‘Humans will be a rounding error on the internet’ says Cloudflare exec

Cloudflare posted Q2 revenue of $696 million (+36% Y/Y) alongside losses of $205.7 million (more than 3x), but investors focused on the upside—shares rose 16% after hours to a new high and are up 68% YTD. CFO Thomas Seifert reiterated its machine-traffic thesis, forecasting non-human traffic could be up to 1,000x human in ~5 years, which he linked to AI-driven growth and rising security threats. Management also emphasized low planned capex at 14–15% of revenue (on $2.865–$2.87B forecast), arguing hyperscalers’ lower utilization makes commodity compute less attractive.

Analysis

NET is the clearest beneficiary because machine-to-machine traffic expands the addressable spend in front of the cloud, not inside it. The second-order effect is important: AI load tends to reward edge routing, bot mitigation, API security, and serverless orchestration, which are higher-uptake software layers than commodity compute; that supports NET’s revenue mix and should improve operating leverage if large-customer adoption keeps translating into usage.

Near term, the stock can keep squeezing higher because the market will likely underwrite a multi-quarter narrative shift from "cyber/CDN" to "AI traffic control plane." But the real catalyst path is 1-3 quarters, not the headline forecast: we need billings, net retention, and margin progression to confirm that traffic growth is monetizing rather than just inflating packet counts. If losses stay wide while growth merely holds, the multiple will be vulnerable after the initial post-earnings momentum fades.

Contrarian take: hyperscalers are not obvious losers. AMZN and GOOGL may actually capture more AI inference demand over 6-18 months, even if utilization is messy today, while Cloudflare’s monetization depends on customers paying for efficiency rather than simply consuming more bandwidth. The thesis breaks if AI traffic routes around NET’s edge or if enterprises optimize spend away from managed services; watch for any next-quarter guide that implies decelerating large-customer additions or weaker dollar-based expansion.

Key Decisions for Investors

• Buy NET on a 5-10% pullback over the next 1-2 weeks; target 15-20% upside into the next earnings cycle, with a stop if growth guidance falls below the current 30%+ pace or margins fail to improve.
• Use a 1-3 month NET call spread instead of outright stock if chasing the post-gap move; the setup is momentum-positive but valuation-sensitive, so defined risk is preferable.
• Consider a relative-value pair: long NET / short AMZN or GOOGL sized dollar-neutral for 1-3 months, betting that the market rewards the edge/security layer faster than it rewards incremental cloud inference.
• Do not short AMZN or GOOGL outright on this headline; if anything, treat them as monitoring targets for whether AI traffic growth is shifting spend to edge software or simply increasing cloud demand.
• Set an alert for the next quarterly print: if large-customer commentary does not show up in billings/RPO and margin trajectory, reduce NET exposure even if traffic anecdotes remain bullish.