Cloudflare Cuts 1,100 Jobs in Agentic AI Bet as Business Hits Multi-Year Highs Across Nearly Every Metric
Q1 2026 Earnings Call, May 7, 2026 — A landmark quarter overshadowed by the company's largest-ever workforce reduction
The Restructuring: 20% Headcount Reduction From a Position of Strength
Cloudflare dropped the most consequential news in its history alongside one of its strongest quarterly reports: the company is laying off more than 1,100 employees, representing approximately 20% of its total workforce. This is not a distress move. With revenue growing 34% year-over-year, free cash flow ahead of expectations, and sales productivity rising for the ninth consecutive quarter, Cloudflare is making a deliberate structural bet that the agentic AI era makes a large share of its existing organizational architecture obsolete.
CEO Matthew Prince was direct: "This isn't a cost-cutting exercise or an assessment of the individual's performance. It's about defining how a world-class, high-growth company operates and creates value in the agentic era." The cuts span all functions and geographies, with one explicit carve-out: quota-carrying sales capacity will not shrink — it will grow. CFO Thomas Seifert explained that as support ratios for account executives improve dramatically through AI, the same cost envelope now funds more customer-facing headcount. "We are freeing up dollars, and within the same spend envelope, you can deploy more quota-carrying capacity towards our market opportunity."
The financial cost is real. Severance and restructuring charges are expected to land between $140 million and $150 million for full year 2026, with roughly $40 million noncash and the bulk hitting in Q2. Importantly, Cloudflare held its full-year free cash flow outlook unchanged, guiding to approximately 25% to 30% of annual cash generation in Q2 and Q3 combined. The company's net headcount ended Q1 at approximately 5,500, meaning the post-restructuring organization will stand somewhere around 4,400 people.
The AI Productivity Inflection That Triggered the Decision
The restructuring is rooted in an internal transformation that Prince described as a genuine tipping point reached last November. Cloudflare's own usage of AI tools has surged more than 600% in the past three months alone. Within the R&D organization, 97% of engineers now use AI coding tools, and 100% of all contributions to production codebases are reviewed by autonomous AI agents. Prince called it "like going from a manual to an electric screwdriver," and said the company has "never seen a quarter-to-quarter increase in new co-generated bug squashed and technical backlog burn down like we did last quarter."
The internal framework underpinning this shift is something Cloudflare calls "Cloudflare OS" — an agentic workflow layer built on its own Workers platform, incorporating dozens of MCP servers, hundreds of centrally managed skill files, and an AI Gateway product that intelligently routes requests to the most appropriate model based on task complexity. Simpler tasks run on Cloudflare's own GPU infrastructure at near-zero marginal cost; more demanding tasks are routed to frontier models. Prince noted that CIOs of other companies, when shown Cloudflare OS in demos, immediately respond: "We want that, too." He suggested the company may increasingly commercialize elements of the internal tooling, consistent with its historical pattern of productizing what it builds for itself.
Q1 Results: A Clean Beat Across the Board
Revenue for Q1 2026 came in at $639.8 million, up 34% year-over-year. The large customer cohort — those paying more than $100,000 annually — grew 25% year-over-year to 4,416, and revenue from these customers reached 72% of total revenue, up from 69% a year ago. Deals over $1 million grew 73% year-over-year, the fastest pace since 2024. The $5 million-plus annual customer cohort grew 50% year-over-year, and the company added as many of these ultra-large customers in Q1 alone as it did in all of 2024. New customer bookings grew at the highest rate since 2023. New pipeline generation grew sequentially at the fastest pace in five years and exceeded internal targets by more than any Q1 since 2021.
Dollar-based net retention was 118%, down two points sequentially but up seven points year-over-year. Seifert flagged that the sequential dip partly reflects acceleration in new customer business, which carries lower initial retention metrics by definition, rather than any softening in the installed base. Gross retention reached its highest level in four years.
Operating income was $73.1 million, or an 11.4% operating margin. Free cash flow was $84.1 million, representing 13% of revenue, up from 11% a year ago. RPO grew 36% year-over-year to $2.543 billion, with current RPO up 34% year-over-year.
Gross Margin Under Pressure — But for Reasons That Cut Both Ways
Gross margin came in at 72.8%, down 210 basis points sequentially and 130 basis points year-over-year. Two dynamics are driving this compression, and neither is straightforwardly negative. First, paid traffic on the network continues to displace what was previously carried as free traffic expensed through sales and marketing. As Prince explained, Cloudflare's long-standing free tier attracted millions of developers, and as developer products now require at least some payment, that traffic is migrating to paid, pushing costs into cost of revenue. "It actually is a sign that more and more people are adopting the paid products," he said. Seifert was blunt: "While that is decreasing gross margin, it's literally a wash from an overall P&L perspective."
