Cloudflare cut 20% of jobs, yet added 45% more engineers by December

Cloudflare cut 20% of its workforce earlier this year. At the same time, its engineering ranks jumped 45% from December, rising from 1,308 engineers in December to 1,894 engineers. BNP Paribas analysts spotted the contradiction by pulling data from LinkedIn, and Cloudflare CEO Matthew Prince confirmed the trend.

So why would a company slash so many jobs while ramping up engineering hiring? Prince’s answer is basically a thesis statement for the AI era: AI is changing the mix of roles inside companies. The headline number looks like a firing spree. The underlying story, Prince argues, is that the “measurer” function is shrinking, while core builders get more capacity.

To understand what is happening, Prince breaks modern companies into three broad groups. First are builders, who create products. Second are sellers, who persuade customers to buy them. Third are measurers, who track, audit, manage, report on, and coordinate the business. In his telling, builders remain safe in an AI world, and in fact could become even more valuable as AI improves productivity.

That matters because the AI conversation often gets reduced to a blunt question: “Will AI kill jobs?” Prince’s framing is more specific. He argues that AI is not a job-killer across the board, but a job reallocator. If Cloudflare’s builders become far more productive with AI, he says, the company will hire even more of them. This is the key twist: layoffs do not have to mean “less work.” They can mean “different work,” with measuring responsibilities increasingly handled by AI.

The data piece reinforces that you are not looking at an isolated Cloudflare quirk. Prince points to broader hiring signals: open tech roles posted by tech companies are up 14% so far this year, according to TrueUp, a tech jobs marketplace. Open hardware engineering roles have jumped 52%. That combination suggests AI might be changing job composition, not shutting down demand for technical talent. In other words, the labor market is getting re-sorted, not erased.

Where the axe landed inside Cloudflare earlier this year was in the measuring layer. Prince says the layoffs focused on middle managers. Operations shrank into a single group. Finance was consolidated through more automation. Marketing was slashed, and he specifically ties it to the reality that marketing can be “riddled with measurers” in many companies. The result, in his description, is fewer people spending time tracking, auditing, and coordinating, with more investment directed at people doing core work to build and run the business.

This is not just an org-chart story. It is also a governance and regulatory story in the making, because measurers are often the folks who help companies satisfy internal controls and external scrutiny. When AI takes over more tracking, auditing, and reporting “constantly and thoroughly,” the internal process changes. Companies can compress layers, but they also need to decide who remains accountable for outcomes, not just outputs. That is a quiet shift executives have to manage, even if the public narrative is only about headcount.

There is also a board-level implication. When layoffs are paired with engineering growth, it can look like management is making a rational, productivity-driven trade, not cutting because demand collapsed. Prince’s logic implies that builders can scale with AI support, while measurers can be consolidated as AI systems handle routine coordination. For leaders, the danger is confusing the numbers. The visible move is “20% fewer jobs.” The strategic signal is “we think AI will increase builder output and automate measurer work.” Boards that understand that distinction are better positioned to judge whether the transformation is disciplined or chaotic.

For any executive or operator watching their own company’s AI experiment, Cloudflare’s pattern provides a high-stakes framework. AI might not simply replace tasks. It might change which tasks a company can afford to keep human. Prince’s “builder, seller, measurer” model suggests the test is not whether jobs disappear. It is whether your company’s AI investments actually reduce the need for measurement and coordination while boosting the output of builders and preserving sales capacity. If it does, the next layoffs may come with fewer panic headlines and more engineering requisitions.