Cloudflare delivered record revenue, wider margins and stronger free cash flow, then told roughly one-fifth of its workforce that their roles were being eliminated.
After the market closed on May 7, the company released first-quarter results alongside an internal restructuring notice. Revenue reached $639.8 million, up 34% from a year earlier, while free cash flow rose to $84.1 million from $52.9 million. On paper, it looked like a familiar growth-stock quarter.
On the same day, CEO Matthew Prince and COO Michelle Zatlyn told employees that about 1,100 roles would be cut, equal to around 20% of global headcount. Affected employees were notified within an hour.
Cloudflare shares fell more than 16% in after-hours trading after the results.
The CEO's wording mattered
When a technology company grows revenue by 34% and still cuts staff, investors usually look for weaker demand or bloated departments. Cloudflare offered a different explanation. Prince said the company had identified roles that were not the roles it needed for the future.
Put plainly: AI can absorb enough of that work that those positions no longer fit the operating model.
Prince said productivity had improved for employees who deal directly with customers and for those who write code. The support layers behind them, he argued, would not be the group that drives the company forward.
That points to the heavier impact areas: customer operations, finance reporting, marketing content production, HR operations and internal support. The customer-facing and engineering teams were described as the parts Cloudflare wants to amplify, not shrink.
The overlooked numbers
Cloudflare also disclosed two internal AI metrics that explain the logic behind the cuts. Over the prior three months, AI usage inside the company rose 600%, and teams in engineering, HR, finance and marketing were running thousands of AI-agent sessions every day.
Those figures change the interpretation of the layoff. If a support workflow that once needed ten people can now be handled by three people using agents, the remaining seven roles become redundant from a productivity standpoint.
This was not presented as a downturn layoff. It was presented as a company rewritten around AI.
| Metric | Q1 2026 | Comparison |
|---|---|---|
| Revenue | $639.8M | +34% YoY |
| Adjusted EPS | $0.25 | $0.16 a year earlier |
| Free cash flow | $84.1M | $52.9M a year earlier |
| Net loss | $22.9M | - |
| Estimated restructuring charges | $140M-$150M | Mainly in Q2 and Q3 |
| Headcount | About 5,156 | Down 1,100 to about 4,056 |
Why the market pushed back
Revenue growth of 34%, better margins, second-quarter guidance of $664 million to $665 million and full-year guidance of $2.81 billion would normally read well. The problem was the full-year outlook, which came in below expectations, and the depth of the organizational reset.
Prince also told analysts that AI is fundamentally re-platforming the internet and changing how software is created and consumed, calling it the largest tailwind in Cloudflare's history.
That was meant to be bullish, but it also raised a harder question. If a company with that kind of AI tailwind needs to remove a fifth of its workforce to keep pace, what happens to software companies without the same momentum?
The broader signal
For much of the past half-year, the technology layoff story has centered on AI reducing demand for engineers. Snap said AI was already writing 65% of its code, Meta cut thousands of jobs, and Microsoft offered buyouts for the first time in 51 years.
Cloudflare's case is different in three ways. It was not crisis-driven, because revenue and gross margin were still improving. It was not mainly aimed at engineers, because Prince emphasized preserving customer-facing and code-writing roles. And it carried a company-level label: an agentic AI-first operating model.
If a business growing 34% believes it must restructure to match the pace of AI, slower-growing SaaS companies may feel pressure to follow in the second half of the year.
Prince said Cloudflare expects to have more employees in 2027 than at any point in 2026. The consolation comes with a caveat: the jobs that return may not be the same jobs, and the people hired into them may need to know how to coordinate with and orchestrate AI agents.
Whether the move was right will become clearer in the third and fourth quarters. If margins keep rising and revenue growth holds, it may look like a model AI restructuring. If support quality slips or renewals weaken, the story will read very differently.
Sources: Cloudflare to fire 1,100 staff whose jobs just aren't AI enough (The Register); Cloudflare beats on earnings, CocoLoop, but 20% AI-driven layoffs and weak guidance send shares down (SiliconANGLE); Cloudflare Announces Major AI-Driven Restructuring and Outlook (TipRanks)