Second, the Workers developer platform — the fastest-growing part of the business — currently carries a lower gross margin than the corporate average. Developers on the platform reached 5.5 million at quarter end, with 1 million added in Q1 alone compared to 1.5 million added across all of 2025. Seifert acknowledged gross margin may continue to drift lower in the medium term but argued operating margin is increasingly the right lens. He pointed to a Rule of 40 figure already exceeding 46% with line of sight to surpassing 50% next year. Network CapEx was 9% of revenue in Q1 and is guided to 14-15% for the full year.
Agentic Traffic: Hundreds of Billions of Requests Per Month and Accelerating
The most strategically important data point Prince shared is that Cloudflare is now processing hundreds of billions of agentic requests per month, and that figure is growing exponentially. This is directly driving the company's core network business. "In this new world of agentic commerce, agentic transactions, our approach is showing the wisdom and durability" of focusing on API and application traffic rather than bandwidth-intensive video, Prince said, drawing a sharp contrast with traditional CDN peers. His point: an AI agent researching a product doesn't watch one video stream — it may visit 5,000 websites in the time a human visits five. Every one of those requests touches infrastructure, and a disproportionate share of that infrastructure is Cloudflare.
The GPU architecture underpinning this is worth understanding. While hyperscalers run GPU utilization in the single digits — a structural consequence of a business model built around leasing physical servers — Cloudflare is pushing GPU utilization toward the 70-80% range it achieves on CPUs, by running its fleet as a unified scheduler across mixed silicon. Each new server deployed now carries CPU, GPU, memory, storage, and network capacity in combination, and workloads are routed dynamically to the most appropriate compute. Prince framed this as a fundamental business model difference: Cloudflare invests behind demand rather than ahead of it, which is why it can serve inference at materially lower cost than hyperscalers.
One concrete signal of how fast agentic adoption is moving: Prince disclosed that a large, unnamed AI studio went from essentially zero Dynamic Workers deployments to over one million running on the platform within just 15 days of engaging with the product. Dynamic Workers — lightweight, near-instant sandboxes for ephemeral agentic tasks — is positioned as a direct alternative to containers, which Prince argued are too slow and too resource-intensive for the response times agentic workloads demand.
Act 4 — Monetizing AI Crawl Traffic — Is Real but Early
On the question of Act 4, Cloudflare's emerging business around controlling and monetizing AI bot traffic and crawler access to internet content, Prince was candid about the horizon. The company has moved from low penetration in the media sector to what he described as dominance — media company executives are telling Cloudflare they are negotiating better licensing deals with AI companies because Cloudflare has given them control over their content. That is the early lighthouse signal. The harder problem is extending that capability to the long tail of the internet — smaller publishers and content creators who cannot negotiate their own deals with frontier AI labs.
Prince sees the eventual solution as some form of micro-payment infrastructure for agentic requests, potentially fractions of fractions of a cent per transaction, but noted that current payment rails cannot handle the transaction volumes required. "Nobody can handle the volumes right now," he said. "We need something that's significantly larger than" what exists. He flagged this as one of Cloudflare's six stated corporate priorities for 2026 and expressed confidence the company would log its first revenue in this category this year. He also projected that by sometime in 2027, non-human traffic will surpass human traffic on the internet in volume — which creates both the urgency and the opportunity for this business to matter.
Second Quarter and Full Year 2026 Guidance
For Q2 2026, Cloudflare guided to revenue of $664 million to $665 million, representing 30% year-over-year growth. Operating income is expected in the range of $90 million to $91 million. For the full year, revenue guidance is $2.805 billion to $2.817 billion, also implying approximately 30% growth at the midpoint. Full-year operating income is guided to $418 million to $421 million, and diluted EPS is expected to be $1.19 to $1.20 on approximately 375 million shares. Free cash flow guidance remains unchanged. The deceleration in the revenue growth rate from Q1's 34% to a guided 30% is worth watching, though the company is absorbing significant organizational disruption in Q2 while explicitly protecting its sales-facing capacity.
Partner Channel and Sovereign Cloud as Emerging Levers
Partners now represent 30% of Cloudflare revenue, a figure that has been rising steadily. Prince attributed this to the go-to-market transformation initiated by President Mark Anderson two years ago, which pushed the company toward a more consultative, channel-led motion particularly for Zero Trust and SASE — products that require deep integration work. Seifert noted that pool-of-funds deals, now in their sixth quarter as a mainstream go-to-market tool, are renewing at Cloudflare's highest-ever gross retention rate, validating the thesis that committed spend contracts create stickiness rather than just pulling forward revenue.
On the sovereign and data-localization opportunity — a topic raised in the context of Fortinet's recent SASE commentary — Prince argued Cloudflare's presence in more than 120 countries and 350 cities, combined with its ability to confine data to specific jurisdictions at granularity no hyperscaler can match, positions it well as both regulatory requirements and practical enterprise risk management push organizations toward geographic data controls. He cited the self-service Zero Trust product as already benefiting from demand generated by OpenAI's ChatGPT-style tools driving individuals and developers to experiment with fine-grained data access controls — and bringing those habits into enterprise purchasing conversations.
Cloudflare, Inc. Deep Dive
The Agentic AI Paradigm Shift
The May 2026 earnings print for Cloudflare represented a watershed moment not just for the company, but for the broader enterprise software sector. While reporting a blistering 34% year-over-year revenue growth to $639.8 million in the first quarter, the company simultaneously executed the termination of approximately 1,100 employees, or 20% of its global workforce. This was decidedly not a distressed cost-cutting measure designed to appease short-term activists. Instead, co-founders Matthew Prince and Michelle Zatlyn executed a deliberate, structural pivot toward an agentic artificial intelligence operating model. Internal telemetry revealed a 600% increase in AI utilization within a three-month window, with core departments running thousands of autonomous AI agent sessions daily. Cloudflare has effectively decoupled headcount from revenue scale, establishing a new software engineering frontier where smaller, AI-augmented teams handle the operational velocity that previously required thousands of additional personnel. This immense structural leverage is expected to push the company's Rule of 40 metric past 46% by year-end, cementing a new baseline for what institutional investors will demand regarding enterprise software profitability.
Business Model: The Architecture of the Connectivity Cloud
Cloudflare operates an elegant, highly disruptive business model that has evolved entirely from a foundational content delivery network into a comprehensive connectivity cloud. The company monetizes by acting as a reverse proxy, intercepting and routing traffic between users and internet properties to improve performance and block malicious activity. Its monetization funnel famously begins with a generous free tier that attracts millions of domains. This massive, unmetered footprint acts as the ultimate threat intelligence sensor, feeding machine learning models that instantly inoculate the paying enterprise tier against zero-day vulnerabilities. Revenue is generated through a hybrid structure of consumption-based billing for developer compute and storage, alongside lucrative, multi-year subscription contracts for enterprise security and network bundles. Large enterprises pay for guaranteed service level agreements, advanced web application firewalls, bot management, and seamless global routing. By unifying network delivery, zero-trust security, and serverless compute into a single control plane, Cloudflare essentially replaces a dozen disjointed legacy IT appliances with a single cloud-native subscription.
Market Share, Customers, and The Competitive Theater
In the content delivery landscape, Cloudflare has systematically dismantled the oligopoly once tightly held by legacy providers. As of early 2026, Cloudflare controls approximately 22% of the global content delivery market by HTTP request volume, sitting shoulder-to-shoulder with Amazon CloudFront at 21% and trailing only Akamai Technologies, which holds roughly 31%. However, Cloudflare's customer mix has aggressively moved upmarket. The company now boasts 4,416 customers paying in excess of $100,000 annually, a cohort that grew 25% year-over-year and now contributes 72% of total revenue. The competitive dynamics vary drastically by product category. In the high-end streaming and legacy enterprise delivery space, Cloudflare battles Akamai, which leverages a massive legacy footprint of over 4,000 points of presence. In the programmable edge compute layer, Fastly is the primary rival, heavily favored by specialized developers who require instant cache purging and bespoke traffic control. Yet, Cloudflare's most critical battleground has shifted to the Secure Access Service Edge market. Here, Cloudflare Zero Trust competes fiercely against Zscaler and Palo Alto Networks. While Zscaler maintains a dominant position in pure-play strict application-level access with over 60% of legacy secure web gateway replacement deployments, Cloudflare is aggressively capturing market share by bundling low-latency network performance natively with its security service edge.
Competitive Moat: The Unified Edge and Egress Economics
The fundamental competitive advantage of Cloudflare lies in its homogeneous architectural philosophy: every product service runs on every server in over 330 cities across the globe. Unlike competitors who acquired discrete security, delivery, and compute companies and stitched them together, Cloudflare built its software stack from the ground up to be natively unified. This single-pass inspection architecture ensures that a request is routed, authenticated, scrubbed for threats, and processed with serverless logic within 50 milliseconds for 95% of the global internet population. Furthermore, Cloudflare is attacking the hyperscale cloud providers at their most vulnerable financial choke point: data egress fees. With the introduction of R2, an object storage service that charges absolutely zero egress fees, Cloudflare has created a powerful economic wedge. Enterprises facing exorbitant bandwidth penalties from Amazon Web Services or Google Cloud are increasingly moving their data gravity to Cloudflare's edge. This low-input-cost advantage is paired with immense scale; processing over 60 million HTTP requests per second grants Cloudflare unparalleled peering leverage with internet service providers, systematically lowering transit costs and defending a robust gross margin of nearly 73%.
Future Growth Vectors: Workers AI and SASE Convergence
The next phase of Cloudflare's growth is anchored in bringing artificial intelligence inference directly to the network edge. Through its Workers AI product, Cloudflare allows developers to run machine learning models, supported by partnerships with platforms like Leonardo.Ai and Deepgram, on graphics processing units deployed physically within its global data centers. This dramatically reduces latency for generative artificial intelligence applications and ensures compliance with increasingly strict regional data sovereignty laws by keeping data processing localized. Workers AI is rapidly transitioning from a developer experiment into a core enterprise revenue driver. Concurrently, the total addressable market for the Secure Access Service Edge is projected to expand from $12.2 billion in 2025 to nearly $39 billion by 2031. As organizations finally abandon legacy hub-and-spoke virtual private networks in favor of identity-aware, cloud-native access, Cloudflare is perfectly positioned. Its Cloudflare One platform natively embeds zero-trust network access without the latency degradation typically associated with routing remote workforce traffic through distant, centralized security appliances.
Opportunities and Threats in a Shifting Industry
While the macroeconomic tailwinds of edge compute and zero-trust security are exceptionally robust, Cloudflare faces tangible industry threats. The enterprise reality of 2026 is dominated by multi-CDN architectures. Massive media companies and global retailers actively route traffic dynamically across Cloudflare, Akamai, and Fastly to avoid vendor lock-in and optimize pure transit pricing. This commoditizes baseline content delivery, forcing Cloudflare to constantly upsell higher-margin edge logic and zero-trust features to maintain its stellar 118% dollar-based net retention rate. Additionally, Cloudflare must navigate the delicate balance of moving upmarket. Its historical reliance on a self-serve, inbound enterprise sales motion means that capturing massive Fortune 500 security contracts requires displacing deeply entrenched, white-glove networking incumbents like Cisco and Palo Alto Networks, which boast decades of legacy relationships with chief information officers.
New Entrants and Disruptive Technologies
Credible threats from well-capitalized new entrants are emerging in the hyper-specialized AI infrastructure space. Purpose-built serverless GPU cloud providers and specialized bare-metal AI inference startups are attempting to undercut generalized edge platforms like Cloudflare by offering lower-cost compute cycles tailored strictly for large language models. Furthermore, sovereign cloud challengers in Europe are gaining traction by offering hyper-localized, jurisdiction-locked edge platforms that appeal to defense and public sector clients. While none of these localized startups currently possess Cloudflare's resilient global anycast network or deep peering agreements, they represent a disruptive pricing threat to the nascent Workers AI segment and demand constant vigilance from Cloudflare's product engineering teams.
Management Track Record: Visionary but Unsentimental
The executive triad of Chief Executive Officer Matthew Prince, President Michelle Zatlyn, and Chief Financial Officer Thomas Seifert has executed a masterclass in scaling global internet infrastructure. Prince and Zatlyn have maintained a founder-led product velocity that is exceedingly rare for a company approaching a $100 billion market capitalization. Their strategic foresight has consistently been years ahead of consensus, from pioneering serverless compute at the edge to predicting the death of the corporate perimeter. Seifert has provided the necessary operational discipline, successfully steering the company toward durable cash flow generation, posting $84.1 million in free cash flow, representing a 13% margin, in the most recent quarter. However, management's recent decision to surgically eliminate one-fifth of the workforce in a period of 34% revenue growth proves they are intensely unsentimental. They view the corporate structure itself as a systems engineering problem to be optimized. While this ruthless adherence to efficiency delights institutional capital, it places immense execution pressure on the remaining workforce to implement a highly complex AI transformation flawlessly.
The Scorecard
Cloudflare occupies one of the most strategically defensible positions in modern enterprise software. By collapsing delivery, security, and compute into a seamlessly unified global network, the company is extracting massive economic rents from the internet's architectural shift toward the edge. The aggressive restructuring into an agentic AI operating model demonstrates a management team that is willing to break its own highly successful operational mold to capture unprecedented leverage. With enterprise large-customer revenue growing at 38% and a deep product pipeline of high-margin SASE and AI inference solutions, the fundamental growth engine is operating with clinical precision.
However, the transition from a beloved developer utility to an indispensable enterprise security behemoth puts Cloudflare squarely in the crosshairs of the industry's most entrenched giants. The execution risks associated with displacing Akamai in enterprise delivery and Zscaler in zero-trust architecture are non-trivial, particularly as sophisticated buyers implement multi-vendor strategies to avoid platform lock-in. Nevertheless, the combination of zero-egress data gravity, sub-50 millisecond global latency, and a wildly cash-generative core business makes this an incredibly formidable asset. Cloudflare has proven it is not merely participating in the re-platforming of the internet; it is actively pouring the concrete for its foundation.