flags before they hit the news
150,000+ tickers
Global coverage across 60+ exchanges
Expected return
Forward return projections for every stock
Real-time data
Live prices, insider trades, news sentiment
Free accounts get 1 report per month. Pro gets unlimited.
Free cash flow has moved from negative to $324 million in three years. Workers AI has become the reference compute layer for inference at the edge. The April 17 rally hinted at what the data has been saying for twelve months.
Oppenheimer calls the concerns 'overblown.' At 30x forward revenue with 35% growth and 78% gross margins, the edge network moat is being underpriced.
At 182x forward earnings and 34x revenue with negative operating margins, Cloudflare's $74B market cap requires a decade of flawless execution that even great companies rarely deliver.