Cloudflare beats Q1 forecasts, announces 20% layoffs; shares slump

Cloudflare beat Q1 expectations with adjusted EPS of $0.25 vs. $0.23 consensus and revenue of $639.8 million vs. $620.83 million, but shares fell more than 16% premarket after the company said revenue growth did not accelerate and it will cut over 1,100 jobs, or about 20% of staff. The company raised full-year 2026 guidance to $1.19-$1.20 in adjusted EPS and $2.805-$2.813 billion in revenue, but Q2 revenue guidance of $664-$665 million slightly missed estimates. Restructuring charges of $140-$150 million and the AI-driven workforce reduction are the key near-term negatives.

Analysis

The market is treating this as a quality-vs-growth reset, but the more important second-order effect is that NET is showing the margin/efficiency step-up that software peers will be pressured to emulate. If AI-driven productivity can support a workforce reduction while preserving sales capacity, investors will start underwriting a higher operating leverage path across infrastructure software names, which is bullish for profitability but bearish for headcount-heavy “growth at any cost” models.

The immediate loser is not just NET shareholders reacting to a softer near-term revenue trajectory; it is the cohort of premium-multiple software names that trade on accelerating growth narratives. A one-quarter deceleration can matter disproportionately when valuation already embeds strong re-acceleration, and the guide miss into the current quarter creates a window where sell-side models likely stay too high for several weeks. That can drive further multiple compression before any fundamental AI efficiency gains show up in reported numbers.

The contrarian read is that the workforce cut is being misread as defensive when it may actually be a signal of management confidence in long-cycle demand elasticity. If the company is preserving quota carriers and reinvesting in sales capacity next year, the near-term hit to sentiment could create an entry point ahead of a margin inflection in 2H and into 2026. The key risk is execution: if AI usage is not translating into faster product velocity or better net retention, the restructuring charge becomes a one-time cosmetic fix rather than a durable operating model shift.

For broader markets, this reinforces a bifurcation inside tech: AI beneficiaries with real productivity gains versus software vendors where AI is still mainly a narrative support. That dynamic should favor companies with clear evidence of incremental gross margin or opex leverage from AI, while punishing those whose AI story is still aspirational. Expect this to matter over the next 1-3 reporting cycles as investors demand proof, not positioning language.

Key Decisions for Investors

• Short NET on strength for a 2-6 week trade or buy put spreads into any post-gap bounce; thesis is further multiple compression as estimate revisions catch down before margins can re-rate. Risk: if management provides credible evidence of accelerating bookings or retention, the short squeezes quickly.
• Pair trade: long MSFT or GOOGL vs short a basket of high-multiple software names with slowing growth profiles, including NET on any rebound; expected outperformance window is 1-2 quarters as AI productivity becomes a gating factor for valuation. Risk/reward favors the long leg because the market is starting to pay for tangible AI monetization.
• For investors seeking a contrarian long, wait 1-2 weeks for guidedown fatigue and consider a starter long in NET only if the stock stabilizes above the post-earnings low and analysts stop cutting estimates; upside could be 15-20% on margin/re-rate recovery into the next print. Stop if revenue re-acceleration thesis is not confirmed by channel checks.
• Consider shorting a basket of software names with similar valuation/growth sensitivity around upcoming earnings, using NET as the template for a broader opex-efficiency trade; this is a 1-3 month relative-value setup. Hedge with a long in profitable mega-cap AI platforms to isolate valuation compression risk.