Cloudflare CEO: How Did I Decide Which Employees to Replace with AI?

Author: Matthew Prince

Translated by TechFlow

Two weeks ago, I laid off more than 20% of our employees. I did not do this because Cloudflare is in trouble. Quite the opposite: our revenue growth hit an all-time high, our cash flow is robust, and the number of new customers we’re acquiring globally has never been higher. I made this decision because the business environment is undergoing dramatic change—and to win the future, Cloudflare must adapt accordingly.

Flip through the annals of U.S. corporate history, and you’ll be hard-pressed to find another publicly traded company that, while sustaining over 30% growth, simultaneously cut 20% of its workforce. Yet what we’ve done over the past two weeks may well become standard practice across the industry over the next year. This is a story about how artificial intelligence (AI) is reshaping everything—but unfortunately, many executives and commentators misunderstand how AI will disrupt business rules, and who will truly be affected.

To understand this, I revisited a classic book first published in 1954—yes, it’s twenty years older than I am: Peter Drucker’s *The Practice of Management*. Widely regarded as the “father of modern management,” Drucker meticulously dissected the diverse roles within organizations. I’ve grouped these roles into three categories: Builders, Sellers, and Measurers.

As their names suggest, Builders build products; Sellers sell them. Measurers, meanwhile, handle everything else: internal audit, revenue recognition (a financial term referring to the accounting process of determining when a sale can officially be recorded as revenue on the company’s books), finance, legal, compliance, middle management, day-to-day operations—the list goes on.

Contrary to pessimistic predictions from some analysts, Builders’ jobs are secure—and they’re going nowhere. If an engineer on my team can now achieve ten times the productivity thanks to AI, I’d hire every such talent I could possibly find.

Sellers likewise face little risk of displacement. Budget-holding humans remain the ultimate decision-makers—and they prefer buying from people who take time to listen, build trust, and step in to resolve issues when things go wrong.

Measurers are equally vital to any organization—but fundamentally different from Builders and Sellers. Top-tier Measurers are rare and invaluable. They work tirelessly behind the scenes, with no appetite for the spotlight (front-of-house role—like a restaurant server directly facing customers, i.e., a visible, publicly recognized position). Ideally, they also maintain objective, independent perspectives separate from other departments. As Drucker observed, measuring business performance matters—but customers are won through Building and Selling. A truly world-class company should allocate its greatest resources to these two core functions.

The AI wave isn’t coming for Builders or Sellers. It’s squarely aimed at Measurers. Tireless, absolutely impartial, highly efficient, and perpetually online—today’s AI systems can monitor and assess a company with an objectivity, granularity, and precision unmatched even by the most elite human professionals of the past.

At Cloudflare, for instance, our internal audit team used to sample only a handful of business risks each quarter. Now, we’re rolling out a new system that conducts continuous, real-time audits across every single business risk. Our financial close (closing our books—the routine process of reconciling accounts and producing financial statements at month-end or year-end) is faster. We make fewer mistakes—and when errors do occur, they’re caught more accurately and reliably. As CEO, I now have unprecedented tools—not just to precisely measure overall operational health, but also to identify rising stars across our teams with remarkable accuracy.

Most of the employees we laid off last week were Measurers. We streamlined middle management across the company, because AI enables managers to oversee more direct reports while still delivering precise performance evaluations and effective coaching. We consolidated scattered operations roles into a unified business support team—and where specialized expertise is needed, AI fills the gap. We also significantly reduced our marketing team—a traditional hotspot for Measurers, just like at most companies. And across our finance function, we identified abundant opportunities to consolidate roles and automate processes.

But the fundamental purpose of this reduction was never simply headcount trimming. In fact, we currently have more open job postings than ever before. I expect our total employee count to continue growing over the next several years. Precisely because Measuring no longer requires so many people, we now have the capacity—and the capital—to invest heavily in the talent that truly drives growth.

This summer, we received nearly one million applications for just 1,111 paid internships. The interns we selected are not only exceptionally talented—they’re AI-native (a generation that has grown up alongside AI, instinctively treating it as both a mindset and a foundational tool). Without exception, they are Builders or Sellers—and we expect the vast majority will receive full-time offers.

They represent the future generation that will invent entirely new ways to power our business. Thanks to AI, we can now measure their contributions more precisely—and identify tomorrow’s leaders with unerring accuracy. Far from being a harbinger of mass youth unemployment, AI is doing exactly the opposite.

AI won’t eliminate all jobs—but it will reshape every enterprise. Ultimately, time will prove Drucker right. AI will dramatically enhance our ability to measure organizational performance—freeing up the living, breathing humans on our teams to focus entirely on where they create and capture real value: Building and Selling